Economic Analysis of Families and Society: The Transformation of Japanese Society and Public

Advances in Japanese Business and Economics 16

Shinji Yamashige

Economic Analysis of Families and SocietyThe Transformation of Japanese Society and Public Policies

Advances in Japanese Business and Economics

Volume 16

Editor in ChiefRYUZO SATOC.V. Starr Professor Emeritus of Economics, Stern School of Business,New York University

Senior EditorKAZUO MINOProfessor Emeritus, Kyoto University

Managing EditorHAJIME HORIProfessor Emeritus, Tohoku UniversityHIROSHI YOSHIKAWAProfessor, Rissho University; Professor Emeritus, The University of TokyoKUNIO ITOProfessor Emeritus, Hitotsubashi University

Editorial Board MembersTAKAHIRO FUJIMOTOProfessor, The University of TokyoYUZO HONDAProfessor Emeritus, Osaka University; Professor, Kansai UniversityTOSHIHIRO IHORIProfessor Emeritus, The University of Tokyo; Professor, National Graduate Institute for Policy Studies(GRIPS)TAKENORI INOKIProfessor Emeritus, Osaka University; Special University Professor, Aoyama Gakuin UniversityJOTA ISHIKAWAProfessor, Hitotsubashi UniversityKATSUHITO IWAIProfessor Emeritus, The University of Tokyo; Visiting Professor, International Christian UniversityMASAHIRO MATSUSHITAProfessor Emeritus, Aoyama Gakuin UniversityTAKASHI NEGISHIProfessor Emeritus, The University of Tokyo; Fellow, The Japan AcademyKIYOHIKO NISHIMURAProfessor, The University of TokyoTETSUJI OKAZAKIProfessor, The University of TokyoYOSHIYASU ONOProfessor, Osaka UniversityJUNJIRO SHINTAKUProfessor, The University of TokyoKOTARO SUZUMURAProfessor Emeritus, Hitotsubashi University; Fellow, The Japan Academy

Advances in Japanese Business and Economics showcases the research of Japanesescholars. Published in English, the series highlights for a global readership the uniqueperspectives of Japan’s most distinguished and emerging scholars of business andeconomics. It covers research of either theoretical or empirical nature, in bothauthored and edited volumes, regardless of the sub-discipline or geographical cover-age, including, but not limited to, such topics as macroeconomics, microeconomics,industrial relations, innovation, regional development, entrepreneurship, interna-tional trade, globalization, financial markets, technology management, and businessstrategy. At the same time, as a series of volumes written by Japanese scholars,it includes research on the issues of the Japanese economy, industry, managementpractice and policy, such as the economic policies and business innovations beforeand after the Japanese “bubble” burst in the 1990s.

Overseen by a panel of renowned scholars led by Editor-in-Chief ProfessorRyuzo Sato, the series endeavors to overcome a historical deficit in thedissemination of Japanese economic theory, research methodology, and analysis.

The volumes in the series contribute not only to a deeper understanding ofJapanese business and economics but to revealing underlying universal principles.

More information about this series at http://www.springer.com/series/11682

http://www.springer.com/series/11682

Shinji Yamashige

Economic Analysisof Families and SocietyThe Transformation of Japanese Societyand Public Policies

123

Shinji YamashigeSchool of International and Public PolicyHitotsubashi UniversityKunitachi, TokyoJapan

ISSN 2197-8859 ISSN 2197-8867 (electronic)Advances in Japanese Business and EconomicsISBN 978-4-431-55907-8 ISBN 978-4-431-55909-2 (eBook)https://doi.org/10.1007/978-4-431-55909-2

Library of Congress Control Number: 2017954470

The Work was first published in 2013 by University of Tokyo Press with the following title: Kazoku toShakai no Keizai Bunseki

© Springer Japan KK 2017This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or partof the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmissionor information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilarmethodology now known or hereafter developed.The use of general descriptive names, registered names, trademarks, service marks, etc. in thispublication does not imply, even in the absence of a specific statement, that such names are exempt fromthe relevant protective laws and regulations and therefore free for general use.The publisher, the authors and the editors are safe to assume that the advice and information in thisbook are believed to be true and accurate at the date of publication. Neither the publisher nor theauthors or the editors give a warranty, express or implied, with respect to the material contained herein orfor any errors or omissions that may have been made. The publisher remains neutral with regard tojurisdictional claims in published maps and institutional affiliations.

Printed on acid-free paper

This Springer imprint is published by Springer NatureThe registered company is Springer Japan KKThe registered company address is: Chiyoda First Bldg. East, 8-1 Nishi-Kanda Chiyoda-ku, 101-0065Tokyo, Japan

To Megumi

Preface

In 2011, the tsunami caused by the Great East Japan Earthquake swallowed ruraltowns in the north-east coast of Japan and took the lives of about 15,000 people. Italso caused the meltdown of the nuclear reactors at the Fukushima Daiichi nuclearpower plants. Japan has been recovering from the damages caused by the earth-quake and tsunami, but has to live with the aftermath of the nuclear meltdown for along time. It will take at least 30 years and U20 trillion to deal with the nucleardisaster.

Furthermore, two big earthquakes are expected to hit two urban areas of Japan inthe near future. An inland earthquake in the metropolitan Tokyo area and one in theNankai Trough in the Pacific Ocean are estimated to occur with 70% probabilitywithin the next 30 years, respectively (Fig. 1). As the two earthquakes are going tohit big cities, the damages caused by them and the resulting tsunami are estimatedto be far more than those of the Great East Japan Earthquake. The earthquake in theTokyo area can have a death total of 23,000 and cost around U95 trillion indamages, whereas the earthquake in the Nankai Trough, which has a very high riskof causing tsunami, is estimated to kill as many as 320,000 (Cabinet Office (2016)Disaster Management in Japan (http://www.bousai.go.jp/1info/pdf/saigaipanf_e.pdf)).

Given such a high probability of the earthquakes and tsunamis, the Japanesegovernment should have a good recovery plan to cope with these disasters.However, the Japanese government has already accumulated a huge public debt.The debt-to-GDP ratio now exceeds 200%, the highest among the OECD countries.When an earthquake hits Japan, how can the Japanese government, with furtherlosses in tax revenues, finance the recovery? Japan seems to have been enjoying theprosperity but its prosperity is built on very fragile grounds.

Although the land of Japan will not be sinking, the Japanese society is sinkingespecially with regard to its population. The Great East Japan Earthquake hit thecountry just after its population peaked in 2008 (c.f. Fig. 1.1 in Chap. 1). Thepopulation has been declining since then. Further, the elderly ratio, which is the

vii

http://www.bousai.go.jp/1info/pdf/saigaipanf_e.pdf

http://www.bousai.go.jp/1info/pdf/saigaipanf_e.pdf

ratio of elderly people above 65 years of age to the total population, is increasingrapidly because the number of the elderly people is projected to be increasing untilaround 2040. The elderly ratio is estimated to reach around 38% in 2050 andstagnate there throughout the rest of this century. This is the future of Japanesedemography that has been forecasted by the Japanese government.

Japan has a generous social security system and has maintained it so far byaccumulating public debt. The decline in younger generations, increase in thenumber of elderly people, and large accumulation of public debt all suggest thatJapan will not be able to sustain the current social security system. In the nearfuture, the Japanese government will face a financial collapse and need to downsizethe social security system to a large extent. When one of the two big earthquakes

Fig. 1 The Anticipated Big Earthquakes. Source: Cabinet Office of Japan (2016) DisasterManagement in Japan, and Wikipedia for a map at https://commons.wikimedia.org/wiki/File:Regions_and_Prefectures_of_Japan_2.svg

viii Preface

hits Japan, it will suffer from a further loss of resources and will face more seriousdifficulties.

The lack of tax revenues and an inability to borrow money with the huge publicdebt will push the country into big troubles (e.g., hyperinflation under the mone-tization of public debt and/or many tragedies under the forced cuts of the gov-ernment expenditure). The government may also face difficulty in finding youngpeople to rebuild the country because almost 40% of the Japanese would already beover 65 years of age by then. Japan will have to accept a large number of foreignworkers. A sudden increase in foreign workers will also create disorders in theJapanese society.

Based on various facts and scientific data, we expect that the prosperity enjoyedby Japan in the past 50 years will decay in the near future unless some unimag-inable events (e.g., big increase in tax burden, big cuts in social security benefits, orsudden decrease in the elderly ratio due to, for example, pandemics) or miracles(e.g., findings of oil/gas fields in Japan or sudden increase in the fertility rate)happen. The collapse may occur within the next several years and will certainlyoccur within the next few decades.

From a scientific point of view, an interesting research question is why Japan hascome to face these problems and has not been able to implement policies to mitigatethem. It was 1979 that Ezra Vogel published a book titled Japan as Number One,which praised the success of the Japanese economy. However, Japan is falling now.

In this book, we try to analyze the rise and fall of Japan before it really expe-riences the ultimate fall. There are two goals of writing such a book. First, we hopethat the book can make some contributions to minimizing the damages of thecollapse and increasing the possibility of its recovery from it. Second, even if it failsin achieving the first goal, we hope that this book contributes to the formulation ofbetter policies in other countries by analyzing how and why Japan failed toimplement public policies to make its society sustainable. The Japanese experiencewill be especially useful for developing countries that have not yet fully developedtheir social security system.

In this book, we pay attention to the roles of families and communities insustaining the society as they form the basis of our society. The most fundamentaland important viewpoint of this book is that public policies can transform familiesand communities. As public policies need to respond to changes in families andcommunities, there exists a cyclical relationship between public policies and fam-ilies and communities. Recognizing this relationship is really important because itsuggests that we need to formulate policies considering their side effects on familiesand communities. In a sense, this is a casebook that illustrates such a claim based onthe Japanese experience. We hope the case will be beneficial for other countries thatare facing or may face similar problems.

To analyze the impacts of public policies on families and communities, we willheavily rely on the economics of families and communities. Although they arerelatively new areas of economics, the economics of families has been growingfairly rapidly since the seminal works of Gary Becker who received the Nobel prizein economics in 1992. There are also many important economic analyses on

Preface ix

communities, which have shed a new light on our understanding of communities.The fact that Elinor Ostrom, who studied how common resources have beenmanaged and maintained in communities across the world, received the Nobel Prizein economics in 2009 together with Oliver Williamson who studied the functionsand structures of corporations, may suggest that “communities” are now wellrecognized in economics as important institutions in our society just likecorporations.

It may be worth noting that the development of the economic analyses offamilies and communities was brought about by the rapid development of gametheory. Most of the transactions in families and communities are not monetary andcould not be analyzed well with the traditional economics. By regarding suchtransactions as games among members of families and communities, economistsrealized that interesting and fruitful analyses can be conducted. Based on economicand game-theoretic analyses of families and communities, we examine the trans-formations and interactions of families, communities, and public policies in Japan.The book consists of three parts.

In Part I, we provide an overview of the transformation of the Japanese society,which occurred mainly in the twentieth century (Chap. 1). Then, we present ouranalytical framework to analyze the social transformation and provide the basicargument on how public policies should respond to such a transformation (Chap. 2).Finally, as a preparation for making in-depth analysis, we introduce readers to thebasics of decision theories (Chap. 3), including game theory, which plays importantroles in analyzing families and communities.

In Part II, we present the economic analyses of families and communities, whichis the core of this book. We first show the economic analyses of family formationssuch as marriages, births, and raising children (Chap. 4). Then, we discuss theeconomic models of resource allocations within families such as bequests and carebetween children and parents (Chap. 5). The second half of Part II is devoted to theintroduction of the economic analyses of communities. We first present the analysesof traditional communities, such as local communities, and see their internalinstability (Chap. 6). Then, we attempt to have a good understanding of new typesof communities, which are expected to substitute and complement the traditionalones. Concepts such as NPOs and social capital play important roles in under-standing the new communities (Chap. 7).

Given the analyses in Part I and Part II, we discuss in Part III how public policiesshould be formulated in response to the transformation of the Japanese society. Weaddress three social problems: population crisis (Chap. 8), poverty (Chap. 9), andregional disparities (Chap. 10). For each problem, we apply our economic analysesof families and communities and derive policy implications from them. Finally, inthe epilog (Chap. 11), we summarize our discussion, try to clarify reasons for whyJapan went wrong, and discuss lessons from the Japanese experiences for devel-oping countries. We argue that current problems in Japan can be explained by the“moral hazard problems” or “free rider problems” under the generous redistributivesystem (c.f. Remark 3.7 in Chap. 3) and the lack of policies to mitigate them.

x Preface

In reading the book, readers who are not interested in theoretical analyses canskip theoretical chapters, from Chap. 3 to 7, and read the rest of the book. In thediscussion on public policies in chapters from Chap. 8 to 11, we provide intuitiveexplanations for theoretical results to justify our analysis and proposals. It should bepossible for the readers to understand the essential arguments presented in this bookwithout having to read the theoretical chapters.

On the other hand, if readers are interested in how economics can be used toanalyze social transformation and public policies, we encourage them to read thetheoretical chapters. We hope that they can help students and researchers to see howeconomics can be useful in the analysis of various social problems and contribute tothe further development of economics.

In this book, we consider a family as a form of communities at times, and try tounderstand the relationship among markets, communities (including families), andthe government, for instance, in Part I. However, we often differentiate familiesfrom other types of communities and analyze them separately, for instance, inPart II. Families are very special forms of communities in most societies.Boundaries and roles of family members are usually defined clearly in legal terms,while other types of communities do not have such clear definitions.

The title of this book “Economics of Families and Society” indicates ouremphasis on “families”. Other types of communities are included in the term“society” which also includes the market system and the government (c.f. Fig. 11.3).From an academic perspective, our economic analyses and discussions on “com-munities” are something that can rarely be found in other books; and thus they can beconsidered as the special characteristics of this book. Although there are someeconomic analyses on communities, as we have argued above, the research in thisfield is still limited. As families and communities are very important elements in thediscussion of public policies, we hope that our book contributes to enriching theeconomic framework for analyzing public policies.

We wish that our analysis on the Japanese experiences will contribute to a betterunderstanding on public policies that have strong powers to change and determinethe future of our society.

Kunitachi, Japan Shinji Yamashige2017

Preface xi

Acknowledgements

The original version of this book was first published in Japanese in 2013. As theoriginal book was intended for the Japanese readers, I have modified the contents toenable readers who are not familiar with Japanese society and fiscal system. Thisversion will also help those who would like to learn from the Japanese experience.I would like to express my gratitude to the several reviewers who have providedcareful and constructive comments on my proposal for the English edition.

As the author of this book, I am now extremely proud and delighted to be able topublish this English edition. I am grateful to all of the people who are involved inpublishing this version. Especially, Dr. Ryuzo Sato, the chief editor in charge of theAdvances in Japanese Business and Economics Series, encouraged me to write thisbook. Ms. Juno Kawakami, the editor at Springer Japan, assisted me at variousstages and Mr. Mitsuharu Soshi, editor at University of Tokyo Press, supported mein publishing both the original Japanese and English editions. Finally, I would liketo express my sincere gratitude to all those who have greatly contributed to myresearch endeavor.

xiii

Contents

Part I Understanding Social Transformation

1 Transformation of the Japanese Society in the 20th Century . . . . . 31.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31.2 Transformation of Families and Communities . . . . . . . . . . . . . . 6

1.2.1 Population and Households . . . . . . . . . . . . . . . . . . . . . 61.2.2 Mutual Aid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

1.3 Economic Transformation in Japan . . . . . . . . . . . . . . . . . . . . . 121.3.1 Regional Disparity . . . . . . . . . . . . . . . . . . . . . . . . . . . 121.3.2 Income Inequality . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

1.4 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

2 Markets, Communities, and Government: AnalyticalFramework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232.2 Functions of Markets, Communities, and Governments . . . . . . . 24

2.2.1 Functions and Limits of Markets . . . . . . . . . . . . . . . . . 242.2.2 Roles and Limits of Governments . . . . . . . . . . . . . . . . 282.2.3 Functions and Limitations of Community . . . . . . . . . . 30

2.3 Interactions of Markets, Communities, and Governments . . . . . 342.3.1 Economic Theories Supporting Our Hypothesis . . . . . . 352.3.2 Evidences in History . . . . . . . . . . . . . . . . . . . . . . . . . 37


3 Introduction to Decision Theories . . . . . . . . . . . . . . . . . . . . . . . . . . 473.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 473.2 Utility Maximization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

3.2.1 Utility Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483.2.2 Utility Maximization Under Budget Constraint . . . . . . . 483.2.3 Decision Under Uncertainty . . . . . . . . . . . . . . . . . . . . 52

xv

3.3 Introduction to Game Theory . . . . . . . . . . . . . . . . . . . . . . . . . . 553.3.1 Simultaneous-Move Games . . . . . . . . . . . . . . . . . . . . . 563.3.2 Sequential Games . . . . . . . . . . . . . . . . . . . . . . . . . . . . 583.3.3 Repeated Games . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 613.3.4 Cooperative Games (Bargaining Games) . . . . . . . . . . . 633.3.5 Tragedy of the Commons . . . . . . . . . . . . . . . . . . . . . . 64


Part II Economics of Family and Community

4 Formation of Families . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 714.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 714.2 Children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72

4.2.1 Settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 724.2.2 Basic Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 734.2.3 Extension of the Basic Model . . . . . . . . . . . . . . . . . . . 774.2.4 Cobb-Douglas Case . . . . . . . . . . . . . . . . . . . . . . . . . . 79

4.3 Marriage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 824.3.1 Settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 834.3.2 Basic Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 834.3.3 Factors Affecting Divorce Probability . . . . . . . . . . . . . 864.3.4 Meaning of Legal Marriage and Divorce . . . . . . . . . . . 874.3.5 Related Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89


5 Resource Allocations within Families . . . . . . . . . . . . . . . . . . . . . . . 935.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 935.2 Gift and Bequest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94

5.2.1 Basic Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 945.2.2 Rotten-Kid Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . 955.2.3 Strategic Bequests . . . . . . . . . . . . . . . . . . . . . . . . . . . 975.2.4 Incomplete Information and Timing of Bequests . . . . . 100

5.3 Support and Care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1035.3.1 Basic Model for Private Support . . . . . . . . . . . . . . . . . 1045.3.2 Extended Model: Sustainability of the Private

Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1065.3.3 Public Support and Fertility Decisions . . . . . . . . . . . . . 108


xvi Contents

6 Traditional Communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1156.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1156.2 Private Provision of Public Goods . . . . . . . . . . . . . . . . . . . . . . 117

6.2.1 Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1176.2.2 Voluntary Provision of Public Goods . . . . . . . . . . . . . 1186.2.3 Crowding-Out Theorem . . . . . . . . . . . . . . . . . . . . . . . 1196.2.4 More Discussion on the Crowding-Out Effects . . . . . . . 122

6.3 Social Norm and Social Punishment . . . . . . . . . . . . . . . . . . . . 1246.3.1 Solutions to Externality Problems by

Communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1246.3.2 Social Norms in Communities . . . . . . . . . . . . . . . . . . . 1256.3.3 Punishment Mechanisms in Japanese Local

Communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1266.3.4 Evolution of the Social Norm . . . . . . . . . . . . . . . . . . . 128


7 New Communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1357.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1357.2 NPO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137

7.2.1 Concepts of NPO . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1377.2.2 Economic Analyses of NPO . . . . . . . . . . . . . . . . . . . . 1387.2.3 Basic Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1407.2.4 Management of NPO . . . . . . . . . . . . . . . . . . . . . . . . . 141

7.3 Social Capital . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1437.3.1 Concepts of Social Capital . . . . . . . . . . . . . . . . . . . . . 1447.3.2 Effects of Social Capital . . . . . . . . . . . . . . . . . . . . . . . 1467.3.3 Accumulation of Social Capital: Theoretical

Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1497.3.4 Accumulation of Social Capital: Empirical

Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1527.4 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154

Part III Social Policies in Japan

8 Population Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1598.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1598.2 Facts and Theories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

8.2.1 Population Crisis in Japan . . . . . . . . . . . . . . . . . . . . . . 1608.2.2 Theories for Understanding the Decline in

Fertility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161

Contents xvii

8.3 Social Security System and Family Policies in Japan . . . . . . . . 1668.3.1 Social Security System . . . . . . . . . . . . . . . . . . . . . . . . 1678.3.2 Family Policies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170


9 Poverty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1779.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1779.2 Theories and Facts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178

9.2.1 Theories for Understanding Poverty . . . . . . . . . . . . . . . 1789.2.2 Increasing Poverty in Japan . . . . . . . . . . . . . . . . . . . . 182

9.3 Public Assistance and Support for the Poor . . . . . . . . . . . . . . . 1859.3.1 Public Assistance Program . . . . . . . . . . . . . . . . . . . . . 1869.3.2 Support for the Poor . . . . . . . . . . . . . . . . . . . . . . . . . . 188


10 Regional Disparity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19510.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19510.2 Facts and Theories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197

10.2.1 Changes in Population Distribution . . . . . . . . . . . . . . . 19710.2.2 Theories for Understanding Regional Disparity . . . . . . 200

10.3 Fiscal Decentralization and Regional Policies . . . . . . . . . . . . . . 20210.3.1 Intergovernmental Relationship . . . . . . . . . . . . . . . . . . 20310.3.2 Regional Policies . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206


11 Epilogue: Social Transformation and Public Policies . . . . . . . . . . . 21311.1 Social Transformation, Public Policies, and Social Norms . . . . . 21311.2 What Went Wrong in Japan? . . . . . . . . . . . . . . . . . . . . . . . . . . 21411.3 Lessons for Developing Countries . . . . . . . . . . . . . . . . . . . . . . 21911.4 Public Policies in the 21st Century . . . . . . . . . . . . . . . . . . . . . 221References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225

xviii Contents

Acronyms

EITC Earned Income Tax CreditGDP Gross Domestic ProductGNP Gross National ProductLDP Liberal Democratic PartiesMOF Ministry of FinanceMOHA Ministry of Home AffairsNGO Non-Governmental OrganizationNIPSSR National Institute of Population and Social Security ResearchNPO Non-Profit or Not-for-Profit OrganizationOECD Organisation for Economic Co-operation and DevelopmentTFR Total Fertility RateVAT Value-Added Tax

xix

Part IUnderstanding Social Transformation

Chapter 1Transformation of the Japanese Societyin the 20th Century

The more you know about the past, the better prepared you arefor the future.

Theodore Roosevelt

1.1 Introduction

To understand the basic nature of the Japanese society in the 20th century, it wouldbe most useful to study the long-run trends of the Japanese population in Fig. 1.1. Inthe 20th century, the population was almost tripled, from about 44 million in 1900 toabout 127 million in 2000. However, such a rapid increase will be almost completelyoffset by the decline in the 21st century. It is estimated that the population will beless than 60 million in 2100.1

Furthermore, Fig. 1.1 indicates that the elderly ratio, which is the ratio of peopleabove 65years of age to the total population, increased slowly in the 20th century andpicked up its pace in the 21st century. It is estimated to reach about 38%by 2050. Thisbig demographic transition has been mainly caused by the change in fertility ratesin Japan in the 20th century, reflecting the transformation of the Japanese societyduring this period.

1The population estimation referred to in this book is the one estimated in 2016 (themiddle casefor fertility and mortality) by the National Institute of Population and Social Security Research,NIPSSR hereafter.

Theodore Roosevelt (1858–1919) served as the 26th President of the U.S. from 1901 to 1909.After having served as the President, he formed the “Progressive Party” and advocated theadoption of social insurance in the U.S., which had already been adopted by many Europeancountries, to protect workers and prevent the spread of socialism. The party won 27% of thepopular vote, the highest third-party vote in American history. Theodore Roosevelt’s vision wasrealized by his cousin, Franklin Roosevelt, in 1935 when the Social Security Act was passedby the Congress as a part of the “New Deal” policies against the Great Depression.

© Springer Japan KK 2017S. Yamashige, Economic Analysis of Families and Society, Advances in JapaneseBusiness and Economics 16, https://doi.org/10.1007/978-4-431-55909-2_1

3

4 1 Transformation of the Japanese Society in the 20th Century

(a) (b)

Fig. 1.1 Demographic transition in Japan. Source: Data published by NIPSSR

Fig. 1.2 Trend of the real GNP. Source: Ohkawa et al. (1974), Economic Planning Agency (2000),National Account (2015)

The social transformation of Japan is in tandemwith its economic transformation.Figure 1.2 indicates the trend of the real GNP (Gross National Product) since 1885.2

The economic transformation of Japan in terms ofmaterial affluencewas as impactfulas its social one in the 20th century, as can be observed in Fig. 1.1.

The social and economic changes after 1945, that is, after the end of the SecondWorld War, were remarkable.3 When we try to understand the transformation of the

2The data from 1999 represent real GNI (Gross National Income), which is equivalent to the realGNP in the old National Account System.3See, for example, Flath (2000) for a good introduction to the development of Japanese economy.

1.1 Introduction 5

Fig. 1.3 Trends of the public debt. Source: Hayashi et al. (2001) and the government statistics

Japanese society, it is necessary to focus on the year 1973, known as the “first yearof the welfare state” in Japan. It was the year when the government declared Japanas a welfare state. Furthermore, it was the year when the first oil shock hit the worldeconomy and ended the rapid growth of the Japanese economy (c.f. Fig. 1.24).

After 1975, the Japanese government had to rely on deficit financing to tacklethe problems of the oil shock, slower economic growth, and the expansion of socialsecurity payments. As a result of its rapid economic growth, the level of Japan’sgovernment bonds was low at that time. Hence, the Japanese government was able toissue a large amount of these bonds. The increase in government expenditure offsetthe decline in the private demand caused by the tight monetary policy for restrainingthe high inflation rate after the oil shock.

The successful policy mix of the Japanese government and the private companies’efforts for technological improvement allowed its economy to stand out in the world.Most of other developed countries suffered from the slow recovery of the first oilshock in 1973 and the second oil shock in 1979. The book Japan as Number One(Vogel 1979) nicely summarizes how theworldwitnessed the success of the Japaneseeconomy during this period. Although the Japanese economy was enjoying goodfortune in the 1980s, the domesticmarketswere actually sustainedbydeficit financingof public expenditure. After the bubble economy, which grew in the late 1980s, burstin 1989, the Japanese economy suddenly exposed its weakness. Despite the bigincrease in the public deficit, the Japanese economy could not recover from the longrecession in the 1990s and the 2000s.

Figure 1.3 indicates that in the last quarter of the 20th century, there was a hugeaccumulation of public debts under government policies. Even today, as of 2016, thepublic debt is issued to financemore than a third of the national government’s budget,and the general government’s long-term debt has exceeded 200% of the GDP.5

4In Fig. 1.2, the slope of the curve labelled as “log GNP” represents the economic growth rate. It iskinked around 1973, which implies that the economic growth rate has fallen since then.5The data in Fig. 1.3 only accounts for the national and local public debts.


The high level of public debt will become a big burden for the Japanese gov-ernment. The social security payment for the elderly is expected to grow further.Furthermore, the metropolitan Tokyo area and the Pacific Ocean area are expectedto be hit by two big earthquakes, each of which has a 70% probability of occurringwithin the next 30years. Moreover, the decline in the population of young people,who are the main contributors to the government budget, is estimated to continue(c.f. Fig. 1.1b). If it does not stop, a population crisis as well as a fiscal crisis isinevitable.

Needless to say, these problemsdid not occur in one day. The Japanese governmentcould have avoided them if it had implemented policies to balance the budget andslow down population ageing. To identify public policies for overcoming such crises,we need to ask what forces initiated them and why the Japanese government failedto mitigate them.

In the rest of this chapter, we look into the socio-economic changes that occurredmainly in the 20th century to consider the transformation of the Japanese societyand its public policies. In Sect. 1.2, we first discuss the social changes in familiesand local communities from the viewpoint of population, households, and mutualaids. Then, in Sect. 1.3, we provide an overview of the economic changes in Japan,which have a close relationship with its social changes, that is, regional disparity andincome inequality. Section1.4 concludes the chapter.

1.2 Transformation of Families and Communities

To understand the transformation of the Japanese society, themost important changesto be considered are the ones that occurred in families and local communities. Theyare closely related with the economic changes that we discuss in the next section.

1.2.1 Population and Households

As Fig. 1.1a indicates, the population in Japan was very small in the old days. Hence,Japan’s going back to a small country again may not be seen as a big problem.However, the process of shrinking and ageing of the population leads to many social,economic, and fiscal problems. To overcome such problems, it is important to identifythe reasons behind the demographic change.

The fundamental reason for the demographic change is the decline in fertility ratein Japan. Figure 1.4 depicts the trend of the total fertility rate, the average numberof children that a woman will have in her lifetime.

In Fig. 1.4, the actual total fertility rate (hereafter TFR) is depicted by the realcurve. It indicates a steady decline after 1973, “the first year of the welfare state” inJapan. Each broken curve represents a government’s estimate for the future fertilityrates in Japan published every five years. These broken curves indicate that until

1.2 Transformation of Families and Communities 7

Fig. 1.4 Trend in total fertility rate. Source: Data published by NIPSSR

2001, the government had optimistic forecasts about the future fertility rates. Suchsystematic forecast errors suggest that the models to estimate future demographyhave serious defects.6

It is well recognized that one of the reasons for the steady decline in Japan’s TFRafter 1973 is the decline in the marriage rate.7 As Fig. 1.5 indicates, the marriagerate started to decline rapidly after 1973, which remains low until today.

Figure 1.5 indicates that the divorce rate has been gradually increasing since the1960s. It is estimated that one third of themarried coupleswill eventually get divorced(Raymo et al. 2004). The figures suggest that the desire for family formation hasbeen declining, especially, after 1973, when the government declared the expansionof social security due to the rapid economic growth in the 1960s.

1.2.2 Mutual Aid

In this book,we define “community” as a network of the peoplewho are notmotivatedto seek profits. In such a community, we observe activities in which various resourcesare exchanged with reciprocity. We call such an activity mutual aid. One of theprominent features of the Japanese society is considered to be the strength of mutual

6These forecasts have been used to determine the contribution rates for the “pay-as-you-go” typepublic pension in Japan. The optimistic forecasts played an important role in delaying the increase inthe pension contribution rate. In 2004, the Japanese government recognized that the public pensionfund was heavily indebted, and thus made a big pension reform.7Another reason is the increase in late marriages. The average age of the first marriage of Japanesewomen was 24.3 in 1973 and is 29.4 in 2015.


Fig. 1.5 Trends in marriage and divorce rates. Source: Data published by NIPSSR

Fig. 1.6 Decline in the share of the elderly people living with children. Source: National census

aid among family members, people in the local communities, and even workers inprivate companies.8

Let us first consider mutual aid among members of a family, which is the mostimportant community in human society. From the viewpoint of mutual aid, children’ssupport for the elderly is very important. It is known that the share of people above65years of age livingwith their children was over 80% in the old days, which is muchhigher than that of other developed countries (c.f. Fig. 2.2). Figure 1.6 indicates that

8Miyajima (1992) is an important work in which social security policies are discussed based on thesolid analysis of mutual aid within families and companies.

http://dx.doi.org/10.1007/978-4-431-55909-2_2


Fig. 1.7 Relationship between three-generation households and dual-earner households. Source:Yamashige (2002)

the ratio has been rapidly declining since the 1960s and the ratio of the householdswith single elderly members or elderly couples has been increasing.9

The high ratio of elderly people living with their children suggests that ties offamilies had been strong under the social normof children taking care of their parents.From the viewpoint of public policies, it is important to recognize that the normallowed the government to save the amount for expenditure on healthcare and long-term care of elderly people. Furthermore, it is also important to realize that youngcouples living with their parents usually means that they can get nursery services fortheir children from their parents (i.e., grandparents of the children). Therefore, theyoung couples can raise their children and work outside their homes without usingdaycare centers. Such a custom has allowed the government to save the amount forexpenditure on childcare.

Yamashige (1998), using the data of 47 prefectures in Japan for the period 1992-1993, demonstrated that the per-capita national health expenditure had a positivecorrelation with the elderly ratio but a negative correlation with the ratio of elderlypeople living with their children. The result suggested that a fair amount of thehealthcare expenditurewas savedbecause of children livingwith their elderly parents.

Using the data of 47 prefectures in 1995, Yamashige (2002) also showed that theratio of the three-generation household (elderly living with their children’s family)[ThreeGenerationHouseholds] has a positive correlationwith the ratio of householdsof working couples [Dual Earner Households], as in the following equation and Fig.1.7.10

9See, for example, Traphagan and Knight (2003) for more analyses and discussion of Japanesefamilies under demographic changes.10The value in each parenthesis is t-value and R2 is the adjusted coefficient of determination.


Dual Earner Households = 20.65 + 0.95 Three Generation Households(20.20) (14.72) R2 = 0.82

The result suggests that elderly parents allow their children and their spouses towork, by providing nursery care for the grandchildren. Such private provisions ofnursery services should allow the government to save on such expenditure.

The mutual aid within families has been changing recently. The ratio of elderlypeople living with the children has been declining since the 1960s. The weakeningtieswithin families can be the cause of poverty and loneliness of the elderly, decliningfertility rates, and the expansion of social security payment.

Now,wemoveon to the analysis of themutual aidwithin local communities knownas chounaikai (neighborhood association) and jichikai (self-governing association).These local communities are neither tied by blood nor religion, but by the area wherepeople live. Although they may originally have been organized voluntarily by thepeople, they were artificially organized by the government during wartime to controlresource allocations in local areas.

Even after the Second World War, they remained as voluntary associations. Astudy by the Ministry of Internal Affairs in 1980 found that there are 274,738 neigh-borhood associations in Japan. The neighborhood association systems covered 100%of the area in 87% of cities and towns, and they covered more than 90% of the areain 96.1% of cities and towns (c.f. Isomura and Hoshino 1990).

They played some important roles in the local communities by organizing activi-ties to maintain the natural and social environment, improve the welfare of childrenand elderly people, and encourage collaboration among community members (c.f.Table1.1).

According to the study, 86.2% of neighborhood associations collected contribu-tions from inhabitants, and 35.5% received some kind of subsidy (consisting of 6.8%of the revenue for an average neighborhood association). Furthermore, 70.3% wereasked to carry out certain tasks by local governments (in return for subsidies), thatis, to act as agencies.11

The functions of such neighborhood associations can be fruitfully discussed usingthe framework of social capital, which has recently received a fair amount of attentionin economics. Social capital is the network of people that can be viewed as theinfrastructure to support our lives in society, just like physical public capital in oureconomy. Enriching such social capital is expected to increase social welfare. Wediscuss social capital in more detail in Chap. 7.

A study by Cabinet Office of Japan (2003), one of the seminal researches on socialcapital in Japan, estimated the social capital index in each prefecture and found thatit has a negative relationship with crime rates (c.f. Fig. 1.8). In areas where the localcommunities have strong ties, the social capital, in general, is rich; and themutual-aidtype activities (e.g., crime prevention) have positive effects on reducing crime rates.

11Japanese local governments have regarded neighborhood associations as useful entities for imple-menting their policies. Even to this day in Tokyo, large subsidies are given to neighborhood asso-ciations.

http://dx.doi.org/10.1007/978-4-431-55909-2_7


Table 1.1 Functions of neighborhood associations (as of 1980)

Function Number of Ratio (%)

Circulating messages from local governments 3095 94.8

Reporting inhabitants’ demands to local governments 2919 89.4

Maintaining community centers 2919 89.4

Cleaning public areas 2782 85.2

Organizing local festivals 2687 82.3

Cultural and sports activities 2687 82.3

Contribution to financial and blood donations 2683 82.2

Maintenance of street lights 2157 66.1

Fire prevention and crime prevention 2073 63.5

Organizing ceremonies and events for the children and elderlypeople

2004 61.4

Traffic controls and safety 1684 51.6

Maintenance of streets 1642 50.3

Mutual help in weddings and funerals 1588 48.6

Publishing community newsletters 525 16.1

Source: Isomura and Hoshino (1990)

Fig. 1.8 Social capital indicies, crime rates and TFR. Source: Data in Cabinet Office (2003)

However, the ties of local communities have been rapidly weakening in Japan.Many people, especially young people, now prefer participating in activities in vol-untary associations (e.g., NPOs and NGOs) whose values and interests can be sharedamong the participants, which is in contrast with the neighborhood associations,where participation is partially compulsory. However, it is important to recognizethat as long as participation in the activities of these associations is voluntary, the level


of mutual aid will be lower than that realized in partially compulsory neighborhoodassociations, as we will see in Chap.6.12

Concerning mutual aid in Japan, many scholars (e.g., Miyajima 1992) haveclaimed that Japanese companies played the role of communities. To maintain theflexible Japanese production system with high total quality management, workersneeded close cooperation based onmutual trust. By providing long-term employmentand a seniority payment system, the Japanese companies created circumstances sim-ilar to the communities to support the workers’ cooperative activities. Such employ-ment systemswere rationalwhen the Japanese economywas growingwith the growthin population.

When the domesticmarkets stagnated and the competition got harsh due to global-ization, innovation and effective use of capitals and inexpensive labor in foreign coun-tries became more important than the cooperation among Japanese workers. Undersuch changes in business environment and strategies, many Japanese companiesreduced the weight of long-term employment and seniority payment. The Japanesecompanies became more contract-based than relation-based (as in the communities).Now, Japanese companies provide neither employment security nor income securityto workers. Thus, the government is required to provide such securities to the people.

1.3 Economic Transformation in Japan

In the last section, we saw that against the backdrop of changes in families andlocal communities, there were economic changes. In this section, we identify thoseeconomic changes that are closely related to social transformation in Japan.

1.3.1 Regional Disparity

The economic growth in the late 20th century leads to regional disparity. One ofthe main reasons for such a change is the outflow of young workers from rural tourban areas.13 Such labor mobility can be easily increased in the process of economicgrowth. In the case of Japan, it was intentionally encouraged.

That is, after the Second World War, intensive public investments were made inthe three largest urban areas (i.e., Tokyo,Osaka, andAichi).Many private companies,

12In areas where neighborhood associations are not working well, such as urban areas, the estab-lishment and activities of the voluntary associations will have positive effects on the communities,because they are playing a role in sustaining the activities that neighborhood associations tradition-ally conducted.13The inflow of the young people increased the attractiveness of the urban areas and acceler-ated the inflow of the young people to urban areas, which created further regional disparities (c.f.Sect. 10.2.1).

http://dx.doi.org/10.1007/978-4-431-55909-2_6

http://dx.doi.org/10.1007/978-4-431-55909-2_10

1.3 Economic Transformation in Japan 13

Fig. 1.9 Trends of public investment per person. Sources: Local Policy Office in MOHA (1972,1998), Statistics Bureau (1997)

ministries of the central government, universities, and research centers were rebuiltin those urban areas; and people, especially young people flowed into these urbanareas from rural areas.

This is an understandable growth policy to rebuild the country destructed bywar. By assembling resources and people from all over Japan in the urban areas, theJapanese government tried to create a rapid recovery and economic growth. Figure 1.9compares trends of the ratio of public investment per person for selected prefecturesto the general average of Japan. It indicates that, until early 1970s, higher publicinvestment had been made in the three largest urban areas.

Figure 1.10 indicates big net inflows of population to the three largest urbanareas until the early 1970s. To generate rapid recovery and growth under the limitedresources in Japan, such a strategic growth policy must have worked well.

However, there were negative side effects of such growth policy. The most seriousones include overpopulation in urban areas and depopulation in rural areas. A seriesof Comprehensive National Development Plan by the government recognized theseproblems and proposed to accelerate regional development after the 1970s. Such achange in development policy can be seen in the increase in public investment inAomori, the representative rural prefecture, in Fig. 1.9.

Responding to such a change, the net inflows of the population to urban areasreduced in the 1970s (c.f. Fig. 1.10). In the 1980s, changes occurred in public invest-ment and migration. Public investment in the Tokyo region started to increase againin the 1980s and the population migration to this area started again. Such changes didnot occur in the Osaka and Nagoya regions. This inflow of population to the Tokyoregion stopped after the economic bubble burst in 1989, but restarted since 1996.The process of unipolar concentration in Tokyo is still underway.


Fig. 1.10 Net inflow of population to three largest urban areas. Source: Statistic Bureau’s annualreports on the migration of population based on basic resident register

Although the correlation between public investment per person and populationinflow does not necessarily imply causality, it suggests that the government’s devel-opment policies since the 1980s have shifted towards the intensive development ofthe Tokyo region, which will continue at least until the Tokyo Olympics in 2020.

The Japanese government’s development policies after the war have createdregional disparities. Tomitigate critical voices against them, the Japanese governmentused the fiscal equalization system known as “local allocation tax system” introducedin 1954. The system automatically provides grants from the national government’stax revenue to local governments whose basic revenues are not enough to cover thebasic expenditures (c.f. Sect. 10.3.1).

Under such a redistribution system, each local government does not have to worryabout the sustainability of the local communities even if the depopulation continuesin rural areas. One of the town mayors of rural towns in the Kumamoto prefectureanswered, in an interview in a Japanese newspaper (Nihon Keizai Shinbun, August15, 1999), as follows.

I am not worried too much about the population decline unless there is a collapse of thefoundation of lives, in which we lose the chance to earn income, fail to sustain healthcareand welfare, or give up our cultural activities.

During the period of rapid economic growth, this redistribution system seemedto have worked well in assembling human resources in urban areas, generating highnational income, and redistributing the fruits of the economic growth to the ruralareas, which provided precious human resources to the urban areas. However, thissystem had the side effect of accelerating the depopulation of the rural areas.

Under such a redistribution system, the foundation of people’s lives is guaranteedin themost rural areas and people there can get income, good healthcare, andwelfare.Therefore, the local governments did not have any incentive to stop the outflow of

http://dx.doi.org/10.1007/978-4-431-55909-2_10


young people because they could receive grants from the local allocation tax and therewas nothing to worry about. The redistribution system to mitigate the depopulationproblem is likely to have accelerated the problem further.

Nowadays, in Japan, many rural areas are disappearing or expected to disappearin the near future. We now find that the redistribution system was not a good systemto encourage independence and autonomy of the local governments for a sustainablesociety.

Furthermore, in addition to the general fiscal redistribution system, there weremany other protective policies and subsidies for the regions where the major indus-tries (e.g., agriculture and coal mining) suffered from the structural changes of theJapanese andworld economy.14 Such protective policies are expected to provide timeto the areas to make a structural change for the promising new industry.

However, the policies did not have the intended effects in many cases. Structuralchanges were not so easy and the subsidies were mainly used, not for investment butfor consumption to support the lives of the people affected by the structural changes.When the subsidies gradually reduced and stopped, people left the areas and a localgovernment went bankrupt after accumulating debt in its attempt to restructure theindustry.

The fiscal redistribution system in Japan may have been a good system from theviewpoint of both efficiency and equity in the period of rapid economic growth.However, the review of the fiscal history suggests that this policy was short-sightedand had serious side effects in the long run.

The acceleration of the depopulation of rural areas is just one of the examples ofthe side effects. In the process of deepening depopulation, for instance, separation offamilies, weakening communities, and increasing regional disparities were commonand Japan is now facing various challenges that the redistributive system broughtabout. The argument here suggests that to solve the problems, we need to make goodreforms in the fiscal systems such as the local allocation tax system, which causedthese problems (c.f. Chap. 10).

1.3.2 Income Inequality

Let us look at the changes in income inequality in Japan, by focusing on the changesin the Gini coefficient as a measure of income inequality. The Gini coefficient takesthe value 0 if the income is equally distributed and the value 1 if one person has allthe income in the society. It increases if the society departs further from the equaldistribution of income.

From the viewpoint of economics, “equal distribution of income” does not nec-essarily mean that the society is fair. Therefore, we cannot say that a society with asmaller Gini coefficient is better. However, it is an important signal for the conditions

14See Honma (1993) for the policies on agriculture and Shimazaki (2015) for the policies on coalmining industry.

http://dx.doi.org/10.1007/978-4-431-55909-2_10


Fig. 1.11 Changes in the Gini coefficient. Source: Isakami (2003, Figure 1–1 and Figure 1–2)

of the society. A society with a high Gini coefficient, for instance, is the one in whichpeople face a high risk of falling into poverty.

In this book, following the convention, if the Gini coefficient goes up, we say thatinequality getsworse, and if it goes down,we say that inequality gets better. However,it is important to understand the meaning of the changes in the Gini coefficient forthe changes in the social structure.

Figure 1.11 indicates the changes in income inequality in Japan.15 The data indi-cate that before the Second World War income inequality was gradually gettingworse. After the war, inequality suddenly dissolved, but it got worse from around1955 to 1960. Then, the inequality situation improved in the 1960s and 1970s butbegan to worsen again in the 1980s.

Concerning the reasons behind the equalization of income after the war, Tanisawaand Minami (1993) point out the following three reasons: (1) Wealth loss due to thewar and hyper-inflation after the war, (2) Dissolution of the Zaibatsu big financialcompanies in Japan, and (3) Property tax and wealth tax.16 On the other hand, whatcould be the reasons for the increasing inequality in the 1980s17?

15It is well known that estimating the income inequality faces many difficulties due to the problemsof data acquisition (c.f. Terasaki 1993). Although the values are different among various estimates,the pattern of changes in income inequality indicated in each estimate seems to be consistent.16The property tax was imposed for six years from 1946 to 1951 and the wealth tax was imposed bythe Shoup recommendation for three years from 1950 to 1952. The property tax was an individualtax on assets with a low tax exemption level and the highest tax rate set at 90%. It naturally had avery high equalization effect on the assets and thus on income. On the other hand, the wealth taxwas an individual tax on net wealth. Contrary to the property tax, the tax exemption level is highand the highest tax rate is 3%; thus, the redistribution effect was considered to be limited.17Economic Planning Agency (1999) and Ohtake (2005) show that the trend continued in the1990s. Concerning the international comparison of the income inequality, some researchers (e.g.,


Ohtake (1997, 2005) and Economic Planning Agency (1999) identify the follow-ing six factors: (1) Increase in wage inequality, (2) Increase in inequality in wealthincome, (3) Increase in the number of the double-income households, (4) Populationageing, (5) Decrease in household size, and (6) Tax reforms since 1987. In the fol-lowing part, we will provide more explanations for the last three factors, which areclosely related with our discussion on public policies.

First, (4) population aging is considered to be one of themost important factors forthe increase in the Gini coefficient in Japan (e.g., Ohtake and Saito 1998 and Ohtake2005, Chap.1) because the income inequality among the elderly is larger than thatamong younger people. In such a case, population aging, under which more weightsare given to the elderly generation, can increase the Gini coefficient in Japan.

Second, in point (5), the decrease in the household size, which was a result ofthe separation of the elderly from their children’s family, increase in divorce rate,and decrease in marriage rate, can contribute to the increase in the Gini coefficientbecause the incomes of the elderly and women are low, on average. Separation ofthese groups fromother groups can increase theGini coefficient (e.g., Terasaki 2000).

Finally, in point (6), the tax reforms in the late 1980s decreased the redistributiveeffects of the tax system in Japan. To see the changes in the redistributive effectsof the tax system, Fig. 1.12 depicts the trends of the redistribution index, whichmeasures how much the tax system decreased the Gini coefficient. It indicates thatafter the war, the index was very large. After 1952, it declined rapidly and remainedlow even after 1957.18

We have already stressed that what is important is not the increase in the Ginicoefficient, but the changes in the Japanese society suggested by it. If the elderly livewith their children’s family, shocks can be shared among householdmembers throughmutual aid. Population aging and decrease in household sizes imply an increase inthe number of people who cannot absorb the shocks in their income and expenditure(e.g., healthcare) independently. We should pay attention to the change in the Ginicoefficients from the viewpoints of not only inequality, but also risks of falling intopoverty in the society (c.f. Sect. 9.2.1).

Indeed, the number of recipients of the public assistance program for the poorpeople has been increasing. Among them, the elderly occupied a large portion, andtheir number has been increasing rapidly (Fig. 1.13).19

(Footnote 17 continued)Tachibanaki 1999, 2005) claim that the Japanese society is one of the most unequal societies whileothers (e.g., Economic Planning Agency 1999) suggest that Japan’s inequality is average. Thecomparison of the degree of inequality is not easy due to the data available in each country, andthus should be done carefully.18In this period, there were a series of tax reforms, such as abolition of the property and wealthtaxes, and reduction in the inheritance tax and the asset income tax (e.g., Sato and Miyajima 1990).Especially, after the fundamental tax reform in 1987, there was a big decrease in the index.19The households with elderly people under the public assistance program comprise 43% of thetotal households.

http://dx.doi.org/10.1007/978-4-431-55909-2_9


Fig. 1.12 Trends of redistribution. Source: Ishi (1993; 2001)

Fig. 1.13 Recipients of the public assistance program. Source: Data published by NIPSSR

Needless to say, if there are ways to share risks, the risk of falling into poverty willbecome lower (c.f. Sect. 9.1). Historically, we have mitigated various risks throughmutual aid among members in families and communities. However, such a system ofmutual aid has become very weak under the transformation of the Japanese society.Hence, the transformation of the Japanese society seems to be one of the mostimportant reasons for income inequality, and thus the risks in our society have beenincreasing in Japan. We need to have a better understanding of the transformation,its consequences, and public policies for the changes.

http://dx.doi.org/10.1007/978-4-431-55909-2_9

1.4 Concluding Remarks 19

1.4 Concluding Remarks

Japan was a tiny island hosting only about twenty thousand people in the early Jomonperiod. Now, more than a hundred and twenty million people are living in Japan, andit has become one of the most developed countries in the world. The rapid social andeconomic development in the 20th century, especially after the Second World War,may be referred to as an Asian miracle.

However, the population in Japan hit its peak in 2008, and started to decline. It isestimated that, if things (e.g., fertility and immigration) do not change very much,the population will be less than sixty million, less than half of the current population,by the end of this century. The rapid decline in population will have serious negativeeffects on the economy.

The most important reason behind the population decline is low fertility. Thequestion is why the fertility rate has become so low in Japan. We claim that thelow fertility is a result of the transformation of Japanese families and society. TheJapanese people are no longer dependent on families or communities, and they havestarted trying to live as free independent individuals.

In the process of economic growth and development of the social security, manyyoung people left their hometowns leaving their elderly parents, and lived in urbanareas. It increased the regional disparity and reduced the mutual aid within familiesand communities, contributing to the increase in income inequality.

Such changes would not have had big impacts on “winners” in the market whohad economic success. However, they had serious negative effects on “losers” in themarket economy, to whom families and communities traditionally provided someaid to mitigate poverty. The safety nets provided by families and communities havebeen disappearing and the dependence on social security, which weakened the tiesof families and communities, was increasing.

In response to the voices of the people, especially the elderly, the Japanese govern-ment expanded social security payment. Under the slowdown of economic growthand fiscal revenue, the government politically failed to increase the tax and socialsecurity contribution, and has accumulated a public debt greater than 200% of theGDP.

Although people’s mindsets have changed, the social norms shared by them,especially the elderly, have not changed very much. For example, the social normthat “husband should work outside home and wife should stay at home” still remainsin the Japanese society (c.f. Fig. 11.2). It has discouraged the expansion of nurseriesthat allow women to work while raising children, thus contributing to low fertilityrates in Japan.

The population and fiscal crises in Japan were caused by the transformation of theJapanese society and were further deepened by inadequate public policies. The finalgoal of this book is to find appropriate public policies under such transformations offamilies and communities. We believe such studies will have important implicationsfor public policies in other countries that face a similar transformation in society.To achieve the goal, we first attempt to understand functional roles and the structure

http://dx.doi.org/10.1007/978-4-431-55909-2_11


of families and communities. Then, we clarify their relationship with the marketsand the governments. In the next section, based on economic analysis, we set up aframework to analyze these issues.

References

Cabinet Office of Japan. (2003). Social capital: Seeking for virtuous cycles of rich human relation-ship and civic activities. (in Japanese).

Economic Planning Agency. (1999). New indices for national lifestyle, 1999.Economic Planning Agency. (2000, 2009). National accounts.Flath, D. (2000). The Japanese economy. Oxford: Oxford University Press.Hayashi, T., Imai, K., & Kanazawa, F. (Eds.). (2001). Handbook of Japanese public finance. Uni-versity of Tokyo Press. (in Japanese).

Isakami, K. (2003). How to see income inequality in Japan. JIL Labor Policy Report No. 3. (inJapanese).

Honma, M. (1993). Japan’s agricultural policy and protection growth. In T. Ito & A. O. Krueger(Eds.), Trade and protectionism. University of Chicago Press. https://core.ac.uk/download/pdf/6483630.pdf.

Ishi, H. (1993). The Japanese tax system, 2nd Ed. Clarendon Press.Ishi, H. (2001). The Japanese tax system, 3rd Ed. Clarendon Press.Isomura, E., &Hoshino,M. (Eds.). (1990). Readings of local autonomy. Toyo Keizai. (in Japanese).Miyajima, H. (1992). Socio-economics in the era of population aging: Family, company, and gov-ernment. Iwanami Shoten.

Ohkawa, I., Takamatsu, N., & Yamamoto, Y. (1974). Long-term statistics 1: National income. ToyoKeizai Shinposha.

Ohtake, F. (1997).Distribution of incomeandwealth in the 1980s.Theoretical EconomicsQuarterly,13(76), 385–402. (in Japanese).

Ohtake, F. (2005). Inequality in Japan: Illusion and Future of Unequal Society. Nihon KeizaiShinbunsha. (in Japanese).

Ohtake, F., & Saito, M. (1998). Population aging and consumption inequality in Japan. Review ofIncome Wealth, 44(3), 361–381.

Raymo, J., Iwasawa, M., & Bumpass, L. (2004). Marital dissolution in Japan: Recent trends andpatterns. Demographic Research,11, Article 14.

Sato, S., & Miyajia, H. (1990). History of tax system after the Second World War. Zeimu KeiriKyokai. (in Japanese).

Shimazaki, N. (2015). Support for workers displaced in the decline of the Japanese coal industry:Formal and informal support. Japan Labor Review, 12(2), 6–27.

Tachibanaki, T. (1999). Changes in wealth value and inequality in wealth distribution. JapaneseEconomic Review, 18, 79–91. (in Japanese).

Tachibanaki, T. (2005). Confronting income inequality in Japan: A comparative analysis of causes,consequences, and reform. Cambridge: MIT Press.

Tanisawa, H., & Minami, R. (1993). Addendum: Factors for equalization of income distributionafter the SecondWorldWar: Changes in conditions for the high income earners and an evaluationfor them. Economic Research, 44(4), 365–373.

Terasaki, Y. (1993). Measurement of income distribution in Japan. Proceedings in Japanese Statis-tical Association, 22(3), 599–612. (in Japanese).

Terasaki, Y. (2000). Life protection functions of household in coresidence of adults. In NIPSSRTransformation and life protection functions of family and household. (in Japanese).

Traphagan, J. W., & Knight, J. (2003). (Eds).Demographic change and the family in Japan’s agingsociety. State University of New York Press, Albany.

https://core.ac.uk/download/pdf/6483630.pdf

https://core.ac.uk/download/pdf/6483630.pdf

References 21

Vogel, E. F. (1979). Japan as number one: Lessons for America. Cambridge: Harvard UniversityPress.

Yamashige, S. (1998). Functions of families and local communities, and roles of the government.Ikkyo Ronso, 120(6), 38–62. (in Japanese).

Yamashige, S. (2002). Families, communities and governments in improving socio-economic per-formances: Japanese experiences. World Bank Institute Working Paper. # 37204.

Chapter 2Markets, Communities, and Government:Analytical Framework

A welfare regime can be defined as the combined,interdependent way in which welfare is produced and allocatedbetween state, markets, and family.

Esping-Andersen, G. (1999, p.34)

2.1 Introduction

In this chapter, we try to clarify functions and limits of the three most important insti-tutions in our society: markets, communities (including families), and governments.We also try to clarify the interactions among those institutions.

In modern societies based on market economies, it is useful to regard communi-ties and the government as institutions for resource allocations that supplement theallocations in markets. Such a way of thinking is a standard in economics, exceptthat we explicitly consider communities as institutions that supplement the resourceallocations by markets and the government.

Roles of a family, one of the most important examples of a community, have beenwell recognized in social sciences other than economics, as suggested by the above-mentioned quote from Espin-Andersen (1999). In this book, we also point out theimportance of various communities other than families, to deepen our understandingon the transformation of our society and the role of the government in it.

The usefulness of such a viewpoint has been put forward in development eco-nomics. Hayami and Goto (2005, p. 311), for instance, argue that “the economicsystem for the development of developing economies must be designed not as acombination of market and state but as a combination of the three organizationsincluding community.” The viewpoint is not only useful in analyzing developingeconomies, but also in understanding the transformation from developing to devel-oped economies (c.f. Aoki and Hayami 2001).


23

24 2 Markets, Communities, and Government: Analytical Framework

2.2 Functions of Markets, Communities, and Governments

We first summarize functions and limits of markets, communities, and governmentsfrom the viewpoint of resource allocations.1

2.2.1 Functions and Limits of Markets

The following first welfare theorem is useful in understanding functions and limits ofthe market system. The “efficiency” in the theorem refers to the condition in whichthere is no waste in the allocation, which is referred to as Pareto efficiency2 namedafter an Italian economist, Pareto (1848–1923).

Proposition 2.1 (First Welfare Theorem) If all goods and services are traded in themarkets, under perfect competition, efficiency of allocations will be achieved.

This theorem is the basis of most economists who claim that the market sys-tem functions well in allocating precious resources efficiently in our society. It alsoindicates the limitations of the market system, by pointing out two conditions forefficiency. First, all goods and services are expected to be traded in markets, whichwe refer to as “universality of markets.” The second condition is the “perfect com-petition” under which economic agents need to take the market prices as given, withno power to change them.

These two conditions are sufficient conditions for efficient resource allocation inthe market. However, it is well known that allocations can be inefficient when eitherof the two conditions is not satisfied, which we refer to as “market failure” problems.As the two conditions are rarely satisfied, our society is full of market failures.

In the following part, we briefly explain sixmarket failure problems: (1) imperfectcompetition, (2) externalities, (3) economies of scale, (4) incomplete information,(5) incomplete contracts, and (6) inequality. Although inequality is not a problemassociated with the first welfare theorem concerning the efficiency of the markets,we regard it as a market failure problem because it is an important problem that themarket cannot solve and often exacerbates.

2.2.1.1 Market Failure: Imperfect Competition

In markets, sellers try to earn some monopoly profits by differentiating their goodsand services from other sellers. Even if such efforts are not made, goods and services

1A good reference for the explanations of functions and limits ofmarkets and governments is Stiglitzand Rosengard (2015).2To be more precise, Pareto efficiency refers to the condition in which it is no longer possible toimprove anyone’s welfare without deteriorating that of others.

2.2 Functions of Markets, Communities, and Governments 25

provided in different places can be seen as different goods and services. Hence,markets are naturally imperfect.

Under imperfect competition, sellers try to set their prices higher than the pricesunder perfect competition, and thus the transactions, in general, become smallerand inefficient. The market system itself usually does not have good mechanisms toprevent such behavior because sellers try to maintain their monopoly powers to earnhigher profits. Government’s intervention to mitigate such problems that the marketsystem fails to solve is justified.

2.2.1.2 Market Failure: Externalities

When some actions have effects on the welfare of other economic agents outsidethe market system, they are said to have externalities or external effects. Creatingnoise, garbage, air pollution, water pollution, and exploitation of resources for freeare a few examples of negative externalities. Provision of good education, goodenvironment, peace and order, and maintenance of a stable financial system areexamples of positive externalities. The existence of such externalities suggests thatthere are many goods and services that are not traded in the markets and are providedfor free.When the “universality ofmarkets” is not satisfied and there exist problemsof“missingmarkets,” resource allocations become generally inefficient. Resources willbe overused under negative externalities and underused under positive externalities.

Such problems usually occur because the property rights of some resources arenot well-defined. For the market system to work efficiently property rights of goodsand services need to bewell-defined and protected; hence, the intervention of govern-ments is required. However, it is difficult to define and maintain the property rightsof all goods and services.

Even if the governments fail to adequately define property rights of some resourcesandmarkets cannot achieve efficient allocations, as long as external effects are limitedto small communities, people will cooperate with each other to improve efficiency inthe communities (c.f. Coase theorem in Sect. 3.1). For example, in our family lives,we usually provide various services that are full of externalities. However, our familylives are not so inefficient, if not perfectly efficient, because we cooperate with eachother to reduce inefficiency.

Although we usually do not expect the government to intervene in family lives,if the externalities are spreading over wide communities, the cooperation within thecommunities will become more difficult. The intervention of the government in suchcases will be demanded and justified. It is well known that the government can miti-gate the problems of externalities by the direct provision of goods and services withexternalities (e.g., education), designing and implementing regulations (e.g., pollu-tions), setting taxes and/or subsidies (e.g., carbon taxes and subsidies for education),and creating new markets (e.g., markets for emission credits).3

3Note that the government policies to mitigate market failures can generate positive or negativeexternalities. For example, the social security system can have negative effects on our fertility

http://dx.doi.org/10.1007/978-4-431-55909-2_3


2.2.1.3 Market Failure: Economies of Scale

Ahigh fixed cost is required to start a business such as the supply of water, electricity,gas, railway services, and telephone services because it requires infrastructure toprovide the service. The average cost for such businesses decreases as the scale ofsupply increases. Under such “economies of scale,” the market price that guaranteesefficiency is often lower than the average cost when the market demand is not largedue to, say, low economic development. If the price, that is, revenue for each unit ofoutput, is lower than the average cost, the profit becomes negative. Hence, no supplywill be made under perfect competition and the market fails to achieve efficiency.

The problem lies in the fact that the market demand is too weak to cover largefixed costs. However, the businesses listed above are essential for our lives. Hence,the governments often set up public enterprises to provide such services and use thetax revenue to compensate for the deficit under efficient supply.

2.2.1.4 Market Failure: Incomplete Information

When there are informational gaps between sellers and buyers,we say that there exists“incomplete information” or “asymmetric information.” In markets, it is difficult toobtain accurate information about qualities and/or risks of goods and services. It isknown that market transactions are not efficient with incomplete information.

The problem is known to exist in the insurance markets, where insurance com-panies have difficulty in obtaining information about the risks of subscribers. If theinsurance premium is set to cover the average risks of subscribers, it is well knownthat it will lead to the “adverse selection problem.” It is a problem that only thosewho have higher risks than average will think that the premium is acceptable andthus purchase the insurance. That is, under incomplete information, subscribers thatthe insurance companies would not want to insure will be adversely selected by themarket. In the worst case, the insurance markets fail to exist (c.f. Akerlof 1970).

If such insurance is very important for most people, it is often justifiable thatthe governments provide it by forcing people to purchase the insurance to avoid theadverse selection problem. The universal public pension and healthcare, in whichall the people in the country are forced to participate, can be justified by the adverseselection problem in the insurance markets.

Incomplete information about the behavior of sellers and buyers can also causemarket failure problems. For example, when people have full insurance for healthcareunder which they do not have to pay at all for the treatment of sickness, they maybehave less carefully than when they have no insurance. Such rational behavior isreferred to as the “moral hazard problem.” The problem is caused by the incomplete

(Footnote 3 continued)behavior, and under the labor income tax system, our labor will have positive externalities on thesociety. Public policies need to be designed by considering such externalities.


information about our behavior in the sense that such wrongful behavior would beprevented if the insurance company knows about it.

2.2.1.5 Market Failure: Incomplete Contracts

Theoretically speaking, there must be some contracts for all market transactions.Hence, for market transactions to be efficiently conducted, the legal and judicialsystem should be well developed so that any failure to comply by the terms of thecontracts should be adequately punished. For the system to work, each contract mustbe complete to handle every possible case that may occur after the market transactionis made. However, there are infinite possibilities in the future. As we cannot writecontracts for infinite contingencies, each contract must be an incomplete contract tosome extent.

When the problem is serious, market transactions will be inefficient. This is alsoa problem that the markets cannot solve and is thus considered a market failureproblem. The problem of incomplete contracts exists not only in market transac-tions, but also in formal and informal contracts in the activities of communities andgovernments, as we will see in this book.

2.2.1.6 Market Failure: Inequality

One of the characteristics of the market allocation is that those who have biggerendowments (e.g., abilities and wealth) can enjoy better lives, and those who havelittle will perish. The market mechanism is cruel. The market system does not satisfyvertical equity in the sense that those who have less will envy those who have morealthough it satisfies horizontal equity in the sense that people need not envy thosewith the same endowments.

However, the secondwelfare theorem suggests that we should not discard themar-ket system to achieve equity in it, because it indicates that under certain conditions,any allocation that is equitable can be achieved in themarket as an efficient allocationif the appropriate redistribution of the initial endowments by lump-sum tax and/orsubsidy is achieved. Although it is very difficult for the conditions to be satisfied,the second welfare theorem suggests that the equality of the market economy canbe improved by the government’s intervention to redistribute the endowments of thepeople and allow people to trade in the markets.

However, the adequate redistribution of the initial endowments is very difficultbecause the government does not have enough information about it. If so, redistribu-tion of outcomes of market transactions, such as labor incomes, may be justified. Inmost countries, there seems to be some consensus, besides guaranteeing the funda-mental human rights in amarket economy, to redistribute incomes that are attributableto the unequal distribution of initial endowments among people.


Remark 2.1 (Equity) In economics, it is considered desirable to achieve efficient andequitable allocation. It is regrettable that many economists place too much emphasison efficiency due to the difficulty in defining equity. In this book, we consider anequitable society as one in which equal opportunities are guaranteed.4

Although “equal opportunity” is also difficult to define, it seems natural, fromthe viewpoint of economics, to define it as the equality of choice sets.5 In standardeconomics, such choice sets are essentially the budget sets (c.f. Remark 3.4). Thebudget sets are dependent on individualwages, that is, the ability to earn labor income,which is generally different for different individuals. Hence, equal opportunity is notguaranteed in a market economy. It can be viewed as one of government’s functionsto reduce the inequality in our opportunities.

2.2.2 Roles and Limits of Governments

How the market failures can be solved has a deep impact on the quality of oursociety. Mainstream economics has regarded the government as an institution thatshould mitigate such problems.

2.2.2.1 Roles of the Government

Whatwould happen if the government disappeared tomorrow?Therewould be chaos.In some societies, it may lead to “the war of all against all” as Thomas Hobbes(1588–1679) described in his Leviathan.6 Theremay be outbreaks of some infectiousdiseases, loss of properties by fire and disasters, or suspension of various activitiesbased on credit transactions. Lives of many people would be lost due to inadequateor the absence of treatment. Markets would not work well, and even if they wouldexist, they would suffer from market failure problems.

To avoid a “war of all against all” and protect our basic human rights, we willengage in a social contract with each other and ask the Leviathan, the powerful seamonster described in the old testament of the Bible (e.g., Job Chap.3), to punishthose who break the contract. In this theory of social contract, the people demandthe government to play the role of Leviathan, who has the power to protect the basichuman rights, and under which markets work well.

4This definition is imperfect as compared to the one for equitable allocation because it does notallow us to compare equity of allocations when equal opportunities are not guaranteed. We candefine the degree of equity if we define a measure of differences of opportunity sets.5The notion of equality adopted in this book has a close relationship with the envy-free allocationas the equitable allocation (c.f. Pazner and Schmeidler (1974) and Yamashige (1997)).6The book, Leviathan or The Matter, Form and Power of a Common Wealth Ecclesiastical andCivil, was published in 1651.

http://dx.doi.org/10.1007/978-4-431-55909-2_3


The government can play many important roles to mitigate market failures inour society. However, Leviathan is not perfect. It will often fail to realize the willsof the people. This is called the “government failure problem.” Hence, even if thegovernment’s intervention in some market failures may be justified, it may not bedesirable.

The desirable roles and the size of the governmentwill be determined by balancingthe benefits of reducing market failures and the costs of government failures, thatis, by maximizing the net benefits of government intervention. The most difficultpart of a democratic society under social contract is that people often have differentperceptions and values about the benefits and costs of government intervention.

2.2.2.2 Government Failure

Human beings have developed democratic societies in which the government willbe operated by the agents elected by the people. As the government is an organiza-tion, it faces organizational problems. Hence, the government failures are essentiallyproblems in governance of the government.7 The question is why the governance ofthe government fails.

There are two sources of failures peculiar to the governments. The first one is thatthe peoplewho are expected to have sovereign power have little interest inmonitoringthe government. It is because the benefit of monitoring the government is very smallcompared to its cost, as each individual is too small to change the behavior of thegovernment even if he/she finds some problem in monitoring. The situation is incontrast to the case in which one owns shares of a private company. Although onlysmall shares of the company are owned, there is an incentive to monitor the companybecause the shares can be profitable if traded.

The difference lies in the fact that each individual cannot choose the governmentwhile he/she can choose the company by selling or buying the shares. Under littleincentive to monitor the government, actors in the governments can pursue their self-interests rather than interests of the people. The essential problem is “incompleteinformation” about the behavior of the government and the low incentive for thepeople to collect the information.

The second source peculiar to the government is that the value of the servicesprovided by the government is difficult to assess, as most of them are provided forfree, that is, not appraised in terms of money. If performance of the organization ishard to evaluate, disciplining the organization becomes difficult. This is in contrastto the case of for-profit companies, whose performance in terms of profit can beeasily evaluated in monetary terms. The contracts of reward and punishment to dis-cipline the organization can be more easily engaged in such a case than in the casewhen the measure of the performance cannot be clearly defined, as in the case of the

7Governance of an organization means that those who have the rights and/or the responsibilities ofcontrolling the organization, discipline it.


government. Due to such a problem of incomplete contracts, the agents of the gov-ernment can pursue their own interests.

In addition to these problems, the complexity of the government, as an organi-zation, contributes to the failure of governance. For example, in most countries, thepower of the governments is divided into three parts to prevent the Leviathan fromabusing it. The structure comprises of an interactive relationship among variousactors of the government, which makes governance more complicated as comparedto the top-down structure in corporate governance. Hence, the probability of thegovernment failing is higher than that of private companies.

Finally, there is another source of government failures, which lies above the orga-nizational problem. It is the difficulty of social choice in democratic societies wherethe government’s decisionsmust bemade by some rule to be based on the preferencesof the people. However, Arrow (1950) demonstrated that it is impossible to design asocial choice rule that satisfies the desirable properties.

Hence, even if the governance of the government works perfectly, democraticdecisions must have some defects that can cause government failures in terms ofefficiency or equity. Furthermore, even if future generations may be affected by somepolicies, they will not be able to participate in the political process to determine thepolicies if they are not born yet, and thus, the welfare of the future generationswill be deteriorated by policies determined by the current generation (e.g., publicdebt accumulation). This limitation of social choices is one of the sources of thegovernment failure wherein the governments worsen the inefficiency and inequalityin society.

From a social perspective, it is desirable to minimize the problems of marketfailure and government failure by considering the roles and size of the government.However, there will be problems that markets and governments cannot mitigate evenif they do their best. In fact, it is families and communities that have been trying tomitigate market failures before the welfare societies were developed, and as we willsee below, they are still playing important roles in mitigating market and governmentfailures.

2.2.3 Functions and Limitations of Community

In this book, communities imply networks connected by blood, location, religion,interests, and so on. Families, neighborhood associations, religious groups, andNPOs(Not-for-Profit Organizations) are some examples of communities.

Communities connected by interests, such as NPOs, are often referred to as “asso-ciations.”We call them“newcommunities” and the others “traditional communities.”In this book, we examine the functions of communities in comparison with markets


and governments, and define communities as the networks of people not connectedby profit-seeking motives.8

The “new communities” have emerged in modern societies in their attempts toovercome the limitations of the “traditional communities.” We defer the discussionon new communities to Chap. 7 and try to summarize functions and limitations offamilies and local communities as good examples of the traditional communities.

2.2.3.1 Functions of Families

We first consider the functions of families, one of the oldest and the most commonforms of communities in the world. Why do we get married, have children, and formnew families? To understand the functions of families, we summarize the benefitsand costs of family formation based on Weiss (1997).9

From the perspective of economics, we can argue that people will get marriedif the benefit of marriage exceeds its cost. There are several benefits of marriage.The first benefit is the economies of scale from living together. That is, the averagecost of living is smaller when living with someone as compared to living alone.For example, couples can share the costs of the house, durable goods, and of raisingchildren.Marriage allows people to consume these goods and services at lower costs.

Second, marriage will produce goods and services that the market cannot. Forexample, love is something that markets fail to provide, which marriage and familiesdo. Another example of something that is not usually sold in the market is children.There are many other goods and services that are either not sold or sold at highprices, especially in rural areas. Families can provide various household services atreasonable costs and compensate for the roles of the market.

An important service often provided within families is financial service. Withincomplete information about our abilities, we often face difficulty in borrowingmoney in the market. When people get married, they may be able to borrow moneyfrom their spouse or children. For example, if a husband borrows money from hiswife to study in a graduate school and earn higher income after graduation, it willloosen the budget constraint of the household.

Furthermore, when we are married, we can avail of some insurance that marketsoften fail to provide. For example, insurance against sickness, injury, long-term care,

8Hayami and Goto (2005, p. 310) defines a community as “a group of people tied by mutual trustbased on intense personal interaction.” Our definition is broader than that as we do not include“mutual trust” nor “intense personal interaction,” but it is narrower at the same time, as we excludegroups of people tied by profit-seeking motives. In Japan, private companies (kaisha) are some-times considered as communities. Although we do not consider them as communities, when weconsider the personal networks of workers or labor unions in the companies, they can be termed ascommunities by our definition. See our discussion in Sect. 2.3.1 about roles and changes of laborunions in the market economy.9See also Becker (1973, 1974, 1985, 1993) and Grossbard-Shechtman (1984, 1993).

http://dx.doi.org/10.1007/978-4-431-55909-2_7


and unemployment are often missing in markets due to adverse selection and/ormoral hazard problems because of incomplete information (c.f. Sect. 2.2.1.4). If wewere living alone, we would face high risks in our lives, but if we have spouses andchildren, they can provide some help when we need it.10

The arguments above suggest that families are working as institutions to mitigatevarious market failures and improve efficiency and equity in our lives.

2.2.3.2 Functions of Local Communities

In Japan, local communities also play important roles. For example, Table1.1 inChap.1 indicates that neighborhood associations have provided various public goodsand services in the region. The punishment mechanism, known as “murahachibu,” inlocal communities provides precious information about how the local communitieshave worked in the lives of the Japanese people (c.f. Sect. 6.3.3). Murahachibu isa norm of ostracizing families who do not follow the rules of the community byrequiring members of the community to withdraw 80% of the mutual assistanceto such families.11 The assistance includes that provided for coming-of-age cere-monies, marriages, new births, sickness, flood damages, religious activities, travels,and building houses. The remaining 20% of mutual assistance still provided includesassistance for funerals and fire-fighting to prevent the spread of contagious diseasesand the spread of fires.

The penal system in murahachibu described above gives us a glimpse into themany types of mutual assistance provided in traditional communities and how essen-tial they were for Japanese families in the past. The local communities have playedtheir role in compensating and supporting families.

As we will see in Chap.6, such mutual assistances have been prevalent in thewestern societies (c.f. Remark 6.7 and Ostrom (1990)). They played an importantrole in avoiding the overuse of communal resources, also known as the tragedy ofthe commons (c.f. Sect. 3.3.5).

Various economic analysis have also clarified that communities have providedinsurances that markets and families fail to provide (c.f. Remark 9.1). For example,risks in the production of agricultural products seem to be absorbed to some extentwithin the village so that the variation in consumption is not too high and it corre-lates with the income of the village. As the lack of insurance markets is a typicalmarket failure problem under incomplete information, these findings strongly sug-gest that market failures have been mitigated by the mutual assistance of the localcommunities.

10See, for example, Kotlikoff and Spivak (1981).11In the word murahachibu, “mura” means village and “hachibu” means 80%.

http://dx.doi.org/10.1007/978-4-431-55909-2_1

http://dx.doi.org/10.1007/978-4-431-55909-2_1

http://dx.doi.org/10.1007/978-4-431-55909-2_6

http://dx.doi.org/10.1007/978-4-431-55909-2_6

http://dx.doi.org/10.1007/978-4-431-55909-2_6

http://dx.doi.org/10.1007/978-4-431-55909-2_3

http://dx.doi.org/10.1007/978-4-431-55909-2_9


2.2.3.3 Community Failure

Communities have functioned as institutions to produce values in our society. As theyare composed of members who know each other well, they can avoid the problems ofincomplete information and high costs of bargaining to solve externality problems,thus mitigating market failures.

However, unlike transactions inmarkets, transactions in communities have usuallyfunctioned based on the norm of reciprocity rather than formal contracts and thusoften fail to be efficient. One of the biggest problems is that community membershave an incentive to free ride on the goods and services that are provided by thecommunity.

In market transactions based on formal contracts, the free rider problem can belegally solved, but the solutions will not be available in the communities. Thus, theyusually have their own punishmentmechanisms such asmurahachibu to punish thosewho try to free ride on the mutual assistance within the communities. If they workwell, communities can succeed in conducting efficient production and transactions.

However, such mechanisms often fail to work well, which we refer to as “commu-nity failure”.12 First, due to the problems of incomplete information and incompletecontracts, the punishment mechanism may fail to provide effective punishments tothose who take inefficient actions (c.f. Remark 6.6). Second, for them to work well,mobility must be strictly restrained. If one can leave the community easily after thepunishment, the punishment mechanism will not work well to prevent the free riderproblem. If mobility is severely restricted, then the closedness of the community willdeteriorate the efficiency of the whole society.13

Finally, from the perspective of equity, informal punishment system such asmura-hachibu may be seen as a problem, in that it is brutal and inhumane to make peopleobey traditional behavioral rules. Although the punishment system can be justified tomaintain the efficiency of local communities, there have been criticisms against theJapanese traditional local communities based on the inhumane murahachibu system(c.f. Sect. 6.3.3).

To overcome the problems of market failure and government failure, the expec-tations riding on the functions of the communities have been increasing. However,communities also have their own limitation and failure in improving efficiency andequity in society.Wewill deepen our understanding on the functions and the structureof communities in Chaps. 6 and 7.

12Hayami and Goto (2005) stated that the communities, which are “critically needed to correctthe failures of the market and the state” (p. 311), are immune to “community failure” in achievingsocially efficient resource allocation (p. 343).13See, for example, Coleman (1990, Chap.22).

http://dx.doi.org/10.1007/978-4-431-55909-2_6

http://dx.doi.org/10.1007/978-4-431-55909-2_6

http://dx.doi.org/10.1007/978-4-431-55909-2_6

http://dx.doi.org/10.1007/978-4-431-55909-2_7


Development of theMarket Economy

� Weakening ofCommunities

��

Expansion ofGovernments

Fig. 2.1 Interactions of markets, communities, and governments

2.3 Interactions of Markets, Communities,and Governments

In the modern society, the governments have started to increase their active rolesin resource allocation to mitigate market failure problems. We inquire into the rea-son that the roles of governments have expanded in our society. We first considerthe dynamic interaction between markets and communities and try to explain theexpansion of governments. Then, we consider the effects of the expansion of thegovernment on communities to present a hypothesis on the mechanism of contin-uous government expansion, which we refer to as the snowball-expansion of thegovernment. The basic idea can be illustrated by Fig. 2.1.14

First, when the market economy develops and allows us to trade many goods andservices, the value of goods and services provided by families and communities maydiminish. Then, people can leave their families and local communities to make aliving. With such mobility, the punishment for those who do not obey the rules toinduce cooperation in communities does not work well and the mutual assistancewithin communities is reduced. Mobility of people indeed increased dramaticallywith the development of the market system. The development of the market systemthus weakened community ties and made people more individualistic. This effect isdepicted in Fig. 2.1 as the causal relationship “Development of the Market Economy→ Weakening of Communities”.

When families and local communities are weakened and mutual assistance isreduced, people are in trouble when they face serious life risks. For example, theincome support and care for the elderly by their children, care for the children bytheir grandparents and local communities, and support for the poor by families andlocal communities diminished. Then, people started demanding the governments forhelp. As the governments cannot abandon the elderly, the children, and the poor,they usually start playing more welfare roles. This effect is depicted in Fig. 2.1 as thecausal relationship “Weakening of Communities → Expansion of Governments”.

The governments need not respond to the voices of the frail people. However,when they do, the increase in social security further weakens families and localcommunities, as people need to rely less on them, thus lowering the efforts to con-tinue mutual assistance. This effect is depicted in Fig. 2.1 as the causal relationship“Expansion of Governments → Weakening of Communities”.

14Needless to state, these three institutions have a more complex relationship than the one depictedin Fig. 2.1. For example, change in the communities should have some impact on the developmentof the market economy, which should be analyzed more carefully.

2.3 Interactions of Markets, Communities, and Governments 35

An interesting consequence of the last two causal relationships in Fig. 2.1 is thatthe continuous snow-ball expansion of the government can happen, especially whengovernments try to respond to the voices of the people.

The expansion of governments in the process of the development of the marketeconomywas observed and analyzed, for instance, byWagner (1883).15 He explainedthat governments needed to increase the expenditure to deal with more complextransactions and social conflicts due to the rise in inequality in a market economy.The fundamental reason for the expansion of the government in most developedcountries is the expansion of social insurance. To explain it, we need to pay attentionto the decline in mutual assistance, which served as an insurance for many people.Our explanation depicted in Fig. 2.1 provides an interesting theory for the expansionof governments, which goes beyond the one given by Wagner.

It also explains the extent to which government expansion depends on the attitudeof the government, that is, the degree of the government’s response to the people’svoices may depend on, say, people’s trust in their government, the stability of thegovernment, and the social norms of families and communities. Such a considerationallows us to explain the difference in the size of governments among developedcountries.

In the following part, we will provide two economic theories to support our abovehypothesis and then provide some historical evidence that is consistent with it.

2.3.1 Economic Theories Supporting Our Hypothesis

Communities such as families and local communities have played important roles inmitigating various market failures. In economics, the government has been expectedto play the role of solving such problems. To understand the relationship betweengovernments and communities, two theories, known in economics as the “Coasetheorem” and “crowding-out theorem,” are useful.

2.3.1.1 Coase Theorem

Our society is full of externalities. For example, there are public spaces in families andlocal communities that are maintained by the efforts and cooperation of the membersof communities. Coase (1960) pointed out that even if there are externalities, peoplewill discuss and bargain to reduce inefficiencies associated with externalities. If thebargaining cost is very small under well-defined property rights in the legal system,bargaining to solve the problems of externalities will reduce the inefficiency of thesociety without government intervention. The idea in Coase (1960) is well illustratedin the following theorem.

15Wagner (1835–1917) is a German fiscal economist who proposed a law of government expansionin a market economy.


Proposition 2.2 (Coase Theorem) If there are no bargaining costs (transactioncosts) among the people involved in a problem of externalities, then efficiency willbe achieved through bargaining (without the governments’ intervention).

Theproof for the theorem is easy onceweunderstand the definition of “efficiency.”When there is inefficiency due to some externality, it means that there is room forimprovement of someone’s welfare without deteriorating the welfare of others. If so,welfare improvement will be realized through negotiation. Bargaining will continueuntil such improvement is impossible, and thus efficiency will be reached when thebargaining cost is zero.

This theorem is usually discussed without explicitly relating to the communities,but, it represents the criticism against traditional welfare economics, a là Pigou(1920), which emphasizes the necessity of government intervention to solve theproblems of externalities. In the theorem, there is some consideration of the non-market and non-government allocation mechanisms to improve the efficiency of thesociety through voluntary bargaining among the people. It suggests the importanceof communities in improving the efficiency of our society.

If we use the term “communities,” the theorem suggests that, as long as the bar-gaining cost in a community is small, efficiency can be achieved through cooperationamong the members. When the community tie is strong, the bargaining cost will belower and many problems can be solved within the community. However, whencommunities are weak they will fail to improve efficiency due to the high costs ofbargaining, and thus government intervention will be demanded. The theorem isuseful in understanding the functions of communities and the relationship betweencommunities and the governments.

2.3.1.2 Crowding-Out Theorem

When the government tries to solve market failure problems, it may crowd out goodsand services provided by families and local communities to mitigate the marketfailures. Such effects, known as the crowding-out effects of government policies,may weaken communities and lead to the expansion of the government.

There are several versions of the crowding-out theorem. Here, we present theversion in which the government’s provision of public goods crowds out the privateprovisionof public goods,where “public goods” refer to the goods and serviceswhosebenefits can be enjoyed by everyone for free.16 Donations and voluntary activitiesto clean up public spaces or to help poor people are some examples of public goodsthat private people often provide.

Proposition 2.3 (Crowding-Out Theorem) If the government provides public goodsthat people voluntarily provide, people will reduce their voluntary provision.

16To be more precise, in economics, public goods are defined as goods and services that are non-excludable and non-rival.


The extreme result concerning this theorem is that the government can completelycrowd out the private provision of public goods in the sense that an increase in publicgoods provided by the government will be completely offset by the reduction inthat provided by the private sector. This theorem is sometimes referred to as the“neutrality theorem” because the increase in the provision of public goods by thegovernment will have no effect on its total level. We shall provide more discussionon this theorem in Sect. 6.2.3.

Although the possibility of complete crowding-out is low, there is much evidenceof the private sector reducing its provision of public goods when the governmenttries to increase the public goods. This theorem provides a theoretical foundation forexplaining the causal relationship between the expansion of the government and thereduction of activities by the communities.

2.3.2 Evidences in History

Based on the economic theories above, wewill show some historical evidences aboutthe structural transformation of our society to support our hypothesis, as illustratedin Fig. 2.1.

2.3.2.1 Development of the Market System and Weakeningof Communities

Let us first see how the expansion of themarket economy affected the family, which isthe most important community. Family, as a social institution, has existed universallyin the long history of human beings. There are extensive studies on the structure offamilies in different places and periods.

Before the emergence of markets, it was not easy for most people to leave com-munities where necessary goods and services must be obtained from the communitymembers. Building cooperation and trust with community members was very impor-tant. After the development of the market system, where people can anonymouslysell and buy goods and services including labor, people started to leave their familiesand local communities to make a living.

Such an economic factor, that is, the development of the market system, can beviewed as an important force that weakened the traditional communities includingfamilies. Many scholars, such as Durkheim, Weber, and Marx recognized the impor-tance of the Industrial Revolution, which created factory workers in urban areas, inunderstanding the transformation of families and local communities.17

17See, for example, Morrison (2006).

http://dx.doi.org/10.1007/978-4-431-55909-2_6


0

10

20

30

40

50

60

70

80

90

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

(%) Austria Denmark Finland France Germany Hungary Iceland Ireland Italy Japan Netherland New Zealand Norway Porland Spain Sweden Switzerland UK USA

Fig. 2.2 Share of the elderly living with children. Source: Data published by NIPSSR for Japanand Sundström (1994) for other countries

Ruggles (1987, p. 13), for example, points out that “[t]oday, the importance ofeconomic influences on the evolution of the family is rarely challenged,” and sum-marizes the standard explanation as follows18:

When production shifted from family to factory, the family economy was destroyed. Thenew industrial system demanded a flexible and mobile family: the stripped-down nuclearfamily prevailed because it was functionally adapted to new economic realities. As part ofthis process, the elderly lost their productive role and became isolated form society and fromthe family.

The change in families suggested above has been accelerated after the SecondWorld War when we observed the very rapid expansion of the market economy. Oneof the most important functions of families has been the provision of care to frailelderly parents. Living together with elderly parents is one way of playing such arole efficiently. Figure 2.2 depicts the trends in the share of the people above 65yearsof age who live with their children in developed countries. It shows that in mostdeveloped countries, many elderly people were used to living with their childrenafter the Second World War.19 However, the share has been declining in most of thecountries.

18Although there are various criticisms against such an explanation, including Ruggles (1987),which indicates that the effects of the development of the market system on the family structure aremore complex than the simple story stated above, the basic explanation above is still convincingtheoretically and empirically as argued by many scholars.19Sundström (1994) who collected data for Fig. 2.2 points out that the decline in the share makesit more difficult for the frail elderly to receive care from their families. However, he also points outthat when the elderly really need care from their children they tend to live close to their children.Thus, the data for the elderly living together in Fig. 2.2 may be overestimating the weakening offamilies in developed countries.


The development of themarket system has also weakened local communities. Themost well-known case is the one in the period of enclosure in 18th-century England.The enclosure of open fields, where farmers had to cooperate with each other, meantthe dissolution of agrarian communities.

It is often explained that the enclosure dissolved the agrarian communities,expelled farmers from the farming lands, provided cheap labor to be employed by thecapitalists in urban factories, and contributed to the economic development.However,the close examination of the open fields suggests that the enclosure is a consequenceof the development of markets, as explained by Allen (2001, pp. 66–67)20:

The open field village fostered solidarity amongst the tenants. ... Open field farming strength-ened bonds between peasants and made them wary of lords. While the social capital of theopen field village raised productivity in that context, it was counterproductive when the yeo-man were eliminated and replaced by tenants at will. ... the great estates that emerged in theeighteenth century required that each farmer identify his fortune with the estate rather thanwith his fellow villagers. ... Enclosure was the result. It was not, therefore, that enclosureoverthrew community and createdmarket capitalism.Rather, the reversewas the case:marketcapitalism undermined the open field community from within and precipitated enclosure.

The explanation above may be well understood if we compare the agrarian com-munity to cooperative labor unions, which have contributed to the increased produc-tivity of the workplace with good relationships with the employers (landlords). Aslong as the workers (yeoman) have long-term employment, the cooperation amongworkers and the employer will work well. When the labor mobility became high asthe market economy developed, workers started to find better opportunities in themarkets, the relationship among workers and employer weakened and its durationshortened, and employers needed to guarantee high returns to the workers (tenants)who would work for them only at their will. Solidarity among workers in the laborunions can sometimes be considered as threats to the employer when the relationshipgets worse. As labor markets and markets for the agricultural products developed, itis understandable that the landlords decided to abolish the open fields and thus theircommunities.

In Japan, the dissolution of local communities in the prewar periods can be wellillustrated by the decomposition of the commons known as “iriaichi” in Japanese. Inthe villages, anyone could access the iriaichi for grass, trees, water, fish, and animals,under formal or informal rules. The rules were clearly set and if people broke them,the severe punishment of “murahachibu” was imposed on them by the village (c.f.Sect. 6.3.3). Members of the local communities were strongly tied by the use and therules for the use of commons.

20The argument need not be denied because the weakening of communities can indeed contributeto the development of the market economy. If we consider such an effect, we should add an arrowin Fig. 2.1 from “Weakening of Communities” to “Development of the Market Economy” andindicate the cyclical expansion of themarket and theweakening of communities.What the followingquote stresses is that the development of the market system is the first cause of the weakening ofcommunities.

http://dx.doi.org/10.1007/978-4-431-55909-2_6


The use of commons has changed drastically after the economic development thatstarted in the late nineteenth century when theMeiji government opened Japan to theworld and made various political and economic reforms. The common lands weredecomposed and distributed to the members of the community. One of the reasonsfor the decomposition of commons was the response to the new government’s taxreform that required clearer definitions of property rights. However, many historianspoint out that it was also a response to the development of the market economyin which farmers sell their products in the markets to earn monetary income. Insuch circumstances, many villages faced difficulties in keeping the common landsas commons and decided to decompose the lands into private farms in the villages.For example, Kainou (1943) wrote as follows:

· · · the decomposition of iriaichi to the villagers and the purely private economic manage-ment of the lands and forests may have contributed to an increase in the income of eachhousehold. However, we cannot ignore the fact that such a change accelerated the dissolu-tion of the villages because they lost a kind of precautionary wealth that the whole villageused to have.

Notice that many Japanese communities had to decompose the commons to indi-viduals to respond to the development of the market economy in which propertyrights need to be well-defined; additionally, the dissolution of the commons, andthus the development of the market economy, weakened the communities.

The weakening of families and local communities in Japan continued with thefurther development of the market economy after the Second World War. Yoshiharaand Wada (1999, Chap.18) describes the increase in social problems after 1955 asfollows:

The distortion of rapid economic growth began to be visible, caused various social problems,and had serious impacts particularly on the lives of the elderly. Due to the shrinking share ofprimary industry and the outflow of people, especially the young, from rural to urban areas,the elderly are left behind in rural areas, resulting in depopulation in those areas.

This clearly describes how economic development changed families in rural areas,thereby rendering the lives of the elderly uncertain. Economic development graduallychanged the family structure in the late 1950s.

2.3.2.2 Weakening of Communities and Expansion of the Government

When communities are weakened with the development of the market economy,mutual assistance within communities will shrink andmanymarket failure problems,which were earlier mitigated by the communities, will occur.

In Japan, given the increasing awareness of the problems affecting the elderly, theLiberal Democratic Party (hereinafter LDP), which was the majority party between1955 and 1993with its important voting blocks in rural areas, responded to the voicesof the elderly. Yoshihara and Wada (1999) state that “with the increasing attentionbeing paid to the problems affecting the elderly, the first thing that showed up wasthe universal national pension system.”


Takeda (2005, p. 117) also describes the changes in the late 1950s as follows:

High growth had destabilized the social units through mobilization of population causedby the transformation in the industrial structure that had, in place of the public purse, beenproviding support for those in need, namely families and communities. As a result of thistransition, the nation-state was required to compensate for functions that families and com-munities normally played.

In 1961, the Japanese government introduced universal public pension and uni-versal public health insurance. Since then, everyone in Japan has been covered bysocial insurance. However, the levels of social security payment were not high in the1960s, and as such, the elderly demanded more social security.

In 1973, the Japanese government declared the “first year of welfare state.” Itincreased the pension payment significantly and made healthcare services for theelderly free. After the declaration of the welfare state in 1973, the Japanese govern-ment continuously increased its social security expenditure.

In western countries, many scholars of sociology have made an argument thatthe expansion of the government has weakened families and communities.21 Forexample, Coleman (1990, p. 607) states the following.

Modern society differs extensively from the village social structure described above. Thefamily is no longer the primary welfare unit. Responsibility for caring for the aged has beentaken over by the society as a whole. Responsibility for medical care has been taken over bythe state and employers. ... The family, greatly shrunken in size and function, has becomeincapable of carrying out many welfare activities. ... The state, greatly expanded in size andfunction, has taken on most welfare functions.

The quote above also reveals that the expansion of the government has crowdedout roles of families and transformed families. We will discuss this further.

2.3.2.3 Expansion of the Government and the Weakeningof Communities

The causal relationship, “weakening of communities → expansion of the govern-ment,” implies that the demand for the welfare services of the government will go upwhen communities are weakened. To what extent does the government respond tothe demand? The size of supply depends on the political decisions in each country.For example, a fairly big expansion of the government occurred in Germany, Eng-land, and Japan. On the other hand, the U.S. government has not responded to sucha demand strongly.

Once the expansion of the government occurs, it will have further effects onthe communities, which is depicted by the causal relationship “expansion of thegovernment → weakening of communities”. History has witnessed many instancesof such a relationship.

21 For example, Cliquet (2006) summarized the changes in families and demography in westerncountries.


For example, Myrdal (1941, p. 5) suggested that the expansion of the governmentworked to weaken the functions of families and lowered the motivation for marriageand children:

The protective function of the family likewise plays a declining role. Hospitalization, insti-tutionalization, and social security measures are taking place of family care for the old andsick. This tends, on the one hand, to decrease the value to the individual of marriage andchildbearings.

The weakening of families and local communities by the expansion of govern-ments has been witnessed by Japanese history as well. We have already seen in Fig.1.5 in Chap.1 that the marriage rates and fertility rates indicated a continuous declineafter 1974, just after the government’s declaration of the welfare state in 1973.

Figure 2.2 indicates that the share of the elderly living with their children, whichwas traditionally high at around 80%, has also started to decline. More and moreelderly people have started to live independently. The risk of their falling into povertybecame greater, particularly when the number of children became smaller.

The expectation of being dependent on children for old-age security continuedto decline as social security expanded (c.f. Fig. 11.1).22 Development of the marketeconomy and the expansion of social security replaced the roles of children, whichled to the decline in fertility rate.


In this chapter, we first considered the functions and limitations of the markets,governments, and communities and then discussed the interactions among thesethree important institutions, to set up an analytical framework for discussing thetransformation of the social structure and the roles of the government in Japan.

We have presented a hypothesis that the development of markets has weakenedcommunities, which demanded the expansion of the welfare role of governments.This in turn further weakened communities when governments responded to thedemand. Then, we forwarded two economic theories and various historical evidencesthat are consistent with our hypothesis.

In the next four chapters, from Chap.3 to Chap.7, we make our argument abovetheoretically sophisticated, and then, in the last part of the book, we apply our frame-work to explain the social transformation and public policies in Japan. Before wepresent the detailed arguments in those chapters, we would like to summarize thebasic viewpoints on how Japan’s public policies should be redesigned to properlyrespond to the transformation of the Japanese society.

22See, for instance, Ogawa andMatsukura (2007, Fig. 10). It also indicates that the norm of childrentaking care of their aged parents changed considerably. The share of thosewho believed that childrentaking care of their aged parents is a “good custom” or “natural duty” declined rapidly after 1986.

http://dx.doi.org/10.1007/978-4-431-55909-2_1

http://dx.doi.org/10.1007/978-4-431-55909-2_11

http://dx.doi.org/10.1007/978-4-431-55909-2_3

http://dx.doi.org/10.1007/978-4-431-55909-2_7


The basic viewpoints have already been indicated in the following statements byMyrdal (1940) who considered public policies to properly respond to the transforma-tion of the Swedish society more than 70years ago.23 The term “social policy” hererefers to public policies, such as social security, public support for raising children,labor policy, and education policies, that closely affect the social lives of people.

[W]hen we transfer the emphasis from curative to preventive social policy we must ceaseto regard the costs of social policy as unprofitable, nonproductive consumption costs on thepublic budget. . . . The tendency, however, is clearly visible in the present activity in thesphere of social reform in Sweden; and the population argument has been the most potentforce behind the speeding up of this important transition in social policy from curative toprophylactic, from consumptive to productive, from the aid of the needy to broad cooperationamong all the people. The solution is: certain services free for all children and for all familiesin all economic classes,with nomeans test, planned as a rational cooperation between citizensin a democratic society on the principle of national solidarity, financed through the nationalbudget, . . .. The intended redistribution is carried out in kind instead of in cash. . . . To alarge extent the needs of families and children can bemet by engaging otherwise unemployedor underemployed factors of production in work and the whole program engineered in thedirection of stabilizing the economic system.

Needless to say, Japan today is quite different from Sweden 70years ago, andwe do not accept all the above proposals as desirable. However, the basic idea ofreforming the social policies “from curative to prophylactic” and “from consump-tive to productive” is surprisingly suggestive in redesigning Japan’s public policiesbecause many social problems in Japan have been caused by curative and consump-tive public policies in Japan.

Gunner Myrdal found the instability of the Swedish socioeconomic system in hisanalysis of the declining fertility rates in Sweden. To determine policies to strengthenits stability, he recognized a simple fact that “[T]he principal part of the wealth of anation always lies in the quality of its population.”

His conclusionwas that, proper social policies to strengthen theSwedish economicsystem should not be curative and consumptive, as in the traditional poor-reliefpolicies, to protect the lives of the frail, but should be prophylactic and productiveto improve the quality of life. In designing such policies, he realized that the in-kindpolicies are more effective to pursue the goals than cash-transfers, which may not beused to improve the quality of life.

As we have seen in this chapter, the development of the market economy hasweakened the families and communities. As a consequence, elderly and poor peoplebecame more frail and the Japanese government expanded the social security systemas a curative policy and the consumptive policies to help the elderly and the poor. Suchcurative policiesmade the people dependent on the government and furtherweakenedfamilies and local communities in Japan. The proper public policies to stabilize andstrengthen the Japanese society should be, as Myrdal pointed out, prophylactic andproductive social policies to empower the people and the communities.

In Japan, the elderly are increasing and the young generation is rapidly shrinking,while the public debt has already exceeded 200% of the GDP. In such a circumstance,

23Quotes from Myrdal (1940, pp. 206–210).


0

10

20

30

40

50

60

70

80

90

100

1519

2024

2529

3034

3539

4044

4549

5054

5559

6064

over

65

( )

Japan (2012) Denmark Finland

France Germany Norway

Sweden Japan(1980)

Fig. 2.3 Patterns of female labor participation. Source NIPSSR and Statistics Bureau

to make the country sustainable, Japan should aim for a society in which potentialabilities of the people will be realized and used to generate high income from whichtaxes are collected to pay the high level of public debt. The keyword is “labor.”When Myrdal said that “the principal part of the wealth of a nation always lies inthe quality of its population” to strengthen the economy, he must have had a visionwhere everyone works for himself/herself and for the society.

Is Japan close to being such a society? Fig. 2.3 indicates the pattern of female laborparticipation by age. It indicates that the female labor participation rate in Japan ismuch lower than, say, in the Nordic countries. Most of the women who are not in thelabor markets since their mid-twenties work at home as housewives. The Japanesepolicies have provided various favorable measures, such as the exemption of pensioncontribution, to dependents of the employed workers. Almost ten million womenhave remained in the category of people who can receive basic pension without theirnor their husbands’ paying the pension contributions (c.f. Sect. 8.3.1).

Under the social norm of sexual division of labor exemplified by such policies(c.f. Fig. 11.2 in Chap.11), female labor participation has been low not only in thetwenties and thirties, but in almost all ages. Although it has been increasing, it is stilllower than that in Nordic countries, as depicted in Fig. 2.3. Such behavior has beendriven by social norms and public policies based on these norms. Thus, whetherpotential abilities of Japanese women are fully utilized is questionable. If highlyeducated women participate in the labor market more, it is very likely that the per-capita income, and thus the tax revenue will increase in Japan, greatly contributingto the sustainability of the Japanese economy.

In the process of rising female labor participation, as Myrdal (1940, p. 210)points out, “the needs of families and children can be met by engaging otherwiseunemployed or underemployed factors of production in work and the whole programengineered in the direction of stabilizing the economic system.”

http://dx.doi.org/10.1007/978-4-431-55909-2_8

http://dx.doi.org/10.1007/978-4-431-55909-2_11

http://dx.doi.org/10.1007/978-4-431-55909-2_11


Such a new socio-economic system will change the traditional way of living inwhich families provide most of the care for the elderly and infants. However, it doesnot imply that all care will be out-sourced. Families will continue to function asthe main institution for providing love and care to members. The prophylactic andproductive social policies will provide more support to such families.

Needless to say, there are many young and elderly people, besides women, whocan contribute to the labor market. It is important for the society to develop andutilize their potential abilities to overcome the current serious problems of Japan.The prophylactic and productive social policies will be also effective in supportingthem.

References

Akerlof, G. (1970). The market for lemons: Quality uncertainty and the market mechanism. Quar-terly Journal of Economics, 89, 488–500.

Allen, R. (2001). Community and market in England: Open fields and enclosures revisited. In M.Aoki & Y. Hayami (Eds.), Communities and markets in economic development. Oxford: OxfordUniversity.

Aoki,M.,&Hayami,Y. (Eds.). (2001).Communities andmarkets in economic development. Oxford:Oxford University.

Arrow, K. J. (1950). A difficulty in the concept of social welfare. Journal of Political Economy,58(4), 328–346.

Becker, G. (1973). A theory of marriage: Part I. Journal of Political Economy, 81(4), 813–846.Becker, G. (1974). A theory of marriage: Part II. Journal of Political Economy, 82(2), S11–S26.Becker, G. (1985). Human capital, effort, and the sexual division of labor. Journal of Labor Eco-nomics, 3, S33–S58. (Reprinted in Becker, G. (1993). A Treatise on the Family, Supplement toChapter 2.).

Becker, G. (1993). A treatise on the family. Cambridge: Harvard University Press.Cliquet, R. (2006). “Major trends affecting families in the new millennium? -Western Europe andNorth America-” in United Nations (pp. 1–26). Major Trends Affecting Families: A BackgroundDocument.

Coase, R. H. (1960). The problem of social cost. Journal of Law and Economics, 3, 1–44.Coleman, J. S. (1990). Foundation of social theory. Cambridge: Harvard University Press.Esping-Andersen, G. (1999). Social foundations of postindustrial economies. Oxford: Oxford Uni-versity Press.

Hayami, Y., & Goto, Y. (2005). Development economics: From the poverty to the wealth of nations(3rd ed.). Oxford: Oxford University Press.

Pazner, E., & Schmeidler, D. (1974). A difficulty in the concept of fairness. Review of EconomicStudies, 41, 441–443.

Grossbard-Shechtman, S. (1984). A theory of allocation of time in markets for labor and marriage.Economic Journal, 94, 863–82.

Grossbard-Shechtman, S. (1993). On the economics of marriage: A theory of marriage, labor, anddivorce. Boulder: Westview Press.

Kaino, M. (1943). Studies on Iriai. Nihon Hyoron Sha.Kotlikoff, L.,&Spivak,A. (1981). The family as an incomplete annuitiesmarket. Journal of PoliticalEconomy, 89, 372–291.

Morrison, K. (2006). Marx, Durkheim, Weber. Thousand Oaks: SAGE Publications.Myrdal, A. (1941).Nation and family: The Swedish experiment in democratic family and populationpolicy. Cambridge: The MIT Press.

Myrdal, G. (1940). Population: A problem for democracy. Cambridge: Harvard University Press.


Ogawa, N., & Matsukura, R. (2007). Ageing in Japan: The health and wealth of older persons.In United Nations, United nations expert group meeting on social and economic implicationsof changing population age structures (199–220). 31 August–2 September, 2005, Mexico City,New York.

Ostrom, E. (1990). Governing the commons: The evolution of institutions for collective action.Cambridge: Cambridge University Press.

Pigou, A. C. (1920). The economics of welfare. Basingstoke: Macmillan.Ruggles, S. (1987). Prolonged connections: The rise of the extended family in nineteenth-centuryEngland and America. Madison: University of Wisconsin Press.

Stiglitz, J. E., & Rosengard, J. (2015). Economics of the public sector. New York: WW Norton &Co.

Sundström, G. (1994). Care by families: An overview of trends. In OECD caring for frail elderlypeople: New directions in care. OECD Publishing.

Takeda, H. (2005). The political economy of reproduction in Japan: Between nation-state andeveryday life. London, New York: RoutledgeCurzon.

Yamashige, S. (1997). Fairness in markets and government policies: A weak equity criterion forallocation mechanisms. Hitotsubashi Journal of Economics, 38, 61–78.

Yoshihara, K., & Wada, M. (1999). History of the Japanese healthcare insurance system. ToyoKeizai Shinposha. (in Japanese).

Wagner, A. (1883). Three extracts on public finance. In R. A. Musgrave & A. T. Peacock (Eds.),Classics in the theory of public finance (1958). Basingstoke: MacMillan & Co.

Weiss, Y. (1997). The formation and dissolution of families: Why marry?Whomarries whom? Andwhat happens upon divorce? In M. Rosenzweig & O. Stark (Eds.), Handbook of population andfamily economics, 1A (pp. 81–123). Amsterdam: Elsevier Science.

Chapter 3Introduction to Decision Theories

[O]ur aim is primarily to show that there is a rigorous approachto these subjects, involving, as they do, questions of parallel oropposite interest, perfect or imperfect information, free rationaldecision or chance influences.

von Neumann J. & Morgenstern, O. (1943, Preface)

3.1 Introduction

In this section, we provide a brief introduction to decision theory to understand ourtheoretical arguments on the social transformation in Part II.1

In the next section, we learn economic theories of utility maximization problemand decisions under uncertainty. In analyzing our daily lives, considering uncertaintyis indispensable. In particular, in analyzing families and communities, we need tohave a good understanding of how we prepare for various risks in our daily lives.

In Sect. 3.3, we will make a brief introduction to game theory to analyze theinteractions among the people. As the quotation from the abovementioned seminalbook on the game theory suggests, game theory is a really powerful analytical tool tounderstand human behavior. Hence, it is important to have a basic understanding ofgame theory to understand non-market transactions among family and communitymembers. At the end of this section, we present a game-theoretic explanation ofthe tragedy of commons and various ways to mitigate the tragedy, which are usefulin understanding functions and structures of families and communities. Section 3.4concludes the chapter.

1Readers who are not interested in the mathematical theories can skip this chapter and Part II toread chapters in Part III where we discuss the public policies used to cope with the challengesof the social transformation in Japan and the lessons that other countries can learn from herexperiences.


47

48 3 Introduction to Decision Theories

3.2 Utility Maximization

In this section, we first explain a basic utility maximization problem and decisionsunder unceainty to discuss how we deal with various risks in our lives.

3.2.1 Utility Function

A useful way of analyzing the behavior of people is “utility,” which is a jargon ineconomics for “satisfaction” and “happiness,” can be expressed using a mathematicalfunction. The idea is that if we express a list of factors that affect our utility by Xthen our satisfaction can be expressed by the function U (X).

As an example of X , let us consider the list X = (Leisure, Consumption), where“leisure” and “consumption,” in economics, imply the time for leisure and the con-sumption of goods and services, respectively. We use the symbols L for leisure(time) and C for consumption. When we do not have any leisure and consumption,an increase in each factor will increase our utility. We assume that the functionU (X)

is an increasing function of X = (L ,C). The rate of increase in utility is likely to fallwith more leisure and consumption because we feel gratified. The rate of increasein utility, that is, the increase in utility with an additional unit of factor is referredto as “marginal utility.” We assume that marginal utility is decreasing in leisure andconsumption.

In the language of mathematics, a marginal change of a function, which representsthe slope of the function, can be expressed as the derivative of that function. Whena function is dependent on two or more factors, it is sometimes useful to calculatethe “partial derivative” of the function, which is the derivative of the function withrespect to the concerned variable, given that other variables are fixed. For example,the partial derivative of the utility function U (L ,C) with respect to L is simply the

derivative of the function with respect to L given C is fixed. It is denoted by ∂U (L ,C)

∂Lor UL(L ,C) and measures the increase in utility when leisure is slightly increased,while consumption is fixed.

3.2.2 Utility Maximization Under Budget Constraint

When we try to increase our utility, we usually need to spend more. The total expendi-ture is limited by our income. Hence, it is useful to consider the utility maximizationproblem subject to the budget constraint. A simple budget constraint correspondingto the maximization problem above can be written as pC = w(H − L) + I , wherep is the price of consumption goods, w is the wage, H is the maximum time onehas for leisure, and I is unearned income. The time for labor is assumed to be thedifference between H and L , that is, the total time and the time for leisure, L ≤ H .

3.2 Utility Maximization 49

The budget constraint can be rewritten as pC + wL = wH + I . The term wL inthe equation suggests that we purchase the time for leisure at pricew. We can interpretthe wage, w, as the price for leisure because we will lose w when we spend one unitof time on leisure, that is, the time for earning wage w will have to be forgone. Thevalue of the opportunity lost to take an action is known as the “opportunity cost” ofthe action. The opportunity cost of leisure is equal to w, which can be seen as theprice of leisure.2

Under this budget constraint, the utility maximization problem can be expressedas follows:

maxL ,C

U (L ,C) s.t. pC + wL = wH + I

The expression maxL ,C U (L ,C) means maximizing the function U (L ,C) withrespect to L and C , and the term “s.t.” is an abbreviation for “subject to.” Hence, theproblem requires us to “maximize U (L ,C) by adequately choosing L and C underthe constraint pC + wL = wH + I .”

Such a constrained maximization problem can be solved by defining the following“Lagrange function.”

L ≡ U (L ,C) + λ(I + wH − pC − wL)

where λ is known as the Lagrange multiplier. The “first order condition” for themaximization problem is given by taking the partial differentiation3 of the Lagrangefunction with respect to L , C , and λ, and setting them equal to zero. Namely, bysolving the following system of three equations

∂L

∂L= 0 : UL − λw = 0 (3.1)

∂L

∂C= 0 : UC − λp = 0 (3.2)

∂L

∂λ= 0 : pC + wL = wH + I (3.3)

we can get the value of the three variables L , C , and λ.4

The following equations that can be obtained by solving (3.1) and (3.2) haveimportant economic meanings.

UL

UC= w

por

UC

p= UL

w(3.4)

2One of the most important perceptions in economics is that our time is limited and thus has a value.The monetary value of time for an individual can be measured by his/her wage.3Differentiation of a function is an act of calculating the derivative, that is, the slope of the function.4In order to find out whether or not the values are a solution to the maximization problem, we needto check the “second order condition.” See Remark 3.3.


Fig. 3.1 Utilitymaximization

0

C

L

Cf

Ce

fe

g

JI0

S

HLe L f

b

a

d

M

300

200

100

The term UCp represents the increase in utility from spending one more yen for con-

sumption. On the other hand, ULw represents the increase in utility from spending one

more yen for leisure. If UCp > UL

w then it is better to spend more on consumption andless on leisure. If the direction of the inequality is opposite, then it is better to spendless on consumption and more on leisure. In either case, the choice is not optimal.Hence, when the utility is maximized, the equality UC

p = ULw must hold as in (3.4).

Remark 3.1 (Illustration of the Utility Maximization Problem) As the budget con-straint can be written as C = −w

p L + wH+Ip , it can be represented by the line with

an intercept of wH+Ip and a slope of −w

p . As consumption cannot be negative andleisure cannot be negative and exceed the maximum time, we have the followingadditional constrains: C ≥ 0 and H ≥ L ≥ 0. Hence, the budget constraint can bedepicted by the line segment aH or bJ in Fig. 3.1, where the latter, bJ , indicates thebudget constraint with a positive unearned income I0. Consumers need to choose abundle (L ,C) of leisure and consumption within the budget constraint. To find oneunder which the utility is maximized, we try to draw an image of the utility functionU (L ,C). As it assigns a value for each bundle (L ,C), the function should look likea mountain on the L − C plane.

To express the image of the function U (L ,C) in the 2-dimensional L − C plane,we draw the contour map of the mountain. Each of the three circles in Fig. 3.1represents a contour of the mountain for a different height. The circles suggest thatthe utility mountain is the highest at the bundle S = (H, M). In economics, thecontour is referred to as the indifference curve because consumers are indifferentbetween any bundle (L ,C) on the curve, because each provides the same utility.5

5In Fig. 3.1, there is a saturation point (H, M) where the utility is maximized. Hence, the util-ity functionU (L ,C) here is not increasing any more after L ≥ H and/orC ≥ M . As consumers will


If we need not consider the budget constraint, we will choose a point S. However,we need to choose from the points on the budget constraint. It is easy to see that sucha point can be found on the budget constraint where an indifference curve touchesit (e.g., point e and f in Fig. 3.1). At the optimal points, the slope of the budgetconstraint (wp ) is equal to the slope of the indifference curve (UC

UL), which corresponds

to our first-order condition in (3.4).

Remark 3.2 (Substitution Effect and Income Effect) When we compare e and fin Fig. 3.1, we find that an increase in income increases optimal consumption andleisure. Goods and services whose demand increase as income increases are called“normal goods.” In this book, we assume that consumption goods and leisure arenormal goods. To see the effects of the change in wage or price of consumptiongoods, we compare f and g. The choice g is the utility maximizing choice on theline segment d J with its slope flatter than the budget constraint bJ .

When the wage increases and the budget constraint becomes the line segment bJ ,the change in the optimal choice from g to f can be decomposed into “g → e” and“e → f .” The latter (e → f ) can be seen as the effect caused by the increase inincome, and thus referred to as the “income effect.” On the other hand, the former(g → e) represents a shift from relatively more expensive leisure to relatively cheaperconsumption goods, and thus referred to as the “substitution effect.”

As an increase in the wage raises the price of leisure, the substitution effect is thereduction in leisure and increase in consumption. The income effect increases bothleisure and consumption. Hence, the increase in the wage will definitely increaseconsumption but may or may not increase leisure.

Remark 3.3 (Optimization Problem and First and Second Order Conditions) In thisbook, we deal with the maximization problem without constraints in some cases. Forexample, consider the problem to maximize the function f (x) with respect to x .When there is a maximum, the shape of the function f (x) would be an inverted-U.The property of the peak of the curve is that we cannot go beyond it. Mathematically,it means that the slope of the function f (x) becomes zero. When the function isdifferentiable and f ′(x) denotes the derivative of the function ( d f (x)

dx ), f ′(x) = 0must hold at the maximum. This is known as the “first order condition.”

However, this condition is also satisfied at the minimum point, because at thebottom we cannot go further. The difference between the peak and the bottom is thatthe slope of the curve decreases from a positive to negative value around the peak,and increases from a negative to positive value around the bottom. When the slopedecreases, as in the former case, the rate of change of the slope f ′(x) is negative.

The rate of change of the slope, defined by d f ′(x)dx or d2 f (x)

(dx)2 , is known as the secondderivative. It is often denoted as f ′′(x). The argument above implies f ′′(x) < 0 atthe maximum and f ′′(x) > 0 at the minimum. These conditions are known as thesecond order conditions for the maximum and the minimum.

(Footnote 5 continued)not choose the bundle (L ,C) beyond the saturation point, only the part below the saturation pointof the indifference curves is usually plotted.


Remark 3.4 (Budget Set and Equal Opportunities) In this section, the budget con-straint in the utility maximization problem is expressed as an equation; however, itcan be expressed as the budget set as follows. Let S be the set of the bundles (L ,C)

which satisfies the inequality pC + wL ≤ wH + I . The set can be mathematicallyexpressed as S = {(L ,C) ∈ R2+ : pC + wL ≤ wH + I, L ≤ H}.6 Then, the utilitymaximization problem can be simply expressed as max(L ,C)∈S U (L ,C).

The budget set S corresponds to the triangle 0Ha or the trapezoid 0HJb. If theutility function is increasing in each argument, individuals will choose only from theupper-right area of the budget set, where the equality holds in the inequality for thebudget set. Hence, under such an assumption, the budget set in effect can be writtenas an equation.

The budget set is the “choice set” when we make decisions. As we have suggestedin the Remark 2.1 in Chap. 2, if such a choice set is the same for all individuals,we can regard that equal opportunities are guaranteed. When equal opportunitiesare guaranteed, even if individuals make different choices no one will envy othersbecause the choices of others were available to everyone. This argument suggestsa close relationship between “equal opportunities” and “envy-free allocations” (c.f.Yamashige 1997).

Unfortunately, such equal opportunities will not be achieved in the marketsbecause the ability to earn labor income, reflected by the wage, is different for indi-viduals and thus the shape of the budget sets differs as well.

3.2.3 Decision Under Uncertainty

In our daily lives, uncertainty plays a very important role. If we fail to prepare wellagainst various risks in our lives, we can easily fall into poverty (c.f. Sect. 9.2.1).The three institutions that we are interested in here; markets, communities, andgovernments, have played important roles in sharing risks that individuals cannotfully hedge by themselves. To analyze the roles of these institutions, we would liketo deepen our understanding on our decisions under uncertainty.

3.2.3.1 Risk Aversion

Let us consider the simplest case of uncertainty in which there are two states, state-0and state-1, that occur with probability π0 and π1, respectively (π0 + π0 = 1). Wedefine the “fair lotteries” as the lotteries whose mathematical expectations are zero.For example, if one gets one million yen under state-0 and loses one million yenunder state-1, and if a lottery chooses each state with the same probability, then it isa fair lottery. Let us call the lottery L1M .

6The mathematical symbol ∈ R2+ means “in the two dimensional space of non-negative realnumbers”.

http://dx.doi.org/10.1007/978-4-431-55909-2_2

http://dx.doi.org/10.1007/978-4-431-55909-2_9


An interesting nature of the human beings is that most of us show tendency toavoid such fair lotteries. For example, when we have one million yen and are askedwhether or not we want to have the fair lottery above, most of us answer “no.” Theanswer usually does not change even if we make the lottery a little more favorable,for instance, by adding ten thousand yen to the prize so that the expected value ofthe lottery becomes positive.7

Although it is known that there are “risk-lovers” who would love to challenge thefair lottery, most people display “risk-aversion,” which is the tendency to avoid fairlotteries. It is interesting to know that people show similar attitudes towards riskseven if their preferences are generally diverse. In economics, such an attitude can bewell explained by the expected utility theory and the “law of decreasing marginalutility.”

To illustrate the idea, let u(W ) be the utility function dependent on wealth W(million yen). Assume that the marginal utility from wealth is decreasing.8 In thiscase, the increase u+ in the utility from gaining one million yen should be lowerthan the decrease u− in the utility from losing one million yen because the marginalutility is decreasing. Hence, even if the expected monetary value of the fair lottery(0.5 × 1M + 0.5 × (−1M)) is zero, the expected utility (0.5 × u+ + 0.5 × (−u−))from the fair lottery becomes negative. If we make decisions based on the expectedutility, it is better to avoid the fair lottery. Namely, under the “expected utility theory,”we usually display risk aversion because we have a utility functions with decreasingmarginal utilities.

Now, suppose that we have one million yen as our wealth and are forced to havethe fair lottery L1M . If we are risk-averse, we want to avoid it and pay a certainamount of money to be free from it. As the inequality 0.5u(2) + 0.5u(0) < u(1)

holds, there is a value ρ∗ > 0 under which 0.5u(2) + 0.5u(0) = u(1 − ρ∗) holds. Itis the maximum amount of money the individual can pay to avoid the lottery and isreferred to as the “risk premium.9”

7The answer can be changed if we were assumed to have a billion yen. In such a case, we maychallenge it. It is well known that the degree of “risk aversion” tends to decline as we have morewealth.8Such a utility function is known as the von-Neumann-Morgenstern utility function. When expen-ditures for consumption are coming from our wealth, the decreasing utility from each unit ofconsumption would mean decreasing utility from the wealth.9When we think of an individual who has two million yen and faces the risk of losing two million yenwith a probability of 0.5, ρ∗ > 0 is the maximum amount of money that he/she can pay in additionto the expected loss of one million yen (= 0.5 × 2 million yen), because 0.5u(2) + 0.5u(0) =u(2 − (1 + ρ∗)) holds. The amount 1 + ρ∗ is the maximum “insurance premium” that he/she canpay to insure against the risk, which can be decomposed into the expected loss and the risk premium.The risk-averse individual’s willingness to pay certain amount of money makes the insurancebusiness profitable.


3.2.3.2 Dealing with Uncertainty

The basic strategy to minimize the fluctuations of our wealth and thus our utility isto offset the fluctuations. We call this “insurance.” When the sources of fluctuations(e.g., sickness, accidents, economic shocks) are clear, we may want to purchasevarious insurances that pay a certain amount of money in the event of a negativeshock.10 As people in general want to avoid risks, if the “insurances” reduce thefluctuations of our utility, people will pay a certain amount of money to have them.

However, such insurances are not provided efficiently in the markets because ofincomplete information of our risks as the Remark 3.6 below explains. Hence, we tryto have various types of “insurances” in our daily lives. One of the important reasonswhy we try to have families and friends in our daily lives is to obtain “insurances”that cannot be obtained from financial market (c.f. Sects. 2.2.3 and 9.2.2).

Remark 3.5 (Insurances) When the timing of negative shocks differs among indi-viduals, by sharing and pooling the risks among the individuals, money can be trans-ferred to those who faced the shocks from those who did not, and they can havefairly stable wealth. Such risk-pooling and risk-sharing are the basic mechanisms ofinsurance. Such a mechanism fails to work if individuals suffer from negative shockssimultaneously because transfers of money cannot be done. In such circumstances inwhich ordinary insurance business fails to work, the government, which is expectedto have a longer life and more authority for better international cooperation thanprivate companies, can play an important role by engaging in intergenerational andinternational risk sharing. See, for example, Gordon and Varian (1988) and Shiller(1999), which discuss the role of governments from this point of view.

Remark 3.6 (Adverse Selection Problem) Individuals usually have more informa-tion about probabilities of negative shocks that they face (e.g., getting sick) andinsurance companies have less information about them than the individuals. Whensuch incomplete information exists, the insurance company will set the insurancepremium based on the average probabilities of the negative shocks. If it were thecase, those who purchase the insurance are the ones whose probabilities of negativeshocks are higher than the average. That is, good customers with smaller risks willnot purchase insurance, but the risky customers will purchase the insurance. Thisproblem is known as “adverse selection” (c.f. Sect. 2.2.1).

When this problem occurs, the average risk of those who purchase the insurancewill become higher than that assumed by insurance companies initially; as such,they need to raise the insurance premium. However, if they raise the insurance pre-mium, the average risk of those who purchase the insurance will rise even higher.This process of rising premium can continue until there is nobody who wants topurchase the insurance, that is, until the insurance markets fail to exist (c.f. Akerlof1970). Families and friends have played important roles to compensate for such seri-ous market failure problems in the insurance markets. When financial markets are

10We may be able to offset the loss of wealth by purchasing various types of financial assets infinancial markets.

http://dx.doi.org/10.1007/978-4-431-55909-2_2

http://dx.doi.org/10.1007/978-4-431-55909-2_9

http://dx.doi.org/10.1007/978-4-431-55909-2_2


developed to mitigate such problems, the roles of families and friends can diminish,indicating the interactions between the development of markets and the communities(c.f. Sects. 2.3.2 and 5.3.2).

Remark 3.7 (Moral Hazard Problem) Even if insurance contracts are engaged,problems of incomplete information related to the behavior of the insured still remain.The problem is that people will not make efforts in avoiding risks when they areinsured. Making little efforts under good insurance coverage is called the “moralhazard” problem as it is socially inefficient and often against our morals.11

Needless to say, insurance companies also do not want to see such behavior andthus try to prevent it. One way is to monitor the behavior of the insured and preventthe moral hazard behavior. However, monitoring is often costly. Another way is toprovide partial insurances under which the insured need to pay some money whenthey face negative shocks. Such partial insurance provides incentives for the risk-averse to take careful actions to avoid negative shocks. Most insurance contractsindeed demand payment for certain costs incurred (as a co-payment) in the event ofnegative shocks or the full cost below a certain level. Providing such partial insurancescan be a strategy to mitigate the moral hazard problem.

The moral hazard problem can occur under a system of redistributions becauseit can be seen as an insurance. It occurs when those who receive benefits from theredistributive system make little efforts to be independent from the system. Theproblem can be also described as free rider problem because recipients, withoutmaking efforts, free ride on resources provided by contributors to the redistributivesystem.12

3.3 Introduction to Game Theory

Game theory can be a very powerful tool to analyze structures and functions offamilies and local communities in which non-market transactions are actively made.In this section, we will pick up its most basic arguments to understand the discussionsof this book.13 We first explain the simultaneous-move games in which players makedecisions simultaneously and then explain the sequential games in which players canmake decisions sequentially. Any standard game can be expressed in either type oras a combination of both.

We illustrate each type of game by using “donation games,” which will be usedfrequently to explain basic ideas of our discussion in the rest of the book.

11See Sect. 3.3.5.2 for a discussion on “morals.”12In economics, the moral hazard problem refers to the problem that occurs in an insurance systemunder incomplete information. Hence, it can be seen as an example of the free rider problems. SeeSect. 9.2.1.2 for more discussion on the moral hazard problem.13As a good introductory textbook on game theory, see, for example, Osborn (2003), and foradvanced concepts, see Osborn and Rubinstein (1994) and Fudenberg and Tirole (1991).

http://dx.doi.org/10.1007/978-4-431-55909-2_2

http://dx.doi.org/10.1007/978-4-431-55909-2_5

http://dx.doi.org/10.1007/978-4-431-55909-2_9


3.3.1 Simultaneous-Move Games

Suppose that we give a 10 thousand yen bill to each of two individuals who do notknow each other and ask them if they would donate the bill in the “donation box.”If the bill is put into the box, then the money will be raised by 60% and distributedto each individual equally, independent of whether each player puts his/her moneyinto the box. This setting corresponds to the idea that the benefits of donations areenjoyed equally by all members of the society, just like pure public goods.

When both individuals donate their money, the money would be raised to 32thousand yen and each will be given 16 thousand yen. If only one individual donateshis/her money, it will be raised to 16 thousand yen. Then, the one who donated themoney will have 8 thousand yen, while the other who did not donate the moneywill have 18 thousand yen after redistributing the donated money. If none of themdonates, then each will keep 10 thousand yen. If you were asked to participate in thedonation game, will you donate the bill?

In this game, the total money that two individuals have would be highest if both ofthem donated. Hence, it is best if both of them donated. However, under an assumptionthat each tries to maximize their own money, game theory predicts that none woulddonate, and many people indeed do not donate their money in experimental studies.

To see the structure of the game more clearly, let us depict the problem in thepayoff matrix presented in Table 3.1. This is a simultaneous-move game in whichtwo players, individual-1 and individual-2, simultaneously choose from two possibleactions, “donation” and “no donation.” The actions of individual-1 are expressed inthe left column, while those of individual-2 are expressed in the first row of the table.The pair of numbers in each cell indicates the money (payoff) each individual canhave at the end of the game under each pair of decisions corresponding to that cell.The left number represents the payoff for individual-1, while the right one representsthe payoff for individual-2.

The table shows, for example, that if individual-1 donates and individual-2 doesnot donate, individual-1 will have 8 thousand yen and individual-2 will have 18thousand yen, as indicated by the pair of numbers in the right most cell of the centerrow. It is easy to confirm that the payoffs will be reversed if the pair of decisions isreversed.

Table 3.1 Payoffs under theDonation Game

Individual-1 \ Individual-2 Donation No donation

Donation 16, 16 8, 18

No Donation 18, 8 10, 10

3.3 Introduction to Game Theory 57

3.3.1.1 Dominant Strategy

Now, let us consider the game from the perspective of individual-1. If she is rational,she will analyze the bet response to each possible action of individual-2. If individual-2 chooses to donate, then individual-1 would choose “no donation” because its payoffof 18 thousand yen is larger than the 16 thousand yen she can get if she chooses“donation.” If individual-2 made no donations, then individual-1 would choose theaction “no donation” because its payoff of 10 thousand yen is larger than the 8thousand yen she can get if she chooses “donation.” That is, in the donation game,choosing “no donation” is always the best response against each possible action ofthe other player.

If there is an action that brings the highest payoff for each possible action of theother player, we call it a “dominant strategy.” If there is a dominant strategy, theplayer will choose it because the action guarantees the highest payoff, independentof what the other chooses.

Hence, game theory predicts that each player will choose “no donation” in thedonation game and each will have 10 thousand yen, which is worse than the casewhen both choose “donation” and earn 16 thousand yen. The game suggests thatindividual rational decisions can easily result in an inefficient outcome where it ispossible to increase the payoff of both individuals from 10 to 16 thousand yen. Theproblem that individual rational decisions result in inefficient outcomes is well knownin game theory, and it is referred to as the “prisoners’ dilemma.”

3.3.1.2 Nash Equilibrium

Now, let us consider a variation of the “donation game” in which each would lose 10thousand yen if neither of them donated. The payoff matrix for the new game willbe depicted as in Table 3.2.

Under the new rule, the action “no donation” is no longer a dominant strategy.To prove it, let us consider the game again from the perspective of individual-1.When individual-2 donates, then choosing “no donation” to earn 18 thousand yen isbetter than choosing “donation” to earn 16 thousand yen. When individual-2 doesnot donate, then choosing “donation” to earn 8 thousand yen is better than choosing“no donation” to earn nothing. Hence, there is no dominant strategy in this case.

In the game, there are two stable pairs of actions. First, when individual-2 chooses“donation,” it is best for individual-1 to choose “no donation;” and when individual-1chooses “no donation,” it is best for individual-2 to choose “donation.” Hence, once

Table 3.2 Payoffs underDonation Game II

Individual-1 \ Individual-2 Donation No Donation

Donation 16, 16 8, 18

No Donation 18, 8 0, 0


the pair of actions of individual-2 choosing “donation” and individual-1 choosing“no donation” is reached, there is no incentive for both individuals to change theiractions. This is the first stable pair of actions. Similarly, we can easily claim that thepair of actions of individual-1 choosing “donation” and individual-2 choosing “nodonation” is another stable pair of actions.

In game theory, such a stable pair of actions (strategies) is known as the Nashequilibrium. As the explanation above suggests, at Nash equilibrium, there is noincentive for players to change their actions because the mutual optimality of theactions is achieved. Formally, Nash equilibrium is defined as follows.

In a two-player game, let A1 and A2 be the set of actions of player-1 and player-2,respectively. The payoff of each individual determined by the pair of actions (a1, a2)

will be denoted by the functions u1(a1, a2) and u2(a1, a2). A pair of actions (a∗1 , a

∗2)

will be called a Nash equilibrium if it satisfies the following two conditions14:

u1(a∗1 , a∗

2) ≥ u1(a1, a∗2) ∀a1 ∈ A1 (3.5)

u2(a∗1 , a

∗2) ≥ u2(a

∗1 , a2) ∀a2 ∈ A2 (3.6)

These two conditions indicate that the action a∗1 is optimal given that a∗

2 is taken, andthe action a∗

2 is optimal given that a∗1 is taken; thus, the pair (a∗

1 , a∗2) satisfies mutual

optimality. Hence, if the payoff functions are differentiable, the following conditionscharacterize the Nash equilibrium (c.f. Remark 3.3).

∂u1(a∗1 , a

∗2)

∂a1= 0 and

∂u2(a∗1 , a

∗2)

∂a2= 0

3.3.2 Sequential Games

In many simultaneous-move games, as we have seen above, multiple equilibria exist.There is no general rule for the selection of the equilibrium. When the decisions aremade sequentially and not simultaneously, there occurs a possibility that one outcomeis selected.

For example, consider a variation of the donation game described in Table 3.2, inwhich individual-1 first chooses an action and then individual-2 chooses one afterobserving individual-1’s action. The new sequential game can be depicted by a gametree in Fig. 3.2.

In such a sequential game, rational decisions should be made by consideringdecisions from the last stage to the first stage sequentially. Such a way of solvingthe problem is known as “backward induction.” For example, in the last stage ofthe game described in Fig. 3.2, individual-2 is expected to make a decision for eachchoice of individual-1.

14The notation ∀a1 ∈ A1, say, means that “for any a1 in the set A1.”


Fig. 3.2 Game tree

1

2

2

Donation

No Donation

Donation

No Donation

Donation

No Donation

16, 16

8, 18

18, 8

0, 0

First, when individual-1 chooses “donation,” it is optimal for individual-2 tochoose “no donation.” This is because the payoff for individual-2 with this deci-sion is 18, as indicated by the number on the right in the pair of payoffs for thepath of individual-1’s choice of “donation” and individual-2’s choice of “no dona-tion,” and it is larger than 16 in the case when individual-2 chooses “donation” afterindividual-1 chose “donation.” Next, when individual-1 chooses “no donation,” it isoptimal for individual-2 to choose “donation” because individual-2’s payoff (8) fromthat decision is larger than the payoff (0) when individual-2 chooses “no donation.”

Having considered individual-2’s decisions in the last stages of the game,individual-1 will choose the best action through backward induction. If she chooses“donation” her payoff will be 8 because she can expect that individual-2 willchoose “no donation” in this case. If she chooses “no donation” her payoff will be 18because she can expect that individual-2 will choose “donation” in this case. Basedon backward induction, she will conclude that she should choose the action “nodonation” to get a higher payoff (18).

In this new game, there is only one reasonable path of rational decisions. Such apath of rational decisions obtained through backward induction corresponds to theoutcome of what we call the “subgame perfect equilibrium.” Subgames are smallgames within a sequential game, and a “subgame perfect equilibrium” is defined as apair of strategies that satisfies the conditions for Nash equilibrium in every subgame.Backward induction allows us to find the optimal decisions in each subgame giventhe other’s strategies, to find an outcome of the subgame perfect equilibrium.

In the game described by Fig. 3.2, there is a unique equilibrium path andindividual-1 can get a higher payoff as the first mover of the game. As individual-2 is at an inferior position in the game, suppose that he suddenly declared that hewill choose “no donation” at the last stage, independent of individual-1’s choice inthe first stage. Individual-2 has good reason to make such a claim before the gamestarts because if individual-1 believes the claim she will choose “donation” becausethe expected payoff, 8, is higher than 0, received when she chooses “no donation.”Hence, indivdial-2’s claim can be seen as a threat to get a better outcome. However,will individual-1 believe in individual-2’s threat?

Game theory predicts that she will not do so as long as it is common knowledgethat players make rational decisions. It is because once individual-1 chooses “nodonation,” individual-2 will choose “donation,” which guarantees a positive payoff(8), while the payoff will be zero if individual-2 sticks to his claim and chooses


“no donation.” Hence, individual-1 should not and will not believe in individual-2’sthreat. If a threat is consistent with rational decisions, it is called a “credible threat,”but the one made by individual-2 above is inconsistent with rational decisions, andthus is an “empty threat.”

Threats are typical strategies that players use to control other players for betteroutcomes. It is important to judge whether the threats made by other players arecredible. In other words, when we make threats it is important to make them credible.By creating circumstances under which our threats become credible, we can win thegame.

Remark 3.8 (Commitment) When there is an opportunity to make our strategy (e.g.,of threat) credible, we say that “commitment” of the strategy is possible. The oppor-tunity that allows us to make the commitment is sometimes referred to as the “com-mitment device.” In our example, if individual-2 gets a chance to buy some goodswith his 10 thousand yen before the game starts, he will not have the money fordonation and thus can commit himself to the strategy of “no donation.” The threatof “no donation” now becomes credible. The opportunity to spend his money tobuy goods before the game starts works as a commitment device. If the goods hepurchased were something that he really wanted to buy, there will be no inefficiency.If individual-2 purchased the goods just to make his strategy credible, then such abehavior can create inefficiency. It is often the case that we make inefficient decisionsjust to make our threats credible to earn higher payoffs in the game.

Remark 3.9 (Soft Budget Problem or Samaritan’s Dilemma) Let us modify thesequential game in Fig. 3.2 to explain the problem of a “soft budget constraint.”The problem is usually illustrated by using cases in which local governments orpublic enterprises accumulate debts under the expectation of bailout from upper-level governments. We illustrate the problem by using the parent-child relationship,in which we are interested.

In the Fig. 3.2, we change the name of individual-1 to child and that of individual-2to parent. The name of the child’s actions “donation” and “no donation” is changed to“effort” and “no effort”. Those of the parent’s actions “donation” and “no donation”must be changed to “no bailout” and “bailout.”

The question is whether the grown-up child makes an effort to become indepen-dent. It is a question because we can think of the possibility that the parent chooses“bailout” when the child makes no efforts to be independent.

When the child makes efforts, the parent can get a higher payoff by choosing “nobailout.” On the other hand, if the child makes no efforts the parent cannot abandonthe child and the payoff will be higher when “bailout” is chosen. Under such anassumption, the payoff structure becomes the same as Fig. 3.2 and the child makes“no effort” and the parent chooses “bailout.” Under this structure of the game, it isuseless for the parent to claim that there will be no bailout even if the child made noeffort, because the child knows that such a claim is an empty threat.

The problem that the agents do not make serious efforts to be independent whenthere are principals who can save them after they fail to be independent is known


as the problem of the “soft budget constraint” because the budget constraints of theagents are soft as they can easily expect to be bailed out. This problem was pointed outby Kornai (1979) as the prevailing problem in the centralized economies of EasternEurope.15

Needless to say, for the parents who want their children to be independent, this isa dilemma. It is called the “Samaritan’s dilemma” (c.f. Buchanan 1975), because itmay be the dilemma faced by the Samaritans who show up in the Bible as the peoplewho help the unfortunate. The problem may be seen as a moral hazard problem (c.f.Remark 3.7) because it occurs under a generous “insurance” provided by the parentsand can be mitigated if they can monitor efforts of their children and discipline themto make efforts.

3.3.3 Repeated Games

When a simultaneous-move game is repeated, it is referred to as a “repeated game.”When it is repeated infinitely, it is well known that an action that would never beentaken in the one-shot simultaneous-move game can be taken. We illustrate this resultby using the donation game described in Table 3.1. We show that the action “donation”can be chosen by the players in the infinitely repeated game.

A strategy in a repeated game is a sequence of actions, each of which is definedbased on the history up to the action. As an example of strategies under which“donation” will be chosen at the equilibrium of the repeated game, we pick twostrategies known as the “trigger strategy” and “tit-for-tat strategy”.

The action “donation” can be viewed as a cooperative action, and thus can bedenoted by C . The action “no donation” can be viewed as the defection from coop-eration and can be denoted by D. The trigger strategy can be described as follows.

(1) Start from C .(2) As long as both players take the action C , continue the action C .(3) If someone takes the action “no donation,” take the action D after the event.

On the other hand, in the tit-for-tat strategy, players also start from C like thetrigger strategy, but take the action that the other player took in the last period (c.f.Remark 6.8). We indicate that it can be a subgame perfect equilibrium that twoplayers take the trigger strategy.

Suppose that two players take the trigger strategies (σ1, σ2). They start with Cand nobody chooses D under the pair of trigger strategies. The payoff under such anoutcome can be calculated as follows.

Ui (σ1, σ2) = ui (C,C) + δi ui (C,C) + δ2i ui (C,C) + . . . = 1

1 − δiui (C,C)

15See, for example, Maskin (1996) for game-theoretic consideration of the problem.

http://dx.doi.org/10.1007/978-4-431-55909-2_6


Here, δi (0 ≤ δi ≤ 1) represents how much individual-i discounts the payoff in thenext period and is called the discount factor of individual-i . As it gets closer toone, it means that individual-i values the future payoffs as highly as the presentone. The payoff calculated above is referred to as the discounted present value offuture payoffs. By using the actual payoffs, the discounted present value above canbe calculated as Ui (σ1, σ1) = 1

1−δi(16).

Now, to find the best alternative strategy when the other player is playing thetrigger strategy, consider pursuing strategy D from the beginning. As the other playeris taking the trigger strategy, the outcome of the game will be as follows:

I ndividual − 1 : C, D, D, D, . . .

I ndividual − 2 : D, D, D, D, . . .

Individual-2 can obtain a higher payoff by choosing D in the first period, but itmakes individual-1 play the action D and never play C . Hence, the optimal actionfor individual-2 is to choose D always.

The payoff under the outcome above, the discounted present value of the payoff ofindividual-i who deviated from the trigger strategy (i = 2 in the explanation above),can be calculated as follows:

ui (C, D) + δi ui (D, D) + δ2i ui (D, D) + . . . = ui (C, D) + δi [ui (D, D) + δi ui (D, D) + . . .]

= ui (C, D) + (δi

1 − δi)ui (D, D)

By putting actual values to the payoffs above, we get

Ui = 18 + (δi

1 − δi)(10).

Given that the other player is choosing the trigger strategy, it is better to choose thetrigger strategy if its payoff is higher or equal to the one under the best alternativestrategy above. In fact, there is no other strategy that provides a higher payoff com-pared to the trigger strategy above. Hence, if the inequalities U1(σ1, σ2) ≥ U1 andU2(σ1, σ2) ≥ U2 hold, each player choosing the trigger strategy is optimal.

Using the actual values of the payoffs, the conditions can be written as follows:

1

1 − δ1(16) ≥ 18 + (

δi

1 − δi) ⇒ δi ≥ 1

4

Hence, if the discount factors are large enough, that is, if players highly valuefuture payoffs, they will choose the trigger strategy to play the action “donation”and enjoy an efficient outcome. The results in the repeated game above suggest thatwhen the individuals behave selfishly, seeking short-run payoff, one can punish theplayer by taking the action (“no donation”) to lower the payoff of the player. If the


selfish player has a high discount factor and he/she values the cost of punishment,he/she will keep choosing the cooperative action “donation” to maximize the payoff.

It is easy to check that, under the conditions above, each player’s playing the triggerstrategy is a Nash equilibrium in any subgame that emerges after each simultaneousgame is played. Hence, the conditions for it to be the subgame perfect equilibriumare satisfied.

Remark 3.10 (Tit-for-Tat Strategy) The trigger strategy is a harsh punishment strat-egy in the sense that it punishes the player who deviates from the cooperation forever.On the other hand, in the tit-for-tat strategy, the punishment (D) will be given onlywhen the other player choose to deviate (D) and no punishment (C) will be givenif the other player corrects his/her action to choose the cooperative action (C). Thepunishment is quite weak as compared to the punishment in the trigger strategy, andthus players need to place a high value on future payoffs for the pair of tit-for-tatstrategies to be the subgame perfect equilibrium.

However, the tit-for-tat strategy has been shown to be a very powerful strategyin our daily lives, for example, by Axelrod (1984). Although it punishes selfishactions, it is a strategy with forgiveness, and thus, it can have longer periods ofcooperation with various types of strategies and can earn high payoffs on average.Its high performance depicted in Axelrod (1984) suggests that both punishment andforgiveness are important in getting along with selfish people in our daily lives. SeeRemark 6.8 for more discussion.

3.3.4 Cooperative Games (Bargaining Games)

In our lives, there are many occasions on which we bargain over how to share theoutcomes of our collaboration. The results of such bargaining have been studied forlong in the theories of cooperative games. One of the most well-known results is the“Nash bargaining solution” (Nash 1950), which can be illustrated as follows.

Consider a variation of the donation game in which the return from the donationbox is given to the two players and they can bargain over how much each will get.Under this assumption, each will donate 10 thousand yen and the total return willbe 32 thousand yen. Now, let (u1, u2) be the allocation of the total return betweenindividual-1 and individual-2. It must satisfy the condition u1 + u2 = 32, which issometimes referred to as the utility frontier.

Given this constraint, the Nash bargaining solution (uN1 , uN

2 ) is defined as thesolution to the following maximization problem:

max(u1,u2)

(u1 − d1)(u2 − d2) s.t. u1 + u2 = 32

where d1 and d2 are payoffs when the bargaining failed, and the pair of payoffs isknown as the disagreement point. In the maximization problem above, if we definev1 ≡ u1 − d1 and v2 ≡ u2 − d2, we can rewrite it as the one to maximize v1v2 with

http://dx.doi.org/10.1007/978-4-431-55909-2_6


respect to v1 and v2 under the constraint v1 + v2 = 32 + d1 + d2. The problem isvery similar to the utility maximization problem. As the maximization problem issymmetric in v1 and v2, the solution under the constraint v1 + v2 = 32 + d1 + d2

has a property v1 = v2. Hence, if d1 = d2, we have uN1 = uN

2 and if d1 > d2, wehave uN

1 > uN2 . This indicates that the Nash bargaining solution has a property that

those who can enjoy high payoffs in a bargain are the ones who have high payoffs atthe disagreement point, that is, when the bargaining collapses. The Nash bargainingsolution is simple and intuitive, and is frequently used in many applications.16

3.3.5 Tragedy of the Commons

3.3.5.1 The Problem

The donation game is one example of the “prisoner’s dilemma game” (c.f. Sect. 3.1).The essential problem is that society can fall into an inefficient state as a resultof selfish behavior of individuals. This problem is known as the “tragedy of thecommons” in the discussions of social problems.

It originally refers to the problem of commons, which are used by all for free,being overexploited to the point of exhaustion unless selfish individuals refrain frompursuing their self-interests (Hardin 1968). There are too many examples of suchtragedy of the commons. It can explain many serious problems in the world, fromthe arms races to the global environmental problems. In the following part, we choosethree examples from various social problems discussed in the book.

Case 1: Mutual Aid

In modern Japan, we are no longer able to expect mutual aid among members oflocal communities even in families because ties of families and traditional localcommunities have weakened. This problem can be interpreted as an example of thetragedy of the commons. Helping those who are in trouble is similar to contributingtowards solving social problems. If each provides aid and assistance, then our societycan enjoy an efficient state. However, such actions are costly and we have incentivesto free ride on the aid provided by others. The consequence of such selfish behaviorwill be the overexploitation of the goodwill of the people and lead to a state where noone provides aid in communities. Although systems of mutual aid in communitiesare wonderful, they cannot survive in general and we need some mechanisms to keepthem alive (c.f. Sect. 6.3.1).

Case 2: Fertility Behavior Under Social Security System

One of the most important arguments in this book is that the expansion of social secu-rity, especially the pay-as-you-go pension system, has been the powerful driving force

16We use the Nash bargaining solution in analyzing how wives and husbands share the returns frommarriage, as given in Sect. 4.3.2.

http://dx.doi.org/10.1007/978-4-431-55909-2_6

http://dx.doi.org/10.1007/978-4-431-55909-2_4


in transforming the Japanese society. This phenomenon can also be explained by thetragedy of the commons. Under the pay-as-you-go system, the pension contributionsby the young people are used for pension payment to the elderly people. Having achild is similar to contributing to the social security system. If everyone has a child,then social security system can be sustainable. However, raising children is costlyand we have incentives to free ride on the contributions made by the children of otherpeople. The consequence of such selfish behavior is declining fertility rates and thecollapse of the social security system. We need a policy to prevent such a tragedy ofthe social security system (c.f. Sect. 5.3.3).

Case 3: Depopulation under the Intergovernmental Transfer System

In the regional development and local government systems, similar problems occur.To reduce regional disparities, the central government pools the tax revenue andredistributes it to rural areas where the tax revenues are small and expenditures onthe elderly are large. Such a system of pooling tax revenues is basically the commonsthat can be used by needy local governments for free. Under such a system, the ruralareas have incentives to free ride on revenues that other local regions contribute. Ruralareas will make little effort to raise tax revenue and let young people go to urbanareas because local governments can get revenues from the commons. Depopulationmay be further deepened because of the commons. However, such selfish behaviorof local governments reduces their contribution to the common, and the revenues inthe redistributive system may run out (c.f. Sect. 10.3.1).

3.3.5.2 Solutions to the Problem

The cases above suggest that the tragedy of the commons, which makes our societyinefficient, can easily occur in our society. When there are such possibilities, effortsto prevent the tragedy are expected. Coase theorem (Sect. 2.3.1) suggests that ineffi-ciency will be perfectly solved by the bargaining among people when the transactioncosts are negligible.

In many societies, the tragedy of the commons has been avoided by setting up amechanism that changes the payoff structure to encourage cooperation among thepeople. In the following part, we explain five basic mechanisms to solve the tragedyof the commons. From the perspective of economics, the tragedy of the commonsgenerates inefficiency because individual behaviors have externalities. Hence, well-known mechanisms in economics to mitigate problems of externalities can be appliedto prevent the tragedy of the commons.

Solution 1: Legal Contracts.

The first way to solve the problem is to have legal contracts among members totake cooperative actions for avoiding the tragedy of the commons. It is expected towork because legal contracts can define and implement punishments against selfishbehavior. Such solutions are frequently used in many societies. For them to workwell, the legal system must be well-established and the costs to engage in effective

http://dx.doi.org/10.1007/978-4-431-55909-2_5

http://dx.doi.org/10.1007/978-4-431-55909-2_10

http://dx.doi.org/10.1007/978-4-431-55909-2_2


contracts must be low. If the costs for proving someone guilty are high, the contractsbecome incomplete and it is difficult for the society to prevent the tragedy of thecommons, as suggested by theories on incomplete contracts.

Solution 2: Long-run Relationship

The game-theoretic analysis in Sect. 3.3 suggested that if there are long-run relation-ships among players with high discount factors, people choose cooperative actionsbecause they expect to suffer from social punishments against selfish behavior. Theargument suggests that although it is difficult to engage in effective contracts, thetragedy of the commons may be avoided if there are social norms to encourage peo-ple to pursue cooperative behavior and enforce punishments on those who deviatefrom the norms. Such a mechanism is suggested to work well if there are long-runrelationships among people. However, in modern societies, long-run relationshipsare being lost because of high labor mobility in the developed free market system.The high mobility seems to be one of the most serious factors that weaken communi-ties and prevent them from mitigating the tragedy of the commons and the increaseddemand for the government’s invovement in the modern Japanese society.

Solution 3: Psychological Punishment or Morals

The social norms, which encourage people to enforce punishment on those whodeviate from the norms, may be gradually disappearing as the mobility of peopleincreases. However, many societies seem to have maintained mechanisms to psycho-logically punish those who deviate from social norms and exhibit selfish behavior viathe education of morals and legends.17 The system of morals can be seen as one ofthe mechanisms that our societies invented to discourage people from taking selfishactions. We can observe various social attempts, for example in the education of theglobal warming, to change the behavior of the people by appealing to their percep-tions, psychologies, and emotions. Although such efforts to foster morals have beenactively made, their effects may be quite limited because the punishments are justpsychological and not monetary.

Solution 4: Taxes and Subsidies

When there are institutions such as governments, it is possible to impose taxes onselfish behavior and give subsides for cooperative behavior. Such taxes and subsidiescan be seen as a mechanism to change the payoff structures to provide incentives forsocially desirable behavior. In public economics, taxes and subsidies to correct theproblems of externalities were proposed by Pigou (1877–1959) and are known asthe Pigouvian taxes and subsidies. The environment tax is an example of Pigouviantax.

17Johnson and Krüger (2004), for example, pointed out that supernatural punishment has played animportant role as a psychological punishment mechanism to discourage selfish people from deviatingfrom the social norm. There are numerous examples of selfish behaviors being discouraged by thefear of punishments by “god” in various religions or legends. They also pointed out that the reasonwhy there are small differences among the religions and legends in different regions are probablybecause problematic behaviors differ.


In the cases of families and local communities that face problems of the tragedyof the commons, certain penalties and rewards given by some members of the com-munities can work just like taxes and subsides by the government do. Whether thetaxes or subsidies are good for preventing inefficiency depends on various conditionsof the communities. In general, subsidies and rewards usually require some budgets,and thus the taxes and penalties that do not need such budgets are usually preferred.However, the latter can lower the expected returns from participating in the com-munities and may encourage people to leave the communities. Communities mayhave to collect some resources from their members to provide subsidies to encouragepeople to take cooperative actions.

Solution 5: Creation of Markets

The first welfare theorem states that “if all goods and services are traded in themarkets, under perfect competition, efficiency of allocations will be achieved” (c.f.Sect. 2.2.1). According to the theorem, the essential problem of the tragedy of thecommons lies in missing markets for services that cause the tragedy of the commons.Hence, by assigning adequate property rights to take actions that cause inefficiencyand allowing people to trade them in perfectly competitive markets, the tragedy of thecommons should be solved. Creating markets for emission trades can be consideredan example of the creation of markets to solve the problem.18

We often trade goods and services within families and communities. For example,we give and take gifts with members of the communities, which can be seen as akind of payment to other members who provide certain services to us. Although itis difficult to realize perfect competition within families and local communities andsuch transactions are far from complete, gift exchanges within communities can beviewed as our attempts at creating certain markets within communities. By receivingpayments in the form of gifts for the services in mutual aid, people are encouragedto retain mutual aid, which contributes towards preventing its collapse, that is, thetragedy of the commons.


In this chapter, we presented a brief overview of basic economic theories of decisionmaking to understand theoretical models and arguments in Part II. As suggestedby the models in this chapter, theoretical economic analyses attempt to understandand derive implications for our society and economy, by solving simple models thatdiscard details of our complex society.

In Part II, by incorporating various elements characterizing families and localcommunities, the models become a little more complex. Such models will give us

18A difficulty associated with this mechanism is the cost of creating and properly maintaining themarkets. For markets to work well, property rights must be well defined and constantly monitored.Whether creating markets is worthwhile depends on its costs and benefits.

http://dx.doi.org/10.1007/978-4-431-55909-2_2


new insights into the functions of families and various communities in our society.However, it is impossible to incorporate all complex elements of our society inthe models to explain and predict everything in our society. We need to realize thelimitations of the theoretical analysis to understand our society and economy.

The theoretical analyses allow us to clarify essential elements and structures thatcause various serious problems in our society. To have a good understanding of ourcomplex society and the problems therein, mathematical models are not enough. Weneed to have a good imagination and “words” that explain actual social phenomenabased on the insights that theoretical models allow us to have. Given the limitationsof theoretical analyses, which manifest from our abilities, we need to apply bothlogic and imagination. The joys of economic thinking may come from such dualapproaches to various social and economic problems.

References

Axelrod, R. (1984). The evolution of cooperation. New York: Basic Books.Buchanan, M. (1975). The Samaritan’s dilemma. In E. S. Phelps (Ed.), Altruism, morality, andeconomic theory. New York: Russel Sage Foundation.

Fudenberg, D., & Tirole, J. (1991). Game theory. Cambridge: MIT Press.Gordon, R., & Varian, H. (1988). Intergenerational risk sharing. Journal of Public Economics, 37,

185–202.Hardin, G. (1968). The tragedy of the commons. Science, 162, 1243–1248.Kornai, J. (1979). Economics of shortage. Amsterdam: North-Holland.Johnson, D. D. P., & Krüger, O. (2004). The good of wrath: Supernatural punishment and the

evolution of cooperation. Political Theology, 5(2), 159–176.Maskin, E. (1996). Theories of the soft budget constraint. Japan and the World Economy, 8, 125–

133.Osborn, M. (2003). An introduction to game theory. Oxford: Oxford University Press.Osborn, M., & Rubinshtein, A. (1994). A course in game theory. Cambridge: MIT Press.Shiller, R. J. (1999). Social security and intergenerational, intragenerational and international risk

sharing. Carnegie-Rochester Series in Public Policy, 50, 165–204.von Neumann, J., & Morgenstern, O. (1943). Theory of games and economic behavior. Princeton:

Princeton University Press.Nash, J. F. (1950). The bargaining problem. Econometrica, 18, 155–162.Yamashige, S. (1997). Fairness in markets and government policies: A weak equity criterion for

allocation mechanisms. Hitotsubashi Journal of Economics, 38, 61–78.

Part IIEconomics of Family and Community

Chapter 4Formation of Families

[A]t the beginning the Creator made them male and female.New Testament: Matthew 19:4

4.1 Introduction

We consider issues of family formations. From a biological viewpoint, the ultimategoal of human beings can be viewed as raising children who possess the same gene.Only the genes that successfully achieved this goal survived.1 Although we may notview it as the goal of our life, it is important to understand how such a biologicalgoal has been pursued by human beings.

In general, marriage between a man and a woman marks the beginning of havingchildren. In Japan, the fundamental reason behind the decline in fertility is consid-ered to be a decline in marriages and an increase in late marriages (c.f. Sect. 1.2.1).However, the data on marriages only reflect legal marriages. The above argument isinadequate because legal marriage is not necessary for human being to bear children.The necessary condition to have a baby is the biological marriage of a man and awoman.2

Legal marriage is simply a social system created by human beings. Once legalmarriage is defined, legal divorce needs to be introduced in the system. The legalsystem for marriage and divorce can be viewed as a system to socially control ourdesire to have children.

In the next section, we will consider factors that affect our decision of havingchildren. In Sect. 4.3, we try to understand the reasons for the existence of the legal

1See, for example, Dawkins (1976).2Under the present technologies, the marriage of a man and a woman is no longer necessary fora baby to be born. Can these technologies solve the problems of the low fertility in developedcountries?


71

http://dx.doi.org/10.1007/978-4-431-55909-2_1

72 4 Formation of Families

system of marriages and divorces and the way in which it controls our behavior ofhaving and raising children.

4.2 Children

Although we as human beings are biologically motivated to have children, we havesome control over the timing for having children as well as the number of children.3

One may claim that most of us behave instinctively and do not calculate the numberof children we want to have. Although such a claim is convincing to some extent,the resources we can use to raise children are limited, and thus we need to controlthe number of children we have. Needless to say, imperfect technologies, biologicallimitations, and various uncertainties and accidents make it difficult for us to achievethe most desirable number of children. Actual fertility decisions are complex, buthere, we set up a simple model to explain the basic elements of our fertility decisions.

4.2.1 Settings

Let us start from an explanation about the basic settings and assumptions of ourmodel. In economics, it has been pointed out that there are basically two types ofmotivations for child-bearing. One is for consumption purposes and the other is forinvestment purposes. The former refers to the idea that we want to have childrenbecause they are cute and they bring us joy.4 On the other hand, the investment-purpose relates to the idea that we want to have children because they will bring ussome benefits in the future. It corresponds, say, to our expectation that children willsupport our lives after retirement when we have no wage income.5

3For example, creation and prohibition of chances for matchmaking, contraception, and abortionwere traditionally available means to control births. Due to the technological progress in contracep-tion and fertility treatment, the controllability of the number of children has been greatly improvedin recent times. The impact of technological progress in birth control on our society is immense.4Such amotivationmay sound that we will have a child just to increase our own level of satisfaction.However, it also corresponds to the motivation that our genes incite in us to nurture their ancestorsby giving us joy in raising our children. Many people say that they want to have children becausethey think it is natural, which we regard as one of the motivations for the consumption-purpose (c.f.Fig. 8.2) because they feel better by having children.5Having children by such a motivation is sometimes referred to as “old-age security hypothesis.”There are very few species that expect their children to support their lives after they are no longer ableto get food. In many species, “adults” actually die soon after they give birth to children. Althoughhaving such expectation from our children to support them after we have no reproductive abilitymay sound very selfish, it may not be so because it may be one of the reasons why human beings, asa species on the earth, have survived for so long time. Unlike other species, our decisions, includingones on marriage and children that are critical for the survival of our genes, are critically dependenton intelligence and wisdom which accumulate as we live longer and can be transferred by “words”

http://dx.doi.org/10.1007/978-4-431-55909-2_8

4.2 Children 73

In our model, we assume that there is no uncertainty and only one period in ourlife, and that the number of children we have in the period need not be an integerbut can be a real number.6 We ignore the fact that men do not have the ability togive birth to a child and assume that each individual in the model makes a decisionabout the optimal number of children given the resource constraints. We assume thatfertility decision will be made before each individual gives birth to a child.

Fertility decisions are usually made as a consequence of bargaining and strategicgames with partners. We omit the analyses of such game-theoretic aspects of ourdecisions with partners and concentrate on each individual’s decision given the deci-sions of their partners.7 As we are interested in the decisions of women on childrenand labor participation in this book, we explain our results on individual decisionsmostly in terms of the decisions of women.

All these settings and assumptions above are unrealistic and too simple to describethe complexity of fertility behavior, but they allow us to have a simple model to showthe basic nature of our fertility decisions.

4.2.2 Basic Model

We set up a simple model for fertility decisions based on the model introduced inChap.3. In the basic model, we assume that the utility of each individual is dependentonly on consumption c and the number of children n and express it as a utility functionU (n, c). To raise a child, time m and a certain amount of goods h will be required.8

For simplicity, we ignore the time for leisure and assume that the total time H ofan individual with n children will be spent either on raising kids mn or on laborH − mn. Under the assumptions and notations above, the budget constraint can beexpressed as follows:

c + qhn = w(H − mn) + I + σ(n) (4.1)

where, q is the price of the goods for children,w is thewage, I is the unearned incomewhich we mainly regard as the income of the partner (husband), and the function

(Footnote 5 continued)to the next generations. Our living longer with the support of our children may contribute to thesurvival and their spread. See Chap.5 for more discussion on intergenerational transfers, that is,interactions between parents and children.6This real number of children here can be seen as the average number of children in one generationin a more sophisticated model with heterogeneous individuals or uncertainty.7In economics, there are many researches that explicitly analyze the bargaining and strategic gamesbetween a husband and a wife over resource allocation that affect the female’s decisions on childrenand labor participation. See, for example, Lundberg and Pollak (1996), Persson and Jonung (1997)and Bergstrom (1997). In Sect. 4.3, we also demonstrate amodel in which bargaining and interactivedecisions of couples are explicitly considered.8For the goods required to raise a child, we can think of composite goods and services that includefoods, clothes, nursery services, and education for each child.

http://dx.doi.org/10.1007/978-4-431-55909-2_3

http://dx.doi.org/10.1007/978-4-431-55909-2_5


σ(n) represents the income dependent on the number of children, in the form of,say, the child allowances and income transfer from their children. We assume thatthe price of the consumption c is normalized to be 1.

The utility maximization problem in the basic model can be expressed as follows:

maxc,n

U (n, c) s.t. c + qhn + wmn = wH + I + σ(n) (4.2)

The first order condition9 for themaximization problem can bewritten as follows.

∂L max

∂c= 0 : Uc = λ (4.3)

∂L max

∂n= 0 : Un = λ(qh + wm − σn) (4.4)

where, λ is the Langrange multiplier for this maximization problem.10 The condition(4.3) says that increasing consumption until the additional benefit, expressed in mon-etary value, from the consumption (Uc

λ) equals the cost (1) is optimal. On the other

hand, the condition (4.4) says that it is optimal to increase the number of childrenuntil the marginal benefit, expressed in monetary value, from increasing the numberof children (Un

λ+ σn) equals the cost (qh + wm) of raising a child.

The marginal benefit from an additional child, observed in the Eq. (4.4), is thesum of an increase in joy (Un) and an increase in the transfer (σn) from having onemore child. On the other hand, the marginal cost from an additional child is the sumof an increase in the cost of purchasing goods for children and an increase in theopportunity cost of spending more time to raise one more child.

We can rewrite these two conditions as the following condition via λ:

Un

Uc= qh + wm − σn (4.5)

This condition says that the marginal rate of substitution must be equal to the relativeprice of raising a child (qh + wm − σn) and consumption (1).

In the following two subsections, to concentrate on the analysis of fertility decisionin relation to changes in wage and labor supply, we ignore the cost (qhn) of raisingn children as well as the return (σ(n)) from them.

9See Sect. 3.2.2.10The Langrange multiplier represents the marginal utility of income, that is, an increase in utilitycaused by an increase in one unit of income. Hence, the value of marginal utility Uc dividedby the Lagrange multiplier (λ) can be interpreted as the monetary value of marginal utility fromconsumption.

http://dx.doi.org/10.1007/978-4-431-55909-2_3

4.2 Children 75

Fig. 4.1 Effects of a wageincrease and work-lifebalance

�

0

g

f

G

F

c

n

mwamwb

Hm

H−Zm

U(na,ca)

U(nb,cb)

Eb

Ea

n nanb

ca

cb

4.2.2.1 Effects of a Wage Increase

We first illustrate the effects of an increase in the wage on the optimal number ofchildren in Fig. 4.1. When the wage rate is wa , the budget constraint is c = (waH +I ) − mwan, which can be expressed as a segment of the line with intercept F ≡waH + I and slope −mwa .11 As the maximum time is given by H , the maximumnumber of children is given by na ≡ H

m . Hence, the budget constraint is defined up tona on the horizontal axis and depicted by the kinked line FEana .

In the figure,we drew a series of indifference curves for the utility functionU (n, c)and depicted the optimal point, where an indifference curve is tangent to the budgetconstraint, at Ea . At the optimal point, the individual does not work at all and useall her time to raise her children. Now, suppose that there is an increase in the wageto wb. In this case, the budget constraint is given by the kinked line GEana and theoptimal choice in Fig. 4.1 is given by the point Eb, where she works for (H − mnb)hours to consume cb. The number of children is now lowered at nb < na .

The result suggests that the rise in the education level of women and thus theirwages can reduce the number of children. The higher wage increases the opportunitycost wm of raising a child and thus reduces the optimal number of children.12

11The mathematical symbol ≡ here means “defined by”.12This is not the only case when the female wage increases. Although the increase in the femalewage has a negative substitution effect, it can have a positive income effect on the number of childrenallowing the couple to raise more children. Hence, higher education of women need not entail adecline in fertility. See Remark 3.2 for the explanation on substitution effect and income effect.

http://dx.doi.org/10.1007/978-4-431-55909-2_3


4.2.2.2 Work-Life Balance

In Japan, long working hours and difficulty in achieving work-life balance is claimedto be one of the reasons for declining fertility rates. In such an argument, it is perceivedthat workers are forced to work for certain amount of time to earn a high wageincome as a full-time regular worker, and the expected working hours are long inJapan. Therefore, the problem lies in the time constraint to raise children. Tomitigatethe problem of declining fertility rates in Japan, allowing workers to choose theirworking hours more flexibly to achieve a work-life balance is effective.

To support such an argument, in Fig. 4.1,we draw a vertical line at n correspondingto the case when workers are required to work more than Z hours, which restrictsthe maximum number of children the worker can have to n ≡ H−Z

w . Now, suppose aworker can choose the working hours freely at a lower wage rate wa as a part-timeworker. Under such a setting, the budget constraint of the worker is given by thekinked line Ea f gG in Fig. 4.1.

With this budget constraint, the individual in Fig. 4.1 will choose the point g atwhich she works for Z and choose to have n children. It is because if she choosespoint g, the indifference curve going through the point will be located on the north-east of the curve forU (na, ca), which goes through point Ea . It means that the utilitylevel is higher at g than at Ea . The optimal number of children n that the individualchooses is lower than nb at the optimal point Eb when there were no time constraintas a full-time regular worker (n < nb). That is, the constraint of long-working hourscan make the number of children inefficiently low. Hence, if the constraints wererelaxed we could expect higher fertility rate.

If we are only interested in the number of children, it is actually higher at Ea

where women does not participate in the labor market (na > nb). However, there area few problems in such a choice. First, the welfare is lower at Ea than at Eb. Second,the individual who has high ability to earn a high wage wb is not working because ofthe time constraint. It means that the investment on education to yield high abilitieshas low return (no return in the case of Fig. 4.1), thus discouraging young girls andtheir parents from investing in education. This will have negative long-run effectson the economy. Considering the fact that not only children but also education andlabor participation have externalities on the society, women choosing a point like Ea

is not likely to be efficient.By relaxing the constraint on working hours and allowing workers to make a

choice close to their optimal ones, society will be able to reduce the inefficiently lowfertility rates and low female labor participation, and increase the efficiency, and inturn welfare of the society. The perception that some of the most serious problems ofthe Japanese society have their roots in the custom of inflexible long working hoursis important. Making the labor market and workplaces more flexible and family–friendly so that workers can balance between work and life seems to be one of themost important policies in Japan.

4.2 Children 77

4.2.2.3 Quantity and Quality of Children

In the arguments above, we have assumed that utilities of parents are dependent onlyon the quantity of children. It has been pointed out that parents care for the quality ofchildren. One of the most interesting empirically observed facts is that people showhigher preferences for quality over quantity. Economists have tried to explain suchempirical facts.13

Such an argument has important implications for Japan because the increase inincome due to higher female labor participation discussed above or higher childallowances may only be used to increase the quality of the children instead of thequantity. The following argument based on Becker and Lewis (1973) confirms thisargument.

Let us nowconsidern the utility fromchildren,which canbe expressed asn ≡ NQdependent on the quantity N and quality Q of children.14 Then, the basic model canbe rewritten as follows:

maxc,N ,Q

U (N , Q, c) s.t. c + qhNQ + wmNQ = wH + I + σ(NQ)

The first order condition here is given by three conditions:Uc = λ,UN = λQ(qh +wm − σn) and UQ = λN (qh + wm − σn).

If we recognize that an increase in income usually has larger positive effects onquality Q than on quantity N , we can demonstrate that it increases Q, and thusreduces UQ . Hence, the marginal rate of substitution UN

UQ= Q

N becomes larger at thenew optimum. That is, an increase in the income will increase n ≡ QN but mayincrease Q and decrease N .15

4.2.3 Extension of the Basic Model

In the basic model above, to have a basic understanding on fertility decisions, weassumed that we need to have a certain amount of time and goods for children.However, we can often reduce the time for raising children by spending more ongoods for children; for example, nursery services.

Hence, to extend our model, we introduce an assumption that time and goods forchildren are substitutable. To be more specific, we introduce an assumption that wecan “produce” n children by a production function n = f (h,m).16

13See, for example, Willis (1973), Becker and Lewis (1973) and Lundholm and Ohlsson (2002).14For simplicity, we are implicitly assuming that the quality Q of each children is the same.15Considering the effects in a diagram with N and Q as the horizontal and vertical axes may beuseful for understanding the argument.16The model here is based on Apps and Rees (2004).


Given that we do not consider the time for leisure, as in the basic model, thebudget constraint can be written as

c + qh + πn = w(H − m) + I + σ(n)

where, we assume that we need to pay π > 0 as a basic cost for each child. Wealso assume that the function σ(n) can be written as σ(n) ≡ (σ + σ0)n where weinterpret σ as the child allowance and σ0 as the transfer from each children. Then,the total net benefit per child can be written as θ ≡ σ + σ0 − π .

The utility maximization problem can be written as follows:

maxc,m,h

U (n, c) s.t. c + qh + wm = wH + I + θn

n = f (h,m)

Note that in the extended model, individuals are not choosing n directly but assumedto choose time m and goods h for children considering their effects on n via thefunction n = f (h,m).

Under the assumptions above, we can solve the problems as follows. First, we fixa number of children at n and find a pair of goods h and time m which minimizesthe cost of producing n subject to the constraint n = f (h,m):

minh,m

qh + wm s.t. n = f (h,m). (4.6)

The optimal goods and time can be written as h(wq : n) and m(wq ; n), respectively.Given these functions, the utility maximization problem can be rewritten as follows:

maxc,n

U (n, c) s.t. c + qh

(w

q; n

)+ wm

(w

q; n

)= wH + I + θn (4.7)

We consider the first order conditions for each problem. Let μ be the Lagrangemultiplier for the minimization problem (4.6). Then its first order conditions aregiven by q = μ fh and w = μ fm . The first equation requires that we purchase goodsfor children until the cost q to purchase the additional goods equals the benefit from itμ fh . A similar interpretation can be given for the second equation in the decision onthe time for children. The two equations can be integrated to form equation w

q = fmfh,

which expresses that relative factor prices must be equal to the marginal rate oftechnical substitution, a well-known condition in economics.

The first order conditions for the maximization problem are given byUc = λ andUn = λ(qhn + wmn − θ), where λ is the Lagrangian multiplier. The first equationrequires that consumption must be increased until the benefit of increasing a smallone unit of consumption equals its cost (1). The second equation requires that thenumber of children must be chosen so that the benefit (Un

λ+ θ ) from an additional

child is equal to its cost (qhn + wmn). These two equations can be integrated into

4.2 Children 79

the following equation:

Un

Uc= qhn + wmn − θ

The condition indicates that, under the optimal choice, the marginal rate of substitu-tion between the child and consumption must be equal to the relative price of rasingthe child (qhn + wmn − θ ) and the consumption (1).

4.2.4 Cobb-Douglas Case

We now consider a simple case in which both the utility function and the productionfunction are expressed as the Cobb-Douglas functions, defined as follows:

f (m, h) = mαh1−α (4.8)

U (n, c) = nβc1−β (4.9)

where the parameters α and β are real numbers between 0 and 1.The results are critically dependent on the assumptions above, but they allow us

to have clear and interesting insights about our model. Needless to say, we will havedifferent results under different assumptions.We interpret and derive the implicationsof our results with some careful attention to our assumptions.

Under the assumption, we have the following solutions to the cost-minimizationproblem (4.6).

h

(w

q; n

)=

(1 − α

α

)α (w

q

)α

n

m

(w

q; n

)=

(α

1 − α

)1−α ( q

w

)1−α

n

Given these two functions, the cost qh + wm of raising a child can be given by

qh

(w

q; n

)+ wm

(w

q; n

)= Awαq1−αn

where A ≡ (1α

)α (1

1−α

)1−α. Hence, the budget constraint can be given by

c + (Awαq1−α − θ)n = wH + I

Now, the cost per child for raising n children is given by γ (w, q, θ) ≡ Awαq1−α − θ ,which we refer to as the “price of the child.” We assume that γ (w, q, θ) > 0 and


analyze the effects of wage rate w, price q of the goods for children, and the childallowance θ on the number of children n and the labor supply H − m.

Using the first order conditions tomaximize the utility functionU (n, c) = nβc1−β

subject to the budget constraint, we can obtain the optimal m and h, and the optimalnumber of children, consumption, goods, and time for children are given as follows:

n∗ = β(wH + I )

γ (w, q, θ)

c∗ = (1 − β)(wH + I )

h∗ =(1 − α

α

)α (w

q

)α (β(wH + I )

γ (w, q, θ)

)

m∗ =(

α

1 − α

)1−α ( q

w

)1−α(

β(wH + I )

γ (w, q, θ)

)

Now, to analyze the effects of changing some variables on the decisions for n∗ andm∗, we take the total differentiation of their optimal conditions and rearrange themto see the effects. For example, the effects of changing the child allowance θ for thenumber of children as well as the time for the children can be written as follows.

dn∗ = β(wH + I )

γ (w, q, θ)2dθ ⇒ ∂n∗

∂θ= n∗

γ (w, q, θ)> 0

dm∗ =(

α

1 − α

)1−α ( q

w

)1−α(

β(wH + I )

γ (w, q, θ)2

)dθ ⇒ ∂m∗

∂θ= m∗

γ (w, q, θ)> 0

They indicate that an increase in child allowance raises the optimal number n∗ ofchildren and the time m∗ for children, and thus reduce labor supply H − m∗. Theeffects of other factors can be analyzed as follows:

∂n∗

∂ I= β

γ> 0 and

∂m∗

∂ I=

(α

1 − α

)1−α ( q

w

)1−α β

γ> 0

∂n∗

∂w= 1

γ(βH − m∗) and

∂m∗

∂w= θβ

w((1 − α)I − αwH) − β Iwα−1q1−αA

∂n∗

∂q= 1

γ(−h∗) < 0 and

∂m∗

∂q= (1 − α)wα−1q−α( α

1−α)1−αβ(wH + I )

γ 2(−θ)

Let us summarize the results above as a proposition. As the budget constraint ofthe government is not considered here, the effects of the changes in policies onlyreflect those on the individual decisions.

Proposition 4.1 In the Cobb-Douglas case, we have the following results:

(1) An increase in non-labor income increases the number of children and reducesthe labor supply.

4.2 Children 81

(2) An increase in the wage rate will reduce the number of children when the laborsupply is small, but will increase it when the labor supply is large. It will raisethe labor supply when θ is small, but lower it when θ is large.

(3) An increase in the child allowance increases the number of children and reducesthe labor supply.

(4) An increase in the price of goods for children decreases the number of children.It will decrease the labor supply when θ > 0, but increase it when θ < 0.

The first result suggests that if we interpret the non-labor income as the incomeof the spouse, wives who have husbands with a high income work less and havemore children than those who do not. The second result relates to the fact that higherwages due to higher education encourage women to work more and discourage themfrom having children, that is, the standard statement to explain the declining fertilityin modern society. However, it is useful to recognize that the statement holds only ifthe female labor supply is small. When women work sufficiently long, an increasein the wage will increase the number of children because the income effect exceedsthe substitution effect (c.f. Remark3.2).

The last two results indicate the effects of public support for families. The increasein the child allowance and the reduction in the price of goods for children are shownto increase the number of children. The effects on female labor supply are different.The child allowance has negative effects as only the income effect work on laborsupply. The subsidy for the goods for children can increase female labor supply if θ

is negative, because the substitution effect from time for children to goods for raisingchildren increases the labor supply. However, when θ > 0 the subsidy will have anegative effect on the labor supply because it increases the number of children, andthus increases the time for children, which cannot be offset by substituting with thegoods for raising children.

If we recognize the differences in the effects of these two public policies onfemale labor supply, we come up with a hypothesis that by increasing the subsidiesfor nursery services by reducing the child allowance, we can encourage female laborsupply and use the tax revenue collected from it to increase the number of children.We show by providing a simple sufficient condition that the hypothesis can be true.17

In the following part, we assume that the government uses the proportional laborincome tax to finance the public support for families. With the tax rate τ , the govern-ment’s budget constraint for such a policy can be written as τw(H − m) = σn + gh,where g is the subsidy for each unit of goods for children and σ is the allowanceper child. The individual budget constraint under the policy can be given byc + (q − g)h = (1 − τ)w(H − m) + I + (σ + σ0 − π)n. In the following part, we

17The argument and results here are based on Apps and Rees (2004, Proposition 3). In general,an increase in the number of children is expected to increase women’s time for raising children.However, if the number of children is increased as a result of an increase in the goods for raisingchildren, and not the increase in the time for children, there will be a rise in the tax revenue that canbe used to increase the subsidy for families. Needless to say, the increase in the number of childrenwill, in general, increase the future tax revenue, and thus the effects of such policy can be evenlarger if we consider the long-run effects.

http://dx.doi.org/10.1007/978-4-431-55909-2_3


assume θ < 0 to consider the case in which subsidy increases female labor supply.Furthermore, to simplify the proof, we assume that there is no subsidy (g = 0) beforethe policy is implemented.

Proposition 4.2 Under assumptions θ < 0 and g = 0, an increase in the subsidyfor the goods for children financed by a reduction in child allowance increases boththe number of children and female labor supply.

Proof By totally differentiating both sides of the government’s budget constraint andassuming that the equality holds, we can consider the effects of the revenue-neutralpolicies. Under the assumption g = 0, the result can be expressed as follows:

(σ + γ

γ

)(n∗dσ + h∗dg) = τw

[−m∗

γ

(dσ + θ

qdg

)]> 0 (4.10)

The right-hand side indicates the effects of the changes in child allowance σ and thesubsidy for the goods g on the tax revenue, which is expressed by the income taxrevenue (τw) per unit of labor supply multiplied by the change in the labor supply.Under our assumptions θ < 0, dσ < 0, and dg > 0, the labor supply increases; thus,the right-hand side becomes positive. This implies that the effects of the revenue-neutral policies on the government’s budget constraint, which can be expressed bythe left-hand side, become positive.

Now, the effects of the policy change on the number of children can be expressedas follows:

dn∗ = ∂n∗

∂σdσ + ∂n∗

∂gdg = ∂n∗

∂θdσ − ∂n∗

∂qdg = 1

γ(n∗dσ + h∗dg)

which is equal to the left-hand side of equation (4.10) divided by σ + γ > 0. Hence,the number of children increases.

4.3 Marriage

As we have already mentioned in the introduction to this chapter, legal marriage isnot a necessary condition for women to have children. However, in many societies,there is a legal or formal marriage system in which couples are recognized by thesociety as married couples. The system seems to have existed because couples’ long-run relationships expected from formal marriage are useful for human beings to raisechildren who require lots of time and resources to grow up. From such a viewpoint,whether the marriages will last for a long time is important for children to grow upsafely.

A high possibility of divorce will have negative effects on marriage and fertilitydecisions. In this section, based on the theories of incomplete contracts, we discussissues concerning marriage and divorce and their effects on our fertility decisions.

4.3 Marriage 83

4.3.1 Settings

We use a variation of the two-period model by Rainer (2007) to explicitly considerthe possibility of divorce.18

In period-1, two individuals, individual-1 and individual-2, get married and eachindividual-i (i = 1, 2) makes some investment Ii in the marital assets such as chil-dren. The value of the marital assets generated by the investment can be writtenas V (I1, I2). We assume that the function V is increasing in each Ii and marginalproductivity is decreasing, that is, ∂V

∂ Ii> 0 and ∂2V

(∂ Ii )2< 0 for each i (i = 1, 2). In

period-2, the two individuals decide whether they want to continue their marriage.When the marriage continues, the utility from the asset V (I1, I2) is expressed by

γi V (I1, I2), while it is lowered to γ Di V (I1, I2) when they get divorced, for each i

(i = 1, 2), where γ Di < γi . Furthermore, their marital assets will be divided, based

on law, by the ratios π1 and π2 (π1 + π2 = 1). Hence, the total utility in the event ofa divorce is given by di ≡ γ D

i πi V (I1, I2) for each i (i = 1, 2).If the information is complete, the couple will not get divorced because the utility

from the divorce is lowered. However, the information is not complete, and thusthe possibility of divorce cannot be denied. We assume that two individuals makedecisions believing that there is a probability p that they get divorced until they reachperiod-2.19

4.3.2 Basic Model

To find a reasonable outcome, we solve the problem backwards, that is, from thelast period.20 In period-2, individuals already invested in the marital assets and theuncertainty of the partner, and thus the marriage is dissolved.What they do is bargainover how much each will get from the marital assets V (I1, I2). We assume that thesolution is given by the Nash bargaining solution (uN

1 , uN2 ), which can be calculated

as the solution to the following maximization problem (c.f. Sect. 3.3.4).

max(u1,u2)

(u1 − d1)(u2 − d2) s.t. u1 + u2 = (γ1 + γ2)V (I1, I2)

18There are many researches on marriage contracts and legal issues of divorces from the viewpointsof contract theories. For example, see Rowthorn (1999), Chiappori et al. (2002) and Weiss andWillis (1985).19The decision of divorce must be determined endogenously in the model; thus, rational individualsshould make decisions by calculating the probability of divorce. However, in our model, we assumefor simplicity that individuals cannot calculate the probability p of getting divorced and take it asgiven.20See Sect. 3.3.2 for backward induction to solve a sequential game.

http://dx.doi.org/10.1007/978-4-431-55909-2_3

http://dx.doi.org/10.1007/978-4-431-55909-2_3


By solving the problem, we get uNi = θi V (I1, I2) for each i (i = 1, 2), where θi is

given by

θ1 =[γ1 + γ2 + γ D

1 π1 − γ D2 π2

2

], θ2 =

[γ1 + γ2 − γ D

1 π1 + γ D2 π2

2

](4.11)

It can be seen as the parameter for the bargaining power over the marital assets. Thebargaining power of i is smaller, if the legally determined division ratio πi and theproportion of the marital assets at divorce γ D

i gets smaller. It is because individual-iwill be forced to have lower utility at the disagreement point where the couples getdivorced, and thus will need to bear with the lower utility in the bargain to avoid sucha bad condition.21

Individuals make decisions about the investment in Period-1, expecting the resultsabove in Period-2. First, the efficient investment (I E1 , I E2 ) for each individual can beobtained by solving the following maximization problem.

maxI1,I2

(γ1 + γ2)V (I1, I2) − I1 − I2

The first order conditions, which set the partial differentiation of the objective func-tion in the maximization problem with respect to I1 and I2 at zero, are given asfollows:

(γ1 + γ2)Vi (IE1 , I E2 ) = 1 (i = 1, 2)

where Vi ≡ ∂V∂ Ii

. These conditions indicate that it is optimal for each to invest untilthe sum of additional benefits (γ1Vi + γ2Vi ) from the investment becomes equal tothe cost of the investment (1).

On the other hand,when individuals seek tomaximize their ownutility, it is reason-able to assume that they will maximize pγ D

i πi V (I1, I2) + (1 − p)θi V (I1, I2) − Iigiven the decisions of their spouses. Under such an assumption, the pair of optimaldecisions (I ∗

1 , I ∗2 ) for which each individual finds no incentive to choose a different

level, must satisfy the following pair of first-order conditions:

(pγ Di πi + (1 − p)θi )Vi (I

∗1 , I ∗

2 ) = 1 (i = 1, 2)

Such a pair (I ∗1 , I ∗

2 ) is known as theNash equilibrium (c.f. Sect. 3.3.1). The conditionsindicate that it is optimal for each individual to invest until the additional expected

21For example, housewives and househusbands who accumulate knowledge and ability that arevaluable mainly for their spouses and children, that is, make relation-specific investment in humancapital, will face low possibility of finding a good job in the labor market after divorce. Hence,the bargaining powers of the housewives and househusbands in their marriages are generally low,which may result in domestic violence and moral harassment in marriage. One of the reasons thatmany Japanese women obtain higher education even if they prefer to be housewives may be to gainbargaining power in the marriage by raising the conditions at divorce.

http://dx.doi.org/10.1007/978-4-431-55909-2_3

4.3 Marriage 85

benefit ((pγ Di πi + (1 − p)θi )Vi ) from the investment equals the additional cost (1)

of the investment.It is easy to check that the following condition holds:

(γ1 + γ2)Vi (IE1 , I E2 ) = (pγ D

i πi + (1 − p)θi )Vi (I∗1 , I ∗

2 ) (i = 1, 2)

As we have γ Di πi < γ1 + γ2 and θi < γ1 + γ2, for any p (0 ≤ p ≤ 1), we have

γ1 + γ2 > pγ Di πi + (1 − p)θi , and thus Vi (I E1 , I E2 ) < Vi (I ∗

1 , I ∗2 ) (i = 1, 2). By our

assumption that marginal productivity is decreasing, we have I Ei > I ∗i (i = 1, 2).

That is, the following proposition holds.

Proposition 4.3 The optimal investment in marital assets at the Nash equilibriumis inefficient.

The problem of an inefficiently small investment in marital assets occurredbecause the two individuals cannot engage in a contract ex ante to make the effi-cient investment. This problem is well known as the “hold-up problem” in the theoryof incomplete contracts.

Notice that the same problem occurs even if the probability of the divorce p = 0.That is, the possibility of divorce is not the essential element of the problem, but itis the inability to engage in a contract that binds the decisions in the second period.However, an increase in the probability of divorce will make the problem worse, assuggested by the following proposition.

Proposition 4.4 An increase in the probability of divorce lowers each individual’sinvestment in marital assets at the Nash equilibrium.

Proof Suppose that the probability of divorce is increased to p + ε (ε > 0). LetI ∗∗i be the optimal investment for each individual-i in this case (i = 1, 2). As uN

i >

di holds at the Nash bargaining solution, we have θi > γ Di πi and thus γ1 + γ2 >

pγ Di πi + (1 − p)θi > (p + ε)γ D

i πi + (1 − p − ε)θi . Hence, Vi (I E1 , I E2 ) < Vi (I ∗1 ,

I ∗2 ) < Vi (I ∗∗

1 , I ∗∗2 ) must hold (i = 1, 2). By the assumption of decreasing marginal

productivity in V , we have I Ei > I ∗i > I ∗∗

i .

Finally, let us consider the marriage decision in period-0. Let u0i be the expectedutility when the individual-i did not get married. The individual-i will get marriedonly if the following condition is satisfied22:

(pγ Di πi + (1 − p)θi )V (I ∗

1 , I ∗2 ) ≥ u0i (4.12)

Now, when the divorce probability p gets higher, (pγ Di πi + (1 − p)θi ) will go

down and the optimal investment will be lowered, and thus the expected utility fromthe marriage, expressed by the left-hand side of the inequality (4.12) above, must

22Notice that an increase in the expected utility from not getting married u0i will decrease themarriage rate. For example, an increase in social security payment or an development of the marketeconomy will raise u0i and thus will lower the marriage rate and the fertility rate.


be lowered. By denoting p∗i as the probability p under which equality holds in the

inequality above,23 we have the following proposition.

Proposition 4.5 When the divorce probability p exceeds p∗i , individual-i chooses

to avoid the marriage.

The theoretical result above suggests the following relations concerning the invest-ment in children.24

(1) If it is not possible for a couple to engage in a contract about the division ofthe marital assets that their children generate, investment in children will beinefficiently low.

(2) If the perceived probability of divorce gets higher, the investment in childrenwill be lowered.

(3) If the perceived probability of divorce gets higher, the number of marriage willfall.

4.3.3 Factors Affecting Divorce Probability

The result above suggests that the increase in the divorce probability can be one ofthe reasons for declining fertility rates, because it discourages people from gettingmarried and having children. The divorce probability is indeed getting higher inJapan (c.f. Fig. 1.5) and is estimated to reach one third in the long run (Raymo et al.2004). Let us consider, based on a simple model, some factors which can increasethe probability.

As in our basicmodel, we assume that the continuation of themarriage is desirablefor each couple. However, couples are not really certain if it is so, and they sometimesget divorced even if they can be better off by keeping the marriage. They are likelyto divorce if it is painful for them to bargain with their partners over the returnsfrom the marriage (marital asset, in the model above), that is, if the bargainingcosts are perceived to be high. There are many examples other than that of marriageand divorce, in which we choose to end the relationship with others because thebargaining costs to keep the relationship with them are quite high, even if we feelthat the relationship with them can generate positive payoffs.

To formalize this idea, let Ci be the bargaining cost that individual-i need to payfor bargaining with the spouse in period-2. We assume that the cost is dependent onthe type of the spouse and that it is drawn from the interval [CL

i ,CHi ] with the same

probability. Individual-i will choose divorce when he/she finds that the bargainingcost Ci satisfies the following condition:

23It is calculated to be p∗i = θi V (I ∗

1 ,I ∗2 )−u0i

(θi−γ Di πi )V (I ∗

1 ,I ∗2 ).

24Here, we basically consider the investment in the quality and quantity of children. As examplesof investment in the quality of children, we can think of the money spent on the children’s educationas well as the time spent with the children.

http://dx.doi.org/10.1007/978-4-431-55909-2_1

4.3 Marriage 87

θi V (I ∗1 , I ∗

2 ) − Ci ≤ γ Di πi V (I ∗

1 , I ∗2 ) (4.13)

If Ci is the bargaining cost under which the equality holds in the above inequality,we have Ci = (θi − γ D

i πi )V (I ∗1 , I ∗

2 ). Then, the probability of divorce is the proba-bility that Ci lies in the interval [Ci ,CH

i ]. Under our assumption of the probability,that is, the uniform distribution over the interval, the probability of divorce is given byp = (CH

i − Ci )/(CHi − CL

i ).25 Now, the divorce probability can be higher if Ci getslower, which can be induced by the increase in the payoff at divorce (γ D

i πi ) and thedecrease in the benefit of keeping the marriage θi V (I ∗

1 , I ∗2 ). Hence, a decrease in the

cost of divorce owing to an increase in femalewages and an increase in public supportfor divorced low-income women26 can be seen as a reason for an increase in divorceprobability. Furthermore, fewer children, which lowers the benefit of marriage, canalso be a reason for an increase in the probability of divorce.

The divorce probability can be also affected by the change in the distribution ofbargaining costs. For example, if the minimum cost CL

i is increased, it will raise thedivorce probability according to the conditions above. For example, the spread ofindividualism or generosity towards the diversity of the people inmodern society willmake bargainingwith othersmore difficult thanwhen people have similar preferencesand wills to cooperate with others. Under such changes in our society, bargainingwith the spouses becomes more difficult, which may be expressed as the increase inCLi , and thus the divorce probability may become higher.Thediscussion abovemayallowus to have the followingpolicy implications. First,

to raise the number of children to an efficient level, itmaybeuseful for the governmentto encourage couples to engage in contracts for the division of marital assets beforetheygetmarried. Itmayalsobeuseful to lower the probability of divorce, byprovidingbenefits, such as in the pension system to couples who keep their marriage, or byfostering themarket formarriage counseling services. It is also important to recognizethat if an increase in the support for raising children succeeds in raising the numberof children, the benefit of maintaining the marriage will increase, thus decreasingthe divorce probability, which in turn can further increase the number of children.27

4.3.4 Meaning of Legal Marriage and Divorce

The existence of the legal marriage system has allowed men and women to engagein marriage contracts and invest in children more easily than when there was nosuch system. If there is no formal marriage system, couples, especially women, will

25We assume CLi < Ci < CH

i .26For example, there are special child allowances and tax measures for single mothers in Japan.27Needless to say, lowering female wages and public support for single mothers, which raises thecost of divorce, can contribute to lowering the divorce probability and raising the number of children.Such policies may not be desirable in Japan if we consider other policy goals. For example, seeBougheas and Georgellis (1999) for analyses of the effects of increase in divorce cost on marriageand divorce rates.


face more uncertainty about the future relationship with their partners and feel moreanxious about investing in children, leading to a fall in the number of children.28

A society without a formal marriage system will not survive the process of naturalselection.

The logic above is the same as the one for explaining why two companies hesitateto make a big investment in a joint venture if there were no formal contracts for thedivision of the outcome, an action that can result in inefficient transactions. Raisingchildren can be viewed as a joint venture business between men and women, andby creating circumstances and institutions in which men and women can engagein a formal marriage contract, the society can make them feel comfortable aboutinvesting in children and encourage them to have children. Considering that havingchildren produce externalities and are essential for the efficiency and sustainabilityof the society, it is no wonder that most of societies have formal marriage systems,which are maintained through their efforts.29

In many societies, some efforts are made towards raising the cost of divorce toencourage couples to maintain their marriages. For example, there are, or used tobe, laws to discourage divorces and certain social stigma associated with divorces.However, if the cost of divorce is too high it can discourage people from gettingmarried (c.f. inequality (4.12) above). Hence, as we have already mentioned above,in modern societies there seem to be some efforts to lower the cost of divorce.

However, if the cost was too low getting divorces would become too easy, whichwill have a negative effect on the marriage and child-bearing. Hence, some countrieshave started increasing the public support for marriage and child-bearing to mitigateside effects of lowering the costs of divorce.

For example, the “sambo system” in Sweden provides the same benefits to unmar-ried couples as married couples, as long as they register their cohabitations. Due tosuch a system, the ratio of out-of-wedlock childbearing is high in Sweden. Such asystem can be understood as one that mitigates the negative effects of the traditionalmarriage system on the marriage ratio by encouraging unmarried couples to havechildren. The Japanese society is also facing a problem of low fertility rates due to adecrease in marriages or an increase in late marriages. Reforms of the legal systemfor marriage and divorce may be required along with an increase in public supportfor families to raise children.

28Despite such conjecture, the number of marriages after couples found to have their babies isincreasing in Japan. Our model suggests that having babies before the couples engage in a marriagecontract is a risky behavior, especially forwomen. Traditionally, women have been expected to avoidsuch behavior because there are high risks of getting an abortion or becoming a single mother. Onecan explain such behavior as a strategy to get a good partner in the competition amongwomen underthe social norm that children must be raised by married couples. Another explanation may be thatthe cost of such behavior has been lowered in modern societies where the cost of “out-of-wedlockchildbearing” is lowered by the increase in social and public support for single mothers and thedecrease in the stigma associated with it. For example, see Willis (1999) and Akerlof et al. (1996)for economic analysis of “out-of-wedlock childbearing”.29For example, see Clark (1999) and Smith (2003) for the analyses of marriage contracts from theperspective of law and economics.

4.3 Marriage 89

4.3.5 Related Literature

In economics, there has been an accumulation of researches on marriage anddivorce.30 In the following part, we will briefly introduce some important discus-sions in relation to our basic models that we could not cover.

First, in the models above, we only considered children as the marital assetsV (I1, I2). However, there are many examples of marital assets as suggested by thediscussion on the benefits of the marriage in Sects. 2.2.3 and 8.2.2. In the analysis onthe distribution of the benefits of marriage (marital assets) within each couple, mostresearches use Nash bargaining solutions, that is, approaches based on cooperativegames (c.f. Lundberg and Pollak 1996, 2003). There are some researchers who claimthat analysis should be based on theories of non-cooperative games (c.f. Lommerud1997), while there are others who attempt to conduct empirical studies on whichtheoretical predictions explain our behavior well (e.g., Weiss 1997).

In our model, the partners are already found and required to make the decisionto get married. However, to get married we generally need to find partners. Thereare many researches based on the sophisticated search models on how the searchfor partners is done and the factors that affect the search (c.f. Bergstrom and Scheni(1996), Boulier and Rosenzweig (1984), Fernandez et al. (2005) and Peters andSiow (2002)). It is true that whether we get married critically depends on whetherwe can meet someone whom we want to get married to. The difficulty in such asearch has been claimed to be one of the most important reasons for the decrease inmarriage and increase in late-marriages these days. The abovementioned researcheshave some important and interesting policy implications to encourage marriage andchild-bearing. Furthermore, there is an accumulation of researches on mechanismsand algorithms to realize socially desirable pairings (e.g., Gale-Shapley 1962).

Further, there are a large number of researches on divorce. The largest volume liesin the analysis of the effects of a change in the legal system of divorce on marriagesand the behavior of the couples. For example, there are many discussions on thedifference between “fault-based divorce law” and “no-fault divorce law” in theireffects on the behavior of people (e.g., Dnes and Rowthorn (2002) and referencestherein). Behind such discussions, lies the basic viewpoint of Coase (1960), thatis, as long as bargaining costs are negligible, the difference in the legal system ofdivorce should not change the efficiency of divorces (c.f. Sect. 2.3.1). Our argumentin this section suggests that, when the investments in the marital assets are not madeefficiently due to the incompleteness of contracts, Coase’s argument does not workwell and the difference in the legal system can affect efficiency of our behavior (c.f.Rowthorn (1999) and Weiss and Willis (1985)).

30The review of literature includes well-known studies such as Keely (1977), Pollak (1985), Cigno(1991), Grossbard-Shecktman (1993), Bergstrom (1997),Weiss (1997), Ermisch (2003), andBryantand Zick (2006, Chap.8).

http://dx.doi.org/10.1007/978-4-431-55909-2_2

http://dx.doi.org/10.1007/978-4-431-55909-2_8

http://dx.doi.org/10.1007/978-4-431-55909-2_2

http://dx.doi.org/10.1007/978-4-431-55909-2_8



In this chapter, we focused on the fact that families are formed for raising childrenand by raising children, and analyzed factors that affect family formation. It takestime and money to raise children, and thus the long-term relationship of the couplesare essential in family formations. The government can have an impact on our deci-sions on family formation by implementing policies to change these factors. Basedon a simple theoretical model, we stated some reasons for the decline in fertilityrates in Japan as well as public policies and institutions associated with the socialtransformation.

One of the important results in this chapter is that decisions on divorce after mar-riage impact fertility decisions. However, important decisions after family formationare not restricted to divorces. We make various decisions about how the resources(marital asset)within the families should be redistributed among the familymembers.Such resource allocations within families can have effects, similar to the decisionson divorces, on fertility behavior. In the next chapter, we focus on resource allocationwithin families, especially between parents and children, and analyze their structuresand consequences.

References

Akerlof, G. A., Yellen, J. L., & Katz, M. L. (1996). An analysis of out-of-wedlock childbearing inthe United States. The Quarterly Journal of Economics, 111(2), 277–317.

Apps, P., & Rees, R. (2004). Fertility, taxation and family policy. Scandinavian Journal of Eco-nomics, 106(4), 745–763.

Becker, G., & Lewis, H. G. (1973). On the interaction between the quantity and quality of children.The Journal of Political Economy, 81(2), S279–S288.

Bergstrom, T. (1997). A survey of theories of the family. In M. R. Rosenzweig & O. Stark (Eds.),Handbook of population and family economics (Vol. 1A, pp. 21–74). Amsterdam: ElsevierScience.

Bergstrom, T., & Schoeni, R. (1996). Income prospects and age of marriage. Journal of PopulationEconomics, 9, 115–130.

Bougheas, S., & Georgellis, Y. (1999). The effect of divorce costs on marriage formation anddissolution. Journal of Population Economics, 12(3), 489–498.

Boulier, B. L., & Rosenzweig, M. R. (1984). Schooling, search, and spouse selection: Testingeconomic theories of marriage and household behavior. Journal of Political Economy, 92, 712–732.

Bryant, W. K., & Zick, C. D. (2006). The economic organization of the household (2nd ed.).Cambridge: Cambridge University Press.

Chiappori, P.-A., Fortin, B., & Lacroix, G. (2002). Marriage market, divorce legislation and house-hold labor supply. Journal of Political Economy, 110(1), 37–72.

Cigno, A. (1991). Economics of the family. Oxford: Oxford University Press.Clark, S. (1999). Law, property, and marital dissolution. Economic Journal, 109(454), C41–54.Coase, R. H. (1960). The problem of social cost. Journal of Law and Economics, 3, 1–44.Dawkins, R. (1976). The selfish gene. Oxford: Oxford University Press.Dnes, A. W., & Rowthorn, R. (2002). The law and economics of marriage and divorce. Cambridge:Cambridge University Press.

References 91

Ermisch, J. F. (2003). An economic analysis of the family. Princeton: Princeton University Press.Fernandez, R., Guner, N., & Knowles, J. (2005). Love and money: A theoretical and empiricalanalysis of household sorting and inequality. The Quarterly Journal of Economics, 120(1), 273–344.

Gale, D., & Shapley, L. (1962). College admission and the stability of marriage. American Mathe-matical Monthly, 69, 9–15.

Grossbard-Shechtman, S. (1993). On the economics of marriage: A theory of marriage, labor, anddivorce. Boulder: Westview Press.

Keeley, M. C. (1977). The economics of family formation. Economic Enquiry, 15(2), 238–50.Lommerud, K. E. (1997). Battles of the sexes: Non-cooperative games in the theory of the family.In C. Jonung & I. Persson (Eds.), The economics of the family and family policies. London:Routledge.

Lundberg, S., & Pollak, R. (1996). Bargaining and distribution in marriage. Journal of EconomicPerspectives, 10, 139–158.

Lundberg, S., & Pollak, R. (2003). Efficiency in marriage. Review of Economics of the Household,1(3), 153–167.

Lundholm, M., & Ohlsson, H. (2002). Who takes care of the children? The quantity-quality modelrevisited. Journal of Population Economics, 15(3), 455–461.

Persson, I., & Jonung, C. (1997). Economics of the family and family policies. London: Routledge.Peters, M., & Siow, A. (2002). Competing pre-marital investments. Journal of Political Economy,110(3), 592–608.

Pollak, R. A. (1985). A transaction cost approach to families and households. Journal of EconomicLiterature, 23, 581–608.

Rainer, H. (2007). Should we write prenuptial contracts? European Economic Review, 51(2), 337–363.

Raymo, J., Iwasawa, M., & Bumpass, L. (2004). Marital dissolution in Japan: Recent trends andpatterns. Demographic Research, 11(14), 395–419.

Rowthorn, R. (1999). Marriage and trust: Some lessons from economics. Cambridge Journal ofEconomics, 23(5), 661–91.

Smith, I. (2003). The law and economics ofmarriage contracts. Journal of Economic Surveys, 17(2),201–226.

Weiss, Y. (1997). The formation and dissolution of families: Why marry? Who marries Whom?And what happens upon divorce? In M. Rosenzweig & O. Stark (Eds.), Handbook of populationand family economics (Vol. 1A, pp. 81–123). Amsterdam: Elsevier Science.

Weiss, Y., & Willis, R. J. (1985). Children as collective goods and divorce settlements. Journal ofLabor Economics, 3(3), 268–92.

Willis, R. J. (1973) A new approach to the economic theory of fertility behavior. Journal of PoliticalEconomy, Supplement, S14–S64.

Willis, R. J. (1999). A theory of out-of-wedlock childbearing. Journal of Political Economy, 107(6–2), S33–S64.

Chapter 5Resource Allocations within Families

[T]ell me which one of you loves me most, so that I can give mylargest gift to the one who deserves it most.

Shakespeare, W. King Lear. [I-1]

5.1 Introduction

In this chapter, we analyze resource allocation within families, especially the trans-fer of resources between parents and children.1 What is the difference between aparent-children relationship and an ordinary relationship between humans? It can beexplained by love or the “altruism”of parents towards their children. Parents’ altruismtowards their children can be well understood from the biological and evolutionaryviewpoints.

In the next section, we first analyze the roles and consequences of gifts and inheri-tances from parents to children. In general, gifts and inheritances can be explained byparents’ altruism for their children. However, we have to consider a fact that parentsalso have self-interest and care for their own life and welfare.2 Gifts and inheritancefrom parents to children, which seem to be altruistically motivated, may be explainedby parents’ self-interest to get loves and resources from their children.

1We can observe resource transfer between spouses and among siblings. Some analyses examineresource transfer between husbands andwives, as suggested by our discussion in Sect. 4.3. However,very few analyses investigate resource transfer among siblings, suggesting that their relationshipdoes not have special characteristics, except for the general rivalry and battle for the love and wealthof their parents, as suggested by Shakespeare’s King Lear and the story of Cain and Abel in the OldTestament. See Sect. 5.2.3 for the discussion on strategic bequest motives and their effects on therelationships among siblings.2See Cartwright (2000, Chap.10) for considerations on the parent–children relationship from theperspective of the biology.


93

http://dx.doi.org/10.1007/978-4-431-55909-2_4

94 5 Resource Allocations within Families

In Sect. 5.3, we discuss transfers of resources from children to parents in the formof care and support. From the perspective of biology, unlike parents’ altruism towardstheir children, it is not easy to understand why grown-up children should providecare and support to elderly parents. We inquire into how the social norms of childrentaking care of their parents be maintained even if there were no laws that requirechildren to support their parents. Will there be any circumstances in which rationalgrown-up children support their elderly parents even if they cannot get any returnfrom them? We show the possibility and the instability of such a social norm. Then,we discuss the problem in relation to the sustainability of the pay-as-you-go pensionsystem, which is basically a system of forced transfers from the young to the elderly.Section5.4 concludes the chapter.

Understanding resource allocation within families is important in understandingthe formation of families discussed in the last chapter. Rational peoplemake decisionsconsidering their consequences in the future. For example, in the decision to investin children, parents will consider what benefits and costs children will bring to them.The variable σ(n) in the last chapter corresponds to the expected net benefits that nchildren will bring to their parents.

Hence, one of the important questions in the decision of family formation iswhether or not the children will really support their parents. If the expected transferfrom the children is not sufficiently large, parents will hesitate to have children andinvest in insurances and other financial assets. The most fundamental question in thischapter is indeed this problem of whether the children will support their parents. Justas Shakespeare’s King Lear suggests, the relationship between parents and childrencreates many tensions between altruism and self-interest.

5.2 Gift and Bequest

We first consider transfers from parents to children. Raising children and providingeducation to children can be viewed as the transfer of resources from parents tochildren. However, in the economic literature, mainly parents’ gifts and bequests tochildren have been analyzed partly because we can often get data on these variablesfor empirical studies. The results obtained in such an analysis can be applied tounderstand parents’ behavior towards raising and educating children.

In analyzing such transfers, we assume that parents have altruism towards theirchildren, while children do not have special altruism towards their parents. We ana-lyze simple games between parents and child under the above assumption.

5.2.1 Basic Model

Our basic model with a child and a parent can be described as follows. Let ck beconsumption by the child (k), cp be consumption by the parent (p). We assume that

5.2 Gift and Bequest 95

both cp and ck are normal goods. Let a be an action by the child, Uk(ck, a) be theutility function of the child,U p(cp, a,Uk(ck, a)) be the utility function of the parent.

As for the budget constraints, we assume that the incomes of the parent andchild, denoted by yp(a) and yk(a), are dependent on the action (a) of the child. Lety(a) ≡ yp(a) + yk(a) be the total income of the family.

As examples of children’s actiona,we can thinkof the children as being supportiveof their parent’s jobs and caring for the parents. Let T be the income transfer fromthe parent to the child. Then, the budget constraints of the child and the parent can bewritten as ck = yk(a) + T and cp = yp(a) − T = yp(a) + yk(a) − ck , respectively.Our basic model is a two-stage model in which the child takes action a first and thenthe parent chooses the transfer T = T (a) dependent on the child’s action a.

5.2.2 Rotten-Kid Theorem

Becker (1974) presented an interesting result, known as the “Rotten-KidTheorem.” Itdemonstrated thatwhen the parent has an altruistic preference for the child, the purelyself-interested child will take an action (e.g., supports and cares) which maximizesthe family income. The idea is that since the parent is altruistic, by taking the actionto maximize the total income yp(a) + yk(a), the selfish child (“Rotten Kid”) can getthe maximum transfer T (a) from the parent.

The essential factor that makes a selfish child exhibit seemingly altruistic behavioris the parents’ altruism towards the child. This suggests that altruism can provide anenvironment in which people have incentive to take cooperative actions. We provethis interesting result in a simplified model.

Becker (1974) assumed that the action of the child does not directly affect theutility of the child and the parent. Their utility functions can be written as Uk(ck)and U p(cp,Uk(ck)) respectively. The child’s action a can change the budget con-straints. Let a∗ be the action that maximizes yk(a) + yp(a), a+ be the action thatmaximizes yk(a), and assume a∗ �= a+. The following proposition indicates thatunder the assumptions here, if parents leave some transfer T (a+) even when thechild takes selfish action a+, then the selfish child will take cooperative action a∗because the parents are altruistic.3

Proposition 5.1 If T (a+) > 0, then the child takes the action a = a∗.

Proof In choosing action a, children will consider the reactions of their par-ents; thus, we solve the problem backwards. Since the parent’s budget constraintcan be written as cp = yp(a) + yk(a) − ck , given an action a, the parent canchoose the pair (ck, cp) to maximize the utility U p(cp,Uk(ck)) under the constraintcp + ck = yp(a) + yk(a). As cp and ck are both normal goods, the child’s optimalconsumption, from the perspective of the parent, can be expressed as the function

3The proposition is slightly different from the original one in Becker (1974), but let us show theessence of the Rotten-Kid Theorem by considering a simple sufficient condition in the proposition.


Fig. 5.1 Illustration of theRotten-Kid Theorem.Source: Hirshleifer (1977,Fig. 1)

ck

cpyp(a+) yp(a∗)

yk(a+)

yk(a∗)

c∗k

c∗p

P

Q

R

J

A

0

ck(yk(a) + yp(a)) that is increasing in the total income yk(a) + yp(a). As the child’sutility Uk(ck) is increasing in ck , it is optimal for the child to choose an action thatmaximizes yk(a) + yp(a) as long asUk(yk(a+)) < Uk(ck(yk(a∗) + yp(a∗))). Now,as T (a+) > 0 by assumption, we have ck(yk(a+) + yp(a+)) > yk(a+), and thusyk(a+) < ck(yk(a+) + yp(a+)) < ck(yk(a∗) + yp(a∗)). Hence, it is optimal for thechild to take the action a = a∗.

The basic idea of the “Rotten-Kid Theorem” can be observed in Fig. 5.1 ofHirshleifer (1977). In the figure, the curve PRQ represents the set of pairs of income(yp(a), yk(a)) generated by all possible actions a of the child. The point R is the pairof incomes (yp(a+), yk(a+)) and the point J indicates the pair (yp(a∗), yk(a∗)).

If point R is chosen, the budget constraint ck + cp = yp(a+) + yk(a+) can bedrawnas a segment passing throughpoint R. If point J is chosen, the budget constraintck + cp = yp(a∗) + yk(a∗) is the one passing on J . Under the assumption, T (a+) >

0, the child can choose to consume more than yk(a+), but the child can enjoy, bychoosing, a∗, a higher level of consumption like c∗

k at the parent’s optimal pointA. It is clear that the condition T (a+) > 0 is a sufficient condition to guaranteec∗k > yk(a+) and it is not a necessary condition for a∗ to be the optimal action.

Remark 5.1 (Sustainability of Altruism) Wintrobe (1981) argued that, for selfish par-ents to induce their child to choose the cooperative action a∗, it is enough to transfera small income to the child just exceeding the income y(a+), because the child isbetter off at a∗ than at a+. With such a transfer, the parent can keep more incomeand will have a higher probability of survival than the altruistic parent who transfersmore than necessary to induce the child to take the action a∗. The expected lowerprobability of survival of the altruistic parents casts some doubt on the sustainabilityof altruism.


Fig. 5.2 Strategic Bequest.Source: Bernheim et al.(1985, Fig. 2)

0

ck

a

D

K1 K2

AB

Cc∗∗k

c∗k c∗

k(a)

ck

a∗ a∗∗

P1

P2

P3

5.2.3 Strategic Bequests

The Rotten-Kid Theorem by Becker (1974) is interesting in the sense that if thealtruistic preference of parents is known to a child, even a selfish child will take theactions that maximize the total income of the family. In such a case, the parent doesnot take strategic actions and Pareto optimality is achieved.

However, Bernheim et al. (1985) showed that such a result does not hold in generalwhen the child’s action directly affects the parent’s utility. In such a case, the parentwill take strategic bequest dependent on the child’s action to induce the child to takedesirable actions.

To see it, let us consider the general utility function in the basic model above. LetUk(ck, a) andU p(cp, a,Uk(ck, a)) be the utility functions of the child and the parent,respectively. Aswe still have a relationship cp + ck = y(a) via the bequest, the utilityfunction of the parent can be rewritten as V p(ck, a) ≡ U p(y(a) − ck, a,Uk(ck, a))

dependent only on ck and a. We assume that the indifference curves of the utilityfunctions V p and Uk can be drawn as depicted in Fig. 5.2.

The indifference curves in Fig. 5.2 indicate the following assumptions.

(1) There exists an optimal pair (a, ck) for the parent (P), and the utility functiondefined over the plain looks like a mountain with D at its peak.4

(2) Although the utility of the child (K ) is higher as consumption ck increases, anincrease in a (e.g., care and support for parents) increases the utility to someextent until the level at which each indifference curve reaches its lowermostpoint and decreases the utility beyond it.

4Imagine a mountain whose contours can be drawn as the indifference curves labeled as P1, P2,and P3.


5.2.3.1 Subgame Perfect Equilibrium and Its Inefficiency

Now, consider a two-stage game in which the child chooses a first and the parentchooses T (a), the bequest depending on the child’s action. We can show that theoutcome of the subgame perfect equilibrium is given by point A in Fig. 5.2.

First, notice that for any action a, by the child, it is optimal for the parent tochoose the point on the line labeled as c∗

k (a), which is the collection of points wherea vertical line is tangent to an indifference curve of the parent. The line c∗

k (a) is the“best response function” of the parent that is dependent on the child’s action.5

Given such a best-response function c∗k (a) of the parent, that is, knowing that the

parent chooses the level c∗k (a) for each action a, the child chooses the action a which

maximizes its utility. Such a choice is given by the point A, where an indifferencecurve is tangent to the line c∗

k (a) because the child cannot find any other point witha higher utility.

Unfortunately, such a point is not Pareto optimal because at any point inside theshadowed lens-shaped area, utility of the child and that of the parent is higher thanthe ones at point A. In other words, at A there remains room for Pareto improvement,that is, room to increase the utility of both the child and the parent, which impliesthat A is not Pareto efficient.

5.2.3.2 Efficiency Under the Strategic Bequests

If the equilibrium outcome is inefficient, there will be some attempt between theparent and the child to improve their utility. Let us consider a new game in whichthe parent offers a contract to the child before the child chooses action a.

Let ck be the level of the child’s consumption when there is no bequest from theparent. In such a case, the child will choose point B, where an indifference curvelabeled as K2 is tangent to the line at ck . The indifference curve indicates the levelof the child’s utility when the parent does not provide any bequest.

Under the constraint that the parent need to guarantee such a level of utility to thechild, the parent will choose point (a∗∗, c∗∗

k ), where the parent’s utility is maximized.PointC in Fig. 5.2, where the parent’s indifference curve (labeled as P1) is tangent tothe child’s indifference curve K2, is the point. This point C is clearly Pareto efficientbecause we cannot find any point where the utility of the parent or the child canbe increased without deteriorating the utility of the other. As the parent’s utility ishigher at C than at A, there is an incentive for the parent to present a contract beforethe child takes an action. The optimal contract should include a threat as follows:If the child chooses action a∗∗, then the parent will give the bequest to fill the gapc∗∗k − ck , but give no bequest otherwise.

5Although it is assumed in Fig. 5.2 to be a simple line where c∗k (a) = c∗

k for all a, it does not haveto be such a line, in general.


If the child believes that the threat is credible then it will choose the action a∗∗and the parent will be able to achieve the highest utility.6 If the child believes that itis an empty threat, then it will choose a∗ at point A, where the child’s utility is muchhigher.7

5.2.3.3 Number of Children and Credibility of Strategy

The credibility of the above threat is determined by whether it is optimal for theparent to leave no bequest when the child does not take the action a∗∗. If it is judgedto be optimal, then the threat is credible; otherwise, the threat will be regarded as anempty one.

Bernheim et al. (1985)made an interesting claim that the credibility of the parent’sthereat depends on the number of the children. That is, they argued that if the parenthas only one child, the threat will be considered empty, but if the parent has morethan one children, the threat will be credible.

When there is only one child, independent of action a of the child, as long asck < c∗

k (a), it is optimal for the parent to give some bequest to the child against thethreat. Hence, the threat is empty.

When there is more than one child, the parent can write a more complex contracttomake the threat credible. To show the result formally, we need to introduce amodelwith multiple children, but its basic idea can be explained as follows.

Suppose that there are two children. Then, the parent can write a contract inwhich the child who chooses the desirable action (e.g., a∗∗ in the case above) getsthe bequest, while the one who does not, gets no bequest. In this case, the parentgiving no bequest to a child who did not take the desirable action can be optimalbecause the whole bequest will be given to the child who took the desirable action.Hence, the threat can be credible.

Under such a contract (“mechanism”) by the parent, selfish children will competefor the bequest from the parent and the parent will be able to get highest utility. Inthis case, having more than one child works as the commitment device to make thethreat credible.

6Note that the child is assumed to choose a∗∗ because its utility is the same as that at B, that is, theminimum utility. If the parent wants to choose the action a∗∗ for sure, the parent can increase thebequest slightly so that the utility of the child becomes strictly higher than the minimum.7The credibility of the parent giving the bequest to the child may be in question. However,we assume here that it is credible because the legal system will allow the parent to committo the statement. That is, the legal system works as the commitment device for the parent. SeeRemark 3.8 for the discussion on the credibility of the threat and the commitment.

http://dx.doi.org/10.1007/978-4-431-55909-2_3


5.2.3.4 Empirical Studies

Interestingly, economic theory suggests that the number of children can change thewelfare of the parent. However, is it possible to confirm the theoretical conjectureempirically? Bernhaiem et al. (1985) show some empirical evidences to check thevalidity of their theories of strategic bequests.

First, they show results of an econometric analysis based on micro-data collectedfrom various types of households. The results indicate that parents with higher wealthcan get higher “care and support” from their children, and children who have nobrothers and sisters tend to provide lesser “care and support” for their parents thanthose with brothers and sisters. In the analysis, the “care and support” for parents aremeasured by the average number of phone calls and/or visits per month. The resultsare consistent with the theory.

They also find some evidence in the literature that parents think of changing thelevels of bequests depending on the actions of their children (Sussman et al. 1970),and Japanese parents traditionally leave all bequests to the eldest son, who, in turn,is fully responsible for taking care of the parents (Horioka 1983).

As this evidence is also consistent with the theory, there is a good reason forus to consider that parents give bequests to their children at least partially basedon strategic motives. In response to such motives and behavior of parents, selfishchildren take altruistic actions to provide care and support to the parents.

It may be worth noting that an increase in families with one child, with decliningfertility rates in developed countries or the one-child policy in China, can change thebehavior of young people in the society, and thus the behavior and political demandsof the elderly people.

5.2.4 Incomplete Information and Timing of Bequests

Following Chami (1996), we now discuss the optimal timing of parents’ givingbequests to their children. In general, it is considered to be optimal for parents togive bequests at the final stage of their lives. The argument here indicates that sucha conjecture is not always correct.

Chami (1996) suggested that when there is incomplete information about theefforts of children, it might be better for parents to make transfers to their childrenmuch earlier than at the final stage of their lives in order to encourage their childrento make efforts.

Let us illustrate the result with a simple model. Let yk be the labor income of achild and yp be the parents’ income given exogenously. It is assumed that yk takesthe value yL or yH (yL < yH ). The function P(e), dependent on the child’s efforte, is the probability that the low income yL is realized. We assume P ′(e) < 0 and


P ′′(e) > 0.8 Namely, we assume that the probability of having low income becomessmaller when the child makes higher effort, but the marginal effects of effort onincome is getting smaller as its level goes up.

We assume that the utility function u(c) is only dependent on consumption c, andit satisfies the ordinary conditions u′ > 0 and u′′ < 0. For simplicity, we assume thatthe utility function of the parent and the child, denoted as u p and uk respectively, arethe same as the function u. The function v(e) represents the child’s disutility fromeffort e, and it satisfies the conditions v′(e) > 0, v′′(e) > 0, and v(0) = 0. The budgetconstraint of the child is given by ck = yk + b(yk) and that of the parent is given bycp + b(yk) = yp, where b(yk) represents the level of transfers from the parent to thechild.

In the following argument, the assumptions that the parent cannot observe thechild’s efforts and the bequest b(yk) will be dependent on the child’s income yk playimportant roles.

5.2.4.1 Ex-Ante Gifts (Pre-commitment)

Let us first consider the case in which parents transfer their wealth to their childrenbefore the children choose their effort levels. Then, the children’s utility can bedescribed by the following expected utility:

EUk(ck, e) = P(e)uk(yL + b) + (1 − P(e))uk(yH + b) − v(e)

If we define ukL ≡ u(yL + b) and ukH ≡ u(yH + b), we can easily show that ukL <

ukH and u′kL > u′

kH . On the other hand, we assume that the utility of the altruisticparent can be expressed by the following expected utility9:

EUp(cp,Uk(ck , e)) = P(e)[u p(yp − b) + ukL ] + (1 − P(e))[u p(yp − b) + ukH ] − v(e)

In this case, as the child chooses the effort level in the second stage, we solve theproblem backwards to find the optimal effort level for each bequest level b. The firstorder condition for the child’s optimization problem, given b, is given by

∂EUk

∂e= P ′(e)[ukL − ukH ] − v′(e) = 0

Let e∗C = e(b, yL , yH ) be the solution to the problem, that is, optimal level of

efforts. The first order condition above indicates that it is optimal for the child to

8P ′(e) and P ′′(e) represent the first derivative and second derivative of the function P . We use thesimilar notations below.9For simplicity, we assume that the child’s utility will not be discounted. Even if we consider the“inter-cohort discount factor,” as long as it is close to one, we can get the same results as thoseobtained here.


increase the effort levels until the marginal benefit of the effort equals its marginalcost.

Now, at the first stage, the parent chooses the optimal bequest level, consideringthe child’s best response e∗

C = e(b, yL , yH ). The first order condition is given by

∂EUp

∂b= u′

p − [P(e∗C )u′kL + (1 − P(e∗C ))u′

kH ] + (P ′(e∗C )[ukL − ukH ] − v′(e∗C ))∂e

∂b= u′

p − [P(e∗C )u′kL + (1 − P(e∗C ))u′

kH ] = 0

Now, under the assumption that the marginal utility is decreasing (u′′ < 0) and thefirst order condition (u′

p is the weighted average of u′kL and u′

kH ), we can provethat u′

k(yL + b∗C) > u′

p(yp − b∗C) > u′

k(yH + b∗C) must hold. Given that u p = uk =

u and the marginal utility is decreasing, we can show that the following inequalityholds:

yp − yH2

< b∗C <

yp − yL2

(5.1)

5.2.4.2 Ex-Post Gifts (The Last Word)

Then, we consider the case in which the parent chooses the level of bequest after therealization of the child’s income. In backwards induction, we first solve the parent’soptimization problem.

As the parent’s utility is given by u p(yp − b) + uk(yL + b) for yL and u p(yp −b) + uk(yH + b) for yH , the optimal pair of bequests (b∗

L , b∗H ) dependent on the

level of the child’s income must satisfy the first order conditions u′p(yp − b∗

L) =u′k(yL + b∗

L) and u′p(yp − b∗

H ) = u′k(yH + b∗

H ). As u p = uk = u and yL < yH ,

b∗H = yp − yH

2< b∗

C <yp − yL

2= b∗

L

must hold (see Eq. (5.1) for b∗C ). That is, it is optimal for the parent to set the bequest

at a low level when the child’s income is high, and a high level when the child’sincome is low. However, the following proposition suggests that such behavior ofthe parent can discourage the child’s effort, resulting in the optimal effort level, e∗

W ,to become lower than the case when the bequest is given before.

Proposition 5.2 e∗C > e∗

W .

Proof Given the pair of the bequests (b∗L , b

∗H ), the child chooses the effort level e∗

W .The first order condition is given by

P ′(e∗W )[uk(yL + b∗

L) − uk(yH + b∗H )] = v′(e∗

W )

On the other hand, when the bequest is given ex-ante, the first order condition isgiven by


P ′(e∗C)[uk(yL + b∗

C) − uk(yH + b∗C)] = v′(e∗

C)

As we have uk(yH + b∗H ) < uk(yH + b∗

C) and uk(yL + b∗L) > uk(yL + b∗

C),

uk(yH + b∗H ) − uk(yL + b∗

L) < uk(yH + b∗C) − uk(yL + b∗

C)

⇐⇒ − v′(e∗W )

P ′(e∗W )

< − v′(e∗C)

P ′(e∗C)

(5.2)

must hold. Now, if we define f (e) ≡ −v′(e)/P ′(e) and calculate the slope of thefunction, we have

f ′(e) = −[v′′P ′ − p′′v′

(P ′)2

]> 0

Hence, the Eq. (5.2) indicates that e∗W < e∗

C must hold.

The result above indicates that it is optimal for the altruistic parent to give thebequest to the children before the child chooses the effort level because the bequestat the final stage discourages the child from making efforts.

Remark 5.2 (Rotten-Kid Theorem and Samaritan’s Dilemma) Based on the two-period overlapping generations model, Bruce and Walden (1990) discussed the dif-ficulty faced by parents when making decisions regarding bequest to their children.When the transfer ismade later, selfish children try tomaximize the family income, assuggested by the Rotten-Kid Theorem. However, they may also try to reduce savingsunder the expectation of a higher transfer from their parents, as the above discussionsuggests. The result that people make little effort and saving with an expectationfor higher wealth transfers from their parents can be viewed as an example of the“Samaritan’s Dilemma” discussed in Remark 3.9. On the other hand, when parentsmake early transfers, children will be encouraged to make efforts, but the childrenwill no longer try to maximize the family income, which reduces the parents’ util-ity and creates inefficiency. We can interpret the abovementioned results by Chami(1996) as follows.When themoral hazard problem of the children’s efforts is serious,it is better to transfer wealth to the children early than at the final stage of life.

5.3 Support and Care

In the last section,we argued thatwhenparents have assets to transfer to their children,even selfish children could provide support and care to their parents. However, thereare many elderly people who have little or no asset to bequeath to their children. Inparticular, in most societies before economic development, most people live fromhand to mouth and rely on their children when they get old. Most young people

http://dx.doi.org/10.1007/978-4-431-55909-2_3


have indeed been providing resources to their elderly parents with no expectation ofreturns from them.

It is easy to explain such behavior as the children’s altruism towards their parents.However, from the perspective of biology or evolution, why do children have altru-ism towards their elderly parents who cannot make any contribution (e,g, wealth,household work, and nurturing grandparents) to their children?

In this section, we first assume that children have altruism and/or social normsto transfer resources to their elderly parents, and present a simple model to showthe effects of children transferring resources to their parents on our fertility deci-sions. Then, we inquire if we can really expect such behavior of the children. Wepresent some conditions under which selfish children make transfers to their elderlyparents who cannot make any contribution towards their children. However, suchconditions are likely to be unsatisfied and many elderly people will be abandoned bytheir children when markets for capital are developed. We claim that it is in such cir-cumstances that in many developed countries, the government introduced the publicpension system and provided public support to the elderly who could not receiveenough support from their children.

However, the introduction of the public pension system has created new socialproblems as its side effects. The side effects are serious especially under the pay-as-you-go pension system. We discuss policies to mitigate the side effects.

5.3.1 Basic Model for Private Support

To analyze the structure of children’s support for their elderly parents, we assumethat the children do not have a direct effect on the parent’s utility, but have indirecteffects via the budget constraint. We assume that it costs q to raise a child and omitdecisions on the labor supply.

Now, let Y be the income in the young period, c1 and c2 be the consumption in theyoung period and the elderly period, respectively. By denoting the interest rate as r ,saving as S, and the transfer from n children as σ(n), we can write the optimizationproblem in a simple two-period model as follows:

maxC1,C2

U (C1,C2) s.t. (1 + r)(C1 + qn) + C2 = (1 + r)Y + σ(n)

Note that the budget constraint above is derived by integrating C1 + qn + S = Yfor the young period and C2 = (1 + r)S + σ(n) for the elderly period via saving S.By solving the first order conditions of the problem above, we get the following twoconditions:

5.3 Support and Care 105

Fig. 5.3 Optimal number ofchildren

�0 n

′(n)1+r

n∗

q

U1

U2= 1 + r (5.3)

σ ′(n)

1 + r= q (5.4)

In Eq. (5.4), which describes the optimal fertility decisions, the left-hand siderepresents the present value of the marginal return from additional children and theright-hand side represents the marginal cost from additional children. If the left-handside is larger, then the parent should have more children. If the right-hand side islarger, then the parent should reduce the number of children. Hence, at the optimalchoice, two sides must be equal.

Now, if we assume that the income transfer σ(n) from n children is increasingin n and the marginal transfer σ ′(n), that is, transfer from the additional child, isdecreasing in n (i.e., σ ′′(n) < 0), the function σ ′(n) can be drawn as the downward-sloping curve depicted in Fig. 5.3. The optimal number of children n∗ is determined atthe point where the discounted value of the marginal transfer is equal to the marginalcost of the child (q).

Now,when the interest rate goes up, it is easy to observe in Fig. 5.3 that the optimalnumber of children will go down. This is because the increase in the interest ratemakes saving more attractive and people prefer investing in financial asset to havingchildren. Needless to say, an increase in the cost q of raising a child will reduce theoptimal number of children.

Such a result is intuitively understandable. However, can we really expect ourchildren to make transfers to us in the future? When the children are grown up, theydo not need to depend on their parents, and thus do not need to support their parentsvoluntarily. If they were rational, would they make any transfer to their parents?

According to Cigno (1991, Chap.9), under certain market conditions, selfish chil-dren can make transfers to elderly parents with no assets for bequests. Such an argu-ment indicates that the children’s transfer to their parents can be explained by theself-interest and not by their altruism.

The essential problem in the explanation by Cigno (1993) is the sustainability ofa norm that children should support their elderly parents. However, such norms caneasily collapse when capital markets are highly developed.


5.3.2 Extended Model: Sustainability of the Private Support

We assume that each individual lives for three periods. People can earn labor incomeonly in the middle age, and need some support in the young period and the elderlyperiod.

Let cti be the consumption in the i th period (i = 1, 2, 3) of an individual born attime t . Let yt and dt be the transfers to the young and the elderly, respectively, bythe people born at t in their middle age.

We denote the number of children of each individual born in t by nt , which is thenumber of people born in period t + 1. The utility function of the t th generation isgiven byUt = U (ct1, c

t2, c

t3) and the budget constraint is given by c

t2 + dt + ct+1

1 nt =yt and ct3 = dt+1nt . Finally, we define ρ t ≡ dt

ct1and name it the “rate of returns from

the generational transfers.”Given these definitions, we can rewrite the budget constraint as follows10:

ct1 + ct2ρ t

+ ct3ρ tρ t+1

= yt

ρ t(5.5)

This budget constraint is similar to the life-time budget constraint when there isaccess to the capital market where the rate of return (= 1 + rt ) at time t is equal toρ t . The first order condition for the utility maximization problem under the budgetconstraint above can be written as follows:

Ut1

Ut2

= ρ t ,Ut

2

Ut3

= ρ t+1

This shows that the optimal resource allocation must satisfy the following condition:

Ut−12

Ut−13

= Ut1

Ut2

= ρ t ≡ dt

ct1

Now, we consider the possibility of establishing the rule of transfers that satisfythe condition above. Notice that there is an incentive for individuals in the middleage to make the transfer to their parents at dt = 0. However, when they set dt = 0they will not have any income in the elderly period if there were no capital market.Hence, they will make the transfer dt > 0 to satisfy the above condition.

10We can derive the equation as follows. First, we rewrite the basic budget constraint as ct2 + dt =yt − ct+1

1 nt , and using the definitionρt ≡ dt

ct1, we derive dt = ρt ct1 and c

t+11 = dt+1

ρt+1 . By substituting

them, we have ct2 + ρt ct1 = yt − dt+1nt

ρt+1 . Now, as we have ct3 = dt+1nt , dividing both sides by ρt+1

and adding them to both sides of the budget constraint above, we have ρt ct1 + ct2 + ct3ρt+1 = yt .

Finally, by dividing both sides of the equation by ρt , we have Eq. (5.5).


Now, suppose that the capital market was developed to guarantee the rate ofreturn at Rt . In this case, there is a possibility that individuals in the middle age setthe transfer at dt = 0 and invest money in the capital markets because they can nowearn income from the capital market in their elderly period. We examine conditionsunder which such a breakdown of the intergenerational transfer system occurs.

When it is possible for individuals born at t to invest in the capital market, thebudget constraint under the rule of the intergenerational transfer will be given by

ct2 + ct3ρ t+1

= yt − dt

where we havemultiplied ρ t to the constraint (5.5) and rewrite it by using the relationct1ρ

t = dt .On the other hand, if the individuals break the rules and invest in the capital market

at rate Rt+1, the budget constraint can be written as follows:

ct2 + ct3Rt+1

= yt

By comparing the two conditions, we know that if ρ t+1 = Rt+1, it is better to invest inthe capital market and get no punishment. Hence, for the rule of the intergenerationaltransfers to survive, the rate of return ρ t+1 from intergenerational transfers must besufficiently high.

That is, it is necessary for the budget constraint under the rule of intergenerationaltransfers must satisfy the level of utility achieved by investing in the capital market,as depicted in Fig. 5.4. As the figure indicates, the rate of return ρ t+1 must be higherthan the market rate Rt+1, but it must be difficult for most of the families to guaranteesuch a high rate. In such a case, the intergenerational transfer systemwill break down,

Fig. 5.4 Conditions for therules of intergenerationaltransfer to be sustainable.Source: Cigno (1991,Fig. 9.1)

0

c3

ct3

c2ct2 yt −dt yt

t+1 Rt+1

P

Q

ρ


and people will stop relying on families and use capital markets to guarantee lives inthe elderly period. Once the system collapses, it is difficult for families to reconstructthe system.

Notice that in such a case people cannot expect to get transfer from their children,which implies σ ′(n) = 0 for any n in Fig. 5.3 and the optimal number of childbecomes zero.

Remark 5.3 (Laws Regarding Obligations to Support Parents) The model abovesuggests that a social norm for children supporting their elderly parents, from whomno return is expected, have been maintained by the children’s fear of the fact thatonce they stop following the norm they need to give up receiving support from theirchildren. With the development of the capital markets, the fear of the collapse offamily bonds has been increased, and thus the norm for intergenerational transfershas gradually faded away in modern society.

Under such structural changes in the family system, if everyone can have sufficientsavings and insurances in the financial markets, we do not have to worry about thewelfare of the people.However, there aremanypeoplewho cannot have suchfinancialassets. For them the collapse of the norm of intergenerational transfers is a greatfear. One way to mitigate their fears is to formalize the norm of the intergenerationaltransfer by creating a law such that thosewho do not obey the obligation of supportingtheir parents will be punished. Under the legal obligations, elderly people feel betterabout their life after their retirement. The law can be viewed as the institution toprevent people from breaking the norm of intergenerational transfers. Japan indeedhas such a law, and elderly people with children who have such an obligation are notentitled to apply for the public assistance (c.f. Sect. 9.3.1).

5.3.3 Public Support and Fertility Decisions

Given one has children, we have discussed about whether one can expect the supportfrom their children at an old age. It was indicated that when capital markets aredeveloped, people tend to use the markets to prepare for life after retirement, and thesocial norm of intergenerational transfer breaks down. It was also suggested that ifpeople do not obtain direct utility from their children, they will not have children, asthat may involve incurring costs with no returns. Hence, the development of capitalmarkets can break down the family support system and reduce fertility rates. Thiscan be viewed as one of the most important reasons for the social transformation indeveloped countries.

However, the capital markets are not sufficiently developed yet. Financial assetsthat fully insure against the risks of long life are not generally provided due to marketfailure problems. Furthermore, markets for care services are not well-developed inmany countries. Thus, many elderly people need to rely on their children to someextent. Hence, people still want to have children to prepare for life after retirementas suggested by the “old-age security hypothesis” (c.f. Sects. 4.2.1 and 8.2.2).

http://dx.doi.org/10.1007/978-4-431-55909-2_9

http://dx.doi.org/10.1007/978-4-431-55909-2_4

http://dx.doi.org/10.1007/978-4-431-55909-2_8


However, there are many children who do not provide enough support and careto their elderly parents due to the collapse of the norm of intergenerational transfers.Life-threatening problems occur in the life of many elderly people who cannot getsufficient support from their children.11

In response to such problems, the public pension system was introduced in manydeveloped countries (c.f. Sect. 2.3.2). We now investigate the effects of an increasein public support for the elderly people on our fertility decisions.

This problem has received the attention of many economists.Wewill demonstratethe basic results in a simple model in which the relationship between public pensionand fertility decision can be analyzed.

To clarify the structural problem, let us consider a simple pension system. Let Nt

be the population born in period-t . Each individual live for two periods, and earnswage w by providing one unit of labor in the young period. Under the pay-as-you-gosystem, young people pay the pension contribution at the rate τ to the labor incomeand receive the pension payment Zt in the elderly period. The budget constraint ofthe pay-as-you-go pension system can be written as Nt Zt = τwNt+1.

Now, we assume that individuals are homogeneous and have the same number ofchildren. If we let nt be the number of children that each individual has, we have thefollowing relationship Nt+1 = nt Nt . The budget constraint of the public pension canbe rewritten as Nt Zt = τwnt Nt and thus Zt = τwnt . That is, the pension paymentin the future depends on how many children each individual has, which means thatchildren have positive externality under the pay-as-you-go pension system.

For simplicity, we assume that the interest rate is zero, and each individual-i triesto maximize the utility dependent on the total consumption Ci

t summing up those inthe young and elderly periods. Hence, the optimization problem for each individualcan be written as follows.

maxCit ,n

it

Ui (Cit , n

it ) s.t. C

it + qnit = (1 − τ)w + Zt

where q is the cost to raise a child.However, the number of children that an individual chooses to have is so small

compared to the total population that the individual will choose the number of chil-dren without considering its effects on the pension budget. Hence, the individualoptimization problem will be solved by assuming that Zt is exogenous, and the firstorder condition for such a problem can be written as Un

UC= q.

On the other hand, in finding a socially optimum number of children, we need tosolve the maximization problem by assuming Zt = τwnt . The first order conditionis now given by Un

UC= q − τw. Figure 5.5 shows the difference between the two

optimality conditions.

11The problem will be especially serious when the traditional norm of children supporting theirparents breaks down in the process of economic development.

http://dx.doi.org/10.1007/978-4-431-55909-2_2


Fig. 5.5 Fertility decisionsunder the pay-as-you-gopension System

�0 n

UnUc

nit n∗t

q

q− τw

The figure indicates that, under the pay-as-you-go pension system, the number ofchildren chosen by the individual (nit ) will be smaller than the socially optimal level(n∗

t ) because each individual does not take into account the positive externality ofraising children. The number of children is inefficiently small under the pay-as-you-go pension system because people can free ride on the children raised by others.

As this is a classic externality problem (c.f. Sect. 3.3.5), we can discuss publicpolicies to mitigate the problem. For example, as suggested by Groezen et al. (2003),providing the child allowance, as a subsidy for raising children can be a good policyfor internalizing the positive externality.

To prove it, let σ be the allowance per child. The budget constraint will be writtenas Ci

t + qnit = (1 − τ)w + Zt + σnit . The first order condition for the individualmaximization problem, in which the level of the pension payment Zt is consideredto be exogenous, is given by Un

UC= q − σ . Now, by letting σ = τw, we can expect

individuals to choose the socially optimal number of children (n∗t ).

The child allowance can be seen as a policy to mitigate the side effects of the pay-as-you-go pension system by providing subsidies for the action (i.e., raising children)that has positive externality under the pension system. The model above suggeststhat the optimal level of the child allowance need to be determined depending on thesize of the pension system.

There are someotherways tomitigate the side effects of the pay-as-you-gopensionsystem.Notice that the positive externality of raising children occur under the pensionsystem because the pension payment is independent of the number of children anindividual raise. Hence, we can mitigate the problem by setting the pension paymentor pension contribution to be dependent on the number of children.12

Remark 5.4 (Model of Public Pension by Groezen et al. (2003)) The model above isuseful in understanding the side effects of the pay-as-you-go pension system, but it istoo simple to understand the total effects of the public pension system on the whole

12In a simple model with perfect capital markets, child allowance and the policy to reduce thepension payment when an individual has a smaller number of children are the same. However, thereare many people who agree with the former policy and disagree with the latter. From the perspectiveof policy design, it is interesting to see that people indicate different attitudes towards policies thatare essentially equivalent.

http://dx.doi.org/10.1007/978-4-431-55909-2_3


economy. From such a perspective, a two-period overlapping generations modelpresented by Groezen et al. (2003) is useful. It allows us to understand the effects ofthe public pension system not only on fertility decisions and labor supply, but alsoon saving decisions and capital formation. The model also allows us to consider thesocially optimal population.

The model first indicates that there is no guarantee that the number of childrenchosen in a market economy is Pareto efficient because children have external effectson per-capita consumption and capital. The introduction of the pay-as-you-go systemactually may increase the efficiency of such an economy, but will not achieve Paretoefficiency because it has its own external effects on the economy. It is indicatedthat Pareto efficiency can only be achieved when the additional policy, such as childallowance, is introduced to fully internalize the externality associated with the publicpension system.13 The model is worth studying as one that clarified the distortionaryeffects of the pay-as-you-go public pension system and indicated the effectivenessof policies such as child allowance in offsetting the distortionary effects of the publicpension system.


In this chapter, we considered the question why the intergenerational transfers, espe-cially the transfers between children and parents, are made within families. One ofthe most important questions is under what conditions the biologically selfish chil-dren will provide some support to their parents. It is an important question becausethe answer has deep implications for the design of an efficient social security system.

The Rotten-Kid Theorem suggested that selfish children will provide support totheir parents because the parents love their children. That is, because of the parents’altruism towards their children,which can be explained froma biological perspective,selfish children behave altruistically to support their parents to obtain the largestbequest from their altruistic parents. We have also seen that when parents are notpurely altruistic, but have some self-interest and wealth to be used for bequest totheir children, they can get care and support from their children by using the bequeststrategically.

Even when parents do not have any wealth, it was shown that the social normof children supporting their parents can be sustained if the capital markets are notwell-developed. However, when people have access to the well-developed financialmarkets our societies can change drastically because they can stop following thenorm and discard their parents by investing the money in the financial market toprepare for their life after retirement.

In the old days, the fear of not being able to get the support of the children once webreak the norm of supporting the elderly parents has been working well to make us

13In the model, it is assumed that the taxation to finance the child allowance does not have anydistortionary effect.


follow the norm.However, the development of the financial markets freed us from thefear because we need not fully depend on our children. The development of financialmarkets had deep impacts on the family relationship and the family formation. Itbrought a big structural transformation of our societies.

However, capital markets are far from complete. They usually fail to provide,say, good private pensions because of market failure problems. Furthermore, somepeople fail to have enough income and savings to make an adequate living after theretirement. Hence, most people have had children to prepare for the elderly periodand made some efforts to maintain the norm of intergenerational transfers withinfamilies. As the norm is very fragile, some countries created a law that requireschildren to support elderly parents; but, selfish children often fail to support theirparents.

With the loss of family supports, those who faced difficulty in making lives afterthe retirement demanded the expansion of public support. When the governmentsexpanded public support in the social security system, it further weakened the neces-sity of the children’s support. In particular, under the pay-as-you-go system, individ-uals can receive public pension without having any children by themselves. Peoplefree rode on the contributions by children of other families, which discouraged thefamily formation further.

The development of the market economy and the expansion of the social securityhave had deep impacts on the roles of families in our society. They reduced the returnsfrom forming families and impacted family formation. That is, the development ofthe market economy and the expansion of the social security system have reducedthe values of raising children and lowered the demand for children and marriage, andthus contributed to the decline in fertility rates. Measures must be implemented tomitigate these side effects of the development of two institutions: the market and thesocial security system. We will have further discussions on public policies in Japanto deal with these side effects in Part III.

References

Becker, G. (1974). A theory of social interactions. Journal of Political Economy, 82(6), 1063–1093.Bernheim, D., Shleifer, A., & Summers, L. H. (1985). The strategic bequest motives. Journal ofLabor Economics, 4(3), S151–182.

Bruce, N., & Waldman, M. (1990). The rotten kid theorem meets the Samaritan’s Dilemma. Quar-terly Journal of Economics, 105(1), 155–165.

Cartwright, J. (2000). Evolution and human behavior. Cambridge: MIT Press.Chami, R. (1996). King Lear’s Dilemma: Precommitment versus the last word. Economics Letters,52, 171–176.

Cigno, A. (1991). Economics of the family. Oxford: Oxford University Press.Cigno, A. (1993). Intergenerational transfers without altruism: Family, market and state. EuropeanJournal of Political Economy, 7, 505–518.

van Groezen, B., Leers, T., & Meijdam, A. C. (2003). Social security and endogenous fertility:Pensions and child allowances as Siamese Twins. Journal of Public Economics, 87(2), 233–251.

References 113

Hirshleifer, J. (1977). Shakespeare vs. Becker on altruism: The importance of having the last word.Journal of Economic Literature, 15(2), 500–502.

Horioka, C. (1983). The applicability of the life cycle model of savings to Japan. Mimeo: KyotoUniversity.

Sussman,M. B., Cates, J. N., & Smith, O. T. (1970). Inheritance and the family. London: Routledge.Wintrobe, R. (1981). It pays to do good, but not to do more good than it pays: A note on the survivalof altruism. Journal of Economic Behavior and Organization, 2, 201–213.

Chapter 6Traditional Communities

If a lazy sparrow intends appropriating the nest which acomrade is building, or even steals from it a few sprays of straw,the group interferes against the lazy comrade.

Kropotkin, P. (1902, Chap.2)

6.1 Introduction

In the following two chapters, we turn to the economic analysis of a community,which we defined in Chap.1 as a network of people who are not motivated to seekprofits. The family we discussed in the last two chapters is an example of the com-munity, but communities extend beyond families.

The fundamental question here is whether goods and services that are not suppliedin markets due to market failure problems can be provided in the communities.This question is important because to overcome market failures, cooperation amongcommunity members is necessary and maintaining this cooperation is not easy.

The most difficult part of maintaining cooperation is preventing the free riderproblems. If communities cannot prevent it, only those who free ride on the benefitsof the cooperation will survive and the cooperation will collapse. The quotation inthe beginning indicates that animals also try to prevent the free rider problem bypunishing those who try to free ride on the efforts of others.

Kropotkin, a Russian philosopher, scientist, and activist, wrote a book MutualAid based on the studies of animal behavior. He demonstrated that, unlike the imagereflected Darwin’s theory of natural selection, we can find ample examples of mutualaid in the animal kingdom. He then argued that human beings should create a societybased on mutual aid.

To maintain the norm of mutual aid in human society, it is necessary to punishthose who try to free ride. Unlike animals, who take actions based on their instincts,human beings are “thinking reed” as a French philosopher Blaise Pascal (1623–1662) put it in his book Pensées and thus imposing punishment on free riders must


115

http://dx.doi.org/10.1007/978-4-431-55909-2_1

116 6 Traditional Communities

be compatible with rationality. Under what conditions can the effective punishmentto induce cooperation be considered as rational? This is the crux of this chapter.

We analyze “traditional communities” in this chapter and “new communities” inthe next chapter. The definitions for “traditional” and “new” are inevitably vague.The new communities refer to the ones in which networks of people are tied bythe strong will of the people, while in the traditional communities the networks areconnected by chance or weakwills. Examples of new communities are NPOs, NGOs,and networks of those who get together for some activities. As we will see in thenext section, these communities may be referred to as “associations” rather than“communities.”

Examples of traditional communities include families, local communities, andreligious communities. Family is an example of a traditional community and it is veryspecial in the sense that formal laws define its various aspects.1 In most traditionalcommunities other than the family, there usually does not exist any law that definesits scope.2

The relationships in traditional communities are generally informal, and thusit is difficult to define formal rules for maintaining cooperation. Furthermore, intraditional communities other than families, altruism is generally not expected, andmonitoring costs and transaction costs are usually high. Hence, it is usually difficultto maintain cooperation in traditional communities.3

In the next section, we consider the case in which self-interested individualsvoluntarily try to provide “public goods,” which are not well provided by the market.We first show that the level of public goods provided will be inefficiently smallbecause each individual tries to free ride on the provisions of others (Sect. 6.2.1). Wewill also show that even if the government increases the provision of public goods,the total level of public goods may not change because private provision of publicgoods may be crowded out (Sect. 6.2.2).

In Sect. 6.3, we discuss how the traditional communities have tried to mitigatethe undersupply of public goods. We first argue that if there are long-term relation-ships among community members efficient public goods provision may be guaran-teed (Sect. 6.3.1). When the relationship becomes short-termed due to, say, highermobility in the market economy, the possibility will be lowered. However, in anevolutionary game, it is shown that communities may be able to achieve the effi-cient outcome through cooperation among members with short-term relationships if

1Couples without children may be categorized as new communities because they are usually con-nected by strong wills. However, children and parents are connected by chance and thus should becategorized as traditional communities.2New communities such as NPOs and NGOs, on the other hand, are often well defined by laws.3It is usually more difficult than in new communities, where relationships are more formal andformal rules to induce cooperation are more easily introduced thanks to the strong wills and lawsthat bind the members.

6.1 Introduction 117

there are sufficient members who punish free riders. Such states, however, are fragilebecause the number of people who impose costly punishment on free riders is likelyto be decreasing (Sect. 6.3.2). Section6.4 concludes the chapter.

6.2 Private Provision of Public Goods

We have already seen, in the donation games in Chap.3, that the voluntary provisionof public goods is inefficiently small.We extend themodel by allowing individuals tochoose an action (e.g., donation) from a budget constraint rather than two actions (i.e.,yes or no). We use the model by Bergstrom et al. (1986) and clarify the problem ofprivate provision of public goods as the essential problem in many communities. Tosimplify the explanations, we consider the simplest community with two individuals,but the basic results can be easily extended to the oneswithmore than two individuals.

6.2.1 Model

Assume that individual-i (i = 1, 2) has income wi and the utility function ui (G, xi )

which is dependent on public goods G and private goods xi . We assume G andxi are both normal goods.4 As the providers of the public goods, we consider thegovernment in addition to the two individuals. Let g1, g2, and g be the public goodsprovided by individual-1, individual-2, and the government, respectively. The levelof public goods is defined by G = g1 + g2 + g. For simplicity, we assume that theprice of public goods is p and private goods is 1. We first assume g = 0 and analyzethe problem of private provision of public goods.

By setting the utility level of individual-2 at u2, we can find the Pareto efficientallocation by solving the following optimization problem5:

maxx1,x2,G

u1(G, x1) s.t. x1 + x2 + pG = w1 + w2, u2(G, x2) = u2

By substituting the resource constraint in the utility function of individual-2, theproblem can be rewritten as follows.

maxx1,G

u1(G, x1) s.t. u2(G, w1 + w2 − x1 − pG) = u2

By defining the Lagrange function associated with the maximization problem asL ≡ u1(G, x1) + λ(u2(G, w1 + w2 − x1 − pG) − u2) we have the following first

4Normal goods are goods whose demand increases with the increase in disposable income.5The definition of Pareto efficiency (c.f. footnote 2 in Chap.2) will be satisfied by fixing the utilityof an individual and maximizing the utility of the other individual under the resource constraint.

http://dx.doi.org/10.1007/978-4-431-55909-2_3

http://dx.doi.org/10.1007/978-4-431-55909-2_2


order conditions:

∂u1

∂G+ λ

(∂u2

∂G− ∂u2

∂x1

)= 0 and

∂u1

∂x1− λ

∂u2

∂x1= 0

By rearranging the two conditions, we have the following condition known as theSamuelson condition for optimal supply of public goods:

MRS1 + MRS2 = p

whereMRSi ≡ ∂ui∂G / ∂ui

∂xiis the “marginal rate of substitution between the public goods

and private goods.” It represents the utility (i.e., benefit) that the additional publicgoods generate, expressed by the quantity of the private goods.6 The price of pub-lic goods p on the right-hand side is the marginal rate of transformation betweenthe public goods and private goods, which is the price (i.e., cost) of producing anadditional unit of public good in terms of the quantity of private goods. Hence, thecondition indicates that it is optimal to increase the public goods until the additionaltotal benefit (MRS1 + MRS2) is equal to the additional cost. We denote the efficientlevel of public goods that satisfies this condition with G∗∗.

6.2.2 Voluntary Provision of Public Goods

Now, we consider a case in which each individual-i chooses private provision ofpublic goods gi to maximize the utility under the budget constraint, given the privateprovision of the other individual. The optimization problem can be written in twoforms:

(I) maxx1,g1

u1(g1 + g∗2 , x1) s.t. x1 + g1 = w1

(II) maxx1,G

u1(G, x1) s.t. x1 + G = w1 + g∗2 , G ≥ g∗

2

Figure 6.1 depicts the problem (II). In this figure, we represent the level of publicgoods on the horizontal axis so that the marginal rate of substitution (MRSi ) is givenby the slope of the indifference curve.

We take g∗2 as given, and draw the budget constraint x1 = −pG + (w1 + g∗

2) asa line segment with slope −p and intercept a, where a ≡ w1 + g∗

2 . Considering thatthe choice of individual-1 must satisfy G ≥ g∗

2 , the budget constraint for individual-1to choose a pair (G, x1) will be given by the segment ba below the point b.

6The numerator represents the increase in utility due to a small increase in public goods G, whilethe denominator represents the increase in utility due to a small increase in private goods xi . Thus,if we divide the former by the latter we can calculate the quantity of private goods xi we need toachieve the increase in utility caused by additional public goods G.

6.2 Private Provision of Public Goods 119

Fig. 6.1 Private provision ofpublic goods

0 G

x1

w1

G∗ G∗∗g∗2

x∗1

x∗∗1

a

ap

−p

b

e

f

The optimal choice will be given at e where an indifference curve for the utilityfunction u1(G, x1) touches the budget constraint. At that point, the marginal rate ofsubstitution is equal to the slope of the budget constraint −p, and thus MRS1 = pholds. Needless to say, the optimal private provision g∗

1 of individual-1 is calculatedby G∗ − g∗

2 .The solution to the same problem for individual-2 must satisfy the condition

MRS∗1 = MRS∗

2 = p at the optimal level of private provisions of public goods G∗ =g∗1 + g∗

2 . As we had MRS∗∗1 + MRS∗∗

2 = 1 at the socially optimal level of publicgoods G∗∗ and themarginal rates of substitution are positive, wemust haveMRS∗∗

1 <

1 and MRS∗∗2 < 1 at the social optimum.

It is easy to verify that G∗ < G∗∗ (c.f. Fig. 6.1) because the marginal rate ofsubstitution decreases with the increase in public goods. Hence, the private provisionof public goods results in undersupply, as is summarized in the following proposition.

Proposition 6.1 When public goods are only privately supplied, the total supply ofpublic goods G∗ is smaller than the socially optimal level G∗.

6.2.3 Crowding-Out Theorem

Now, we introduce the government in the model above. Suppose that the governmentimposes a tax τi ≥ 0 on individual-i (i = 1, 2) and provides public goods g > 0under the budget constraint τ1 + τ2 = g. The total supply of public goods is givenby G = g1 + g2 + g. The budget constraint of individual-i is assumed to be givenby xi + gi = wi − τi .

The maximization problem of individual-1, for example, can be expressed by thefollowing one in which public goods supplied by individual-2 and the governmentare given.

maxx1,g1

u1(g1 + g∗2 + g, x1) s.t. x1 + g1 = w1 − τ1


Fig. 6.2 Illustration of the crowding-out theorem

In this case, we can show that if the amount of public goods supplied by thegovernment is small, it cannot change the total supply of public goods (case (i)in Proposition 6.2 below), because the private provisions will be reduced by thesame amount as the increase in the tax payment (c.f. Fig. 6.2).7 This result indicatesthe existence of a “crowding-out effect” (c.f. Sect. 2.3.1) in that, the government’sprovision of public goods can crowd out the private provision of public goods.8 Ithas been empirically shown that there exist such crowding-out effects (c.f. Remark6.4).

Proposition 6.2 (Crowding-Out Theorem) Let G∗(g) be the total public goodswhen the government provides g. Then the following property holds.

(i) If g ≤ G∗(0) and τi ≤ g∗i for every i , then we have G∗(g) = G∗(0) at the

equilibrium.(ii) If g ≤ G∗(0) and τi > g∗

i for some i , then we have G∗(g) > G∗(0) at theequilibrium.

(iii) If g > G∗(0), then we have G∗(g) > G∗(0) and g∗1 = g∗

2 = 0 at the equilib-rium.

Proof See Bergstrom et al. (1986; Theorem 4) for the proof for the general case.9

The theorem suggests that the government can increase the total public goodswith their public provision only if some taxpayers provide no public goods (c.f. caseof (ii) and (iii) in Proposition 6.2).

The result implies that the lack of private provision of public goods need not beviewed in a negative light because it means that the government is providing public

7The result may be well-explained if we consider that individuals think that they made their contri-bution to public goods through tax payment, thus reducing their private provision.8See Warr (1982) and Bergstrom et al. (1986) for more details and discussions.9Althoughwe omit the proof, we recommend readerswho are interested in the theoretical researchesto try solving a proof based on Fig. 6.2.

http://dx.doi.org/10.1007/978-4-431-55909-2_2


goods at the optimal level or at least beyond the inefficient level attained by privateprovisions.

In Japan, it has been pointed out that voluntary actions and activities of NPOs andNGOswere quite low compared to, say, the U.S. However, the crowding-out theoremsuggests that it may be because the government has been providing sufficiently highlevels of public goods so that people do not feel that they should provide the publicgoods. In other words, societies in which public goods were provided actively byvolunteers and NPOs are not really good societies, because the levels of public goodsare inefficiently low.

The story will completely change if the government faces a serious budget con-straint or government failure problems. In such a case, the government’s provisionwill become lower than the optimal level that people want, and it will be very impor-tant to encourage voluntary provision of public goods to communities. Recent socialattempts to encourage the voluntary provision of public goods in Japan can be under-stood from this perspective.

When the governments are providing insufficient public goods, the crowding-outtheorem suggests that the reduction in government’s provision of public goods doesnot lower the total supply of public goods because private provision will be increasedto offset the reduction in public provision. Hence, if the government can no longerprovide some public goods sufficiently and private agents started providing themvoluntarily, the crowding-out theorem suggests that, considering the distortions oftaxation, it may be better for the government to stop its provision of public goods.

Hence, we may be able to argue that, from the perspective of efficiency, if thegovernment can supply sufficiently high levels of public goods, it should play agreater role; otherwise, it should play a minimal role. The crowding-out theoremseems to have an important implication in the discussion on whether the governmentshould be big or small in modern society.

Remark 6.1 (Neutrality Theorem) In amodel with private provision of public goods,there exists an interesting theoretical result, known as the “neutrality theorem,” thata small change in the redistribution policy will not change the welfare of the people(Warr 1983). The basic reason for such a result can be inferred from the crowding-outtheorem. Consider a policy to increase the tax on rich individual-1 by τ to transfer itto poor individual-2. Given that τ ≤ g∗

1 , the argument in the crowding-out theoremsuggests that the individual-1 will decrease the private provision of public goods tog∗1 − τ and individual-2will increase it to g∗

2 + τ becausebydoing so, the equilibriumallocation (x∗

1 , x∗2 , G∗) before the policy change is achieved. As the allocation was

optimal for everyone before the policy change it should be optimal after the policychange. Note that the change in the redistributive policy did not change the totalpublic goods nor individual consumption. When the poor people are also makingprivate provision of public goods (e.g., donations), a small increase in the subsidy tothe poor financed by the tax on the rich will not raise the welfare of the poor becausethey will increase the private provision of public goods to offset the reduction ofthose by the rich.


Remark 6.2 (Extension of the Model) Let us extend the model to consider the effectsof increasing the number of community members. Suppose that there are n membersin the community, where n is a natural number greater than one. For simplicity, weassume that each individual-i has the same budget constraint xi + gi = w and theutility function u(G, xi ). Just like in the basicmodel, social efficiency can be achievedwhenMRS∗∗

1 + . . . + MRS∗∗n = nMRS∗∗ = p and the individual rationality requires

MRS∗ = p for each individual. Hence, we have

nMRS(G∗∗, x∗∗) = MRS(G∗, x∗)

As MRS(G, x) ≡ uG (G,x)

ux (G,x)decreases as G increases, we have G∗ < G∗∗ and G∗ gets

smaller and smaller as the n number of community members increases, which sug-gests that the inefficiency of the private provision of public goods increases as thecommunity becomes larger.

Remark 6.3 (Voluntary Provision of Public Goods in the War of Attrition Game)Bliss and Nalebuff (1984) analyzed the problems of private provision of publicgoods by using a game of the war of attrition well-known in auction theories. Theirargument can be summarized as follows. Imagine a crowded ball-room in which theair quality is low. The air quality can be improved if someone provides a public goodto open the windows for a few minutes. Although the number of individuals in theball-room is quite large, we can expect that it will not take long before someoneprovides the public goods. In this case, the inefficiency associated with the privatepublic good will decrease as the number of individuals increase. This result can beobtained because the individual who provides public goods in this case is the onewhoincurs the least cost of providing it. Further, as the number of individuals increasesthe least cost of individuals in the community, the time taken by someone to providethe public goods is likely to reduce further. The argument suggests that the results ofthe inefficiency of providing private public goods depend on the model specificationand the type of public goods. It also suggests that by introducing time we can getsome interesting insights into our public goods provision.

6.2.4 More Discussion on the Crowding-Out Effects

Although the crowding-out theorem has some interesting suggestions with respectto our society, it is based on the assumption that individuals are only interested in thetotal level of public goods. In such a case, if the number of community members islarge, we expect people to try to free ride on the provisions by others and provide nopublic goods as their contribution to the total level becomes negligible (c.f. Remark6.2).

However, in reality it is indicated that some people do provide public goodseven if their contributions to the whole are negligible (e.g., Andreoni 1988). One ofthe possible explanations for such a behavior is that they have altruism or “impure


altruism” with which selfish people can feel great joys by giving to others. In suchcases, it is easy to show that perfect crowding-out will not occur.

Furthermore, it has been pointed out the government’s provision of public goods,say, to NPOs and universities can induce private donations to them. This is knownas the “crowding-in effect.” An explanation for such effects is that the governmentproviding support to some institutions plays the role of signaling the high qualityof services provided by the institutions, which increase the donations of those withincomplete information about the quality of services provided by them.

Such crowding-in effects have been empirically confirmed (c.f. Remark 6.4below). If an increase in the government’s subsidies crowds out private donations,thosewho receive subsidies will try to hide this information. However, inmany cases,they appeal the fact to the public, suggesting the existence of crowding-in effects.10

Remark 6.4 (Empirical Studies on the Crowding-Out Effects) The crowding-outeffect of public provision of public goods on private provisions has gained the atten-tion of many researchers, and thus, many empirical studies have been conductedon it. Many studies indicate that crowding-out effects are not perfect, but partiallyobserved. Although experimental studies tend to indicate large effects, empiricalstudies using data of actual donations tend to report small or no effects (c.f. Manzoorand Straub 2005). There are also studies that show the existence of the crowding-ineffects discussed above.11 These empirical and experimental studies suggest that theprivate provision of public goods such as donations depends on individual prefer-ences and information about benefits and costs of public goods.Although it is difficultto derive a general result about the crowding-out effects, we believe that recognizingthe interactions between the public and private provision of public goods and thepossibility of crowding-out or crowding-in effects is important in considering rolesof the government in our society.

Remark 6.5 (A Numerical Example of Voluntary Provision of Public Goods) Itseems useful to confirm the results in this section by using the simple model andnumerical examples. Let us consider a community with two individuals. Supposethat individual-i (i = 1, 2) has the same utility function u(G, xi ) ≡ (G)

12 (xi )

12 and

the budget constraint xi + gi = 3, where G = g1 + g2 and assume the price of publicgoods is set at p = 1. As a resource constraint we consider the simple sum xi + x2 +G = 6 of the budget constraints of two individuals.

In this model, we can show that (g∗∗1 , x∗∗

1 , g∗∗2 , x∗∗

2 ) = ( 32 ,32 ,

32 ,

32 ) is a Pareto

efficient resource allocation.12 On the other hand, the resource allocations underoptimal private provision of each individual is given by x∗

1 = x∗2 = 2 and g∗

1 = g∗2 =

10Some analyses show that whether crowding-out effects or crowding-in effects occur depends onhow governments provide public goods. See, for example, Rege (2004).11See, for example, Rose-Ackerman (1986), Khanna and Sandler (2000), Payne (2001) and Heutel(2009).12However, this is not the unique Pareto efficient allocation. For example, (g∗∗

1 , x∗∗1 , g∗∗

2 , x∗∗2 ) =

(2, 1, 1, 2) is an allocation that satisfies the conditions x1 + x2 = 3 and G ≡ g1 + g2 = 3. Thereare infinite Pareto efficient resource allocations.


1. This numerical example indeed indicates that the private provision of public goodsG∗(=2) is much lower than the socially optimal level G∗∗(=3). We can also confirmthe crowding-out effects illustrated in Fig. 6.2, which we leave for readers.

6.3 Social Norm and Social Punishment

We have seen that the voluntary provision of public goods by self-interested indi-viduals becomes inefficiently small, implying that mutual aid within communitiescannot be expected very much. The essential problem is the externalities associatedwith the public goods provided by individuals. We then considered the possibility ofthe government increasing its provision of public goods to mitigate the problem. Theresults suggested that it would not improve the problem much unless it increases thelevel of public goods a lot.

The Japanese government indeed expanded its roles in its attempt to respond tothe decline in mutual aid in traditional communities. However, it has probably con-tributed towards reducing the mutual aid in traditional communities further throughits crowding-out effects. The question is whether there were ways to maintain mutualaid within the communities. To answer this question, we need to find out how tra-ditional communities have prevented the free rider problems and maintained mutualaid within communities.

6.3.1 Solutions to Externality Problems by Communities

The question is how the communities have mitigated the problems of externalities.In Sect. 3.3.5, we have shown five ways to solve them: (1) legal contracts, (2) psy-chological punishment (or morals), (3) long-run relationship, (4) tax and subsidies,and (5) creation of markets. Among them, we quickly review the first three, whichare important in understanding the mechanisms of communities.

The first solution requires us to have legal contracts for taking socially desirableactions and accepting legal punishments if we fail to take such actions. This is a veryeffective way to prevent the free rider problem and has been used frequently in manycommunities.

However, the solution works only when the legal system is well established andthe cost of using it is small, which is often not the case especially in underdevelopedor developing countries. Furthermore, there are so many externality problems inour communities that it is almost impossible to engage in legal contracts to inducemembers to take desirable actions.

Another popular solution to the problem is the use of psychological punishments.Inmost communities, there is a system ofmoral, religious, or ethical codes that definedesirable and undesirable actions, whichmake thosewho fail to take desirable actionsfeel guilty. Ifmembers of the community indeed feel guilty, this type of psychological

http://dx.doi.org/10.1007/978-4-431-55909-2_3

6.3 Social Norm and Social Punishment 125

punishment will work to discourage them from taking undesirable actions such asfree-riding.

We can find numerous examples in which communities make their membersbelieve in such codes via education and stories of various folklore and legends.However, for such a solution to work well community members need to share andbelieve in the same morals, religion, and ethics, and feel guilty about deviating fromthe codes. As science develops and many natural phenomena are explained, peoplehave difficulty in believing in supernatural stories. Although relying on psychologicalpunishment is an inexpensive way to mitigate the problem, its effects seem to havebecome weaker.

To discourage people from taking inefficient actions, we probably need punish-ments that impose some real costs on the defectors. In fact, there are many actionsthat we can take to impose real costs on others. Taking violent actions or deprivingthem fromwealth can be considered as good examples. However, taking such actionsis often illegal and risks receiving legal punishments. Hence, informal punishmentsneed to be weak.

Social ostracism of defectors, wherein members of the community do not providelove, help, and services to those who take undesirable actions, has been a popularway of punishment. For such weak punishment to work, long-run relationships arenecessary. If, say, the defectors can leave their communities easily, then ostracismdoes not work well.13

In Japanese traditional communities,we argue thatmutual aid has beenmaintainedthrough social ostracism, known as “murahachibu” in Japanese.Wefirst theoreticallyshow how suchweak punishment works to prevent undesirable actions of communitymembers and then give an example of murahachibu that has been used in Japanesecommunities when the mobility of the people was not high. We will also see thatthe conditions for the communities to retain mutual aid have been lost as the marketeconomy prevails in Japan.

6.3.2 Social Norms in Communities

The basic strategy for communities tomaintainmutual aid by norms and punishmentscan be explained by repeated games. In the numerical example in Remark 6.5 above,the desirable action for the communitymemberswas g∗∗

1 = g∗∗2 = 3

2 ,while the selfishaction that lowers the efficiency of the society was given by g∗

1 = g∗2 = 1. In an

infinitely repeated game, the community may be able to make each communitymember follow the social norm to take the action g∗∗

i .A strategy in the repeated game specifies an action for the history of each player’s

actions in the game. Consider the following trigger strategy (σ T Ri ) of individual-i :

13In Sect. 6.3.4 below we depict a model of evolutionary games in which the long-run relationshipis not necessary to prevent selfish behavior.


(1) Start from the action gi = g∗∗i (= 3

2 ).(2) As long as each takes the action gi = g∗∗

i in each period, take the action gi = g∗∗i .

(3) If someone takes the action gi �= g∗∗i in some period, take the action gi = g∗

i (=1) after that.

As we have seen in Sect. 3.3.3, if players place high weights on future payoffs,that is, if they have sufficiently high discount factors, each player choosing the triggerstrategies becomes the subgame perfect equilibrium. By applying the calculation tothe case here, we can easily show that the condition for the discount factor δi is givenby δi ≥ 9 − 6

√2 = 0.514 · · · .

If we define the social norm such that each member of the community chooses thesocially desirable action, the social norm will be maintained only if each member isafraid of other members starting to take punitive actions when he/she did not followthe social norm. If there is a long-run relationship in traditional communities and ifeach member highly values the payoffs in the future, the communities can maintain asocially desirable state by making the rules of the punishment on defectors commonknowledge in the communities.

Remark 6.6 (Punishment Rules in Repeated Games) In Sect. 3.3.3,we have seen thatstrategies that impose punishments on undesirable actions, such as trigger strategiesand tit-for-tat strategies, work to make players take desirable actions. We can thinkof various punishment rules that we can take in our strategies, but Abreu (1988)showed that, to consider whether players choose a desirable action, it is enough toconsider “simple strategy profiles” concerning the penal code that will be imposedwhen a player deviates from the desirable actions.

What is interesting in this argument is that the simple strategy profile specifiesthe rule of the punishment on those who do not punish the deviators. When thereare only two players, we do not have to consider such a possibility. When there arethree or more players, a player may want to avoid punishing the defectors becauseimposing the punishment requires incurring some cost and players have incentivesto free ride on the punishment by other players. For the community to make everymember take desirable actions, it needs to punish those who try to avoid punishingthe defectors. We can indeed find such an example in the penal codes of the Japaneselocal communities for inducing desirable actions by community members.

6.3.3 Punishment Mechanisms in Japanese LocalCommunities

Let us look into the mechanism for encouraging people to take desirable actions intraditional communities in Japan.Watanabe and Hojo (1975) describe a problem thatoccurred in a small community in Meji era.

To explain the punishment mechanism in a small village (buraku), let us explain it usingan example of an incident in the Meiji period and the punishment involved in a village

http://dx.doi.org/10.1007/978-4-431-55909-2_3

http://dx.doi.org/10.1007/978-4-431-55909-2_3


called Hirano. . . . the incident started from . . . the day when a “community detective” foundthat one of the dominant person in Hirano was working to cut grasses on a Sabbath daydetermined by the village every year.

The “community detective” is sometimes called the “field guardian.” Five distin-guished young men are selected as detectives from among the community firemen.They are “at the end of the punishment mechanism of Hirano” and expected to playtwo roles. The first is policing the “criminal activities against the village norm” andthe second is policing the “inadequate use of the iriaichi (commons)14 which violatesthe rule.15”

The detective tried to impose a fine on the person who did not follow the norm,but the deviator not only refused to pay the fine but also tried to avoid it. Hence, theproblemwas reported to the community fire brigade, which is the higher organizationthat the detectives belong to. It was formally judged that the violator should pay asmall fine. The violator refused to pay the fine. Hence, the problem was sent to thehigher decision body of the community and then to the village government. The finewas raised gradually as the problemmoved through higher decision bodies. However,the violator refused to pay the fine.

At this point, the violator was sentenced to get the “murahachibu,” which involved a 100%rather than 80% reduction of communal mutual assistance. His relationship with the villagewas completely cut off. It was a very severe punishment in which no one was allowed tohelp in funerals and fire-fighting at his house. It is said that when his house caught fire afterthe incident, nobody helped put out the fire in fear of receiving the same punishment.

The decision about the murahachibu was made at the general assembly of thecommunity, which is the “highest organ of the punishment system of the villagecommunity.”When the punishment by the highest organ is not enough, it can imposeleagal charges. However, it is said that such a case is very rare. Watanabe and Hojo(1975) summarize the incident and the punishment as follows:

As the case above clarifies, there were norms set by the village community of Hirano andthe punishment mechanism for those who violated it. Against the “criminal” actions, thepunishment mechanism of the village community is activated, and the community tried tosolve the problem only through the punishment mechanism to the extent possible.

Remark 6.7 (Discussion by Ostrom) Ostrom, who received the Nobel Prize in Eco-nomics, has demonstrated many cases in which communities have solved problemsof the “tragedy of the commons.” It was found that self-organizations similar tolocal communities discussed in this section have played important roles in achievingefficient resource allocations in our society. Ostrom (1990, Chap.3), for example,suggests that there are common characteristics in self-organizations that help themsucceed in avoiding the “tragedy of the commons,” which can be summarized intoseven principles.

14See Sect. 2.3.2 for “iriaichi.15The detectives were said to be given some rights to judge the violator and were allowed to chargefines in the form of either money or sake (a Japanese rice wine) to cover the cost of the policingactivities.

http://dx.doi.org/10.1007/978-4-431-55909-2_2


Among such principles, monitoring is considered to be important, to find anyonewho tries to break the rule to avoid the tragedy of the commons and impose ade-quate punishments on them when they are found. Various examples are presented toillustrate that point, which include the punishment mechanism of the Japanese localcommunities, management of meadows in Switzerland, and the self-organizationsystem in the management of the water resources in Spain and Philippines. In eachcase, it is indicated that there exist some clever rules to avoid the tragedy of the com-mons and give adequate punishments to the defectors. In most cases, punishmenton the defectors takes the form of fines, and the complex punishment mechanismlike murahachibu in Japanese local communities is rare. It is also shown that thereare cases in which quarrels occur between those who caught the deviator and theone who was caught. Such cases suggest that the cost of finding and punishing thedefector can be quite large and that it is not so easy to prevent the selfish behavior thatcauses the tragedy of the commons. We can find many examples that are consistentwith the predictions of our model.

6.3.4 Evolution of the Social Norm

We have so far repeatedly claimed that there must be a long-term relationship forthe punishment to work well on those who did not follow the social norm. As themarket economy prevails and the labor markets become highly developed, we canno longer rely on the long-run relationship because we can leave communities andfind jobs easily to make a living. Such an economic change has made it difficult forcommunities to maintain mutual aid.

In such societies with high mobility, psychological punishment seems to be theonly way other than legal punishment to prevent the socially undesirable selfishactions. However, when we look at our society many people are not really pursuingselfish behavior even if they do not seem to believe in any particular religions, morals,and ethics.

One of the reasons why we do not behave in a very selfish manner is that somepeople, with whom we do not have a long-run relationship, often criticize and attackour selfish behavior, and we do not want to get such an attack from others.16

16In Japanese local communities, the “punishment system based on the norm” was established andused to induce cooperative behavior. The “voluntary punishment” discussed here has also beenfrequently used. With the weakening traditional local communities, the society may have to rely onthe voluntary punishment of some people to discourage socially undesirable actions. However, thearguments here suggest that such voluntary punishment may be disappearing as rational individualsfind it costly to impose. As the last resort, the Japanese society may have to use the legal processmore. However, recently there seem to be some social attempts at applauding and admiring sociallydesirable actions such as volunteering and social contributions. If voluntary praise for desirableactions is more effective than punishment for undesirable ones, and if the cost of praising others islower than that of punishing others, society may succeed in expelling socially undesirable actions.Such a mechanism may be effective in the new communities as we discuss in the next section.


Although such blame and scolding seem to have been working as social punish-ment to discourage people from taking socially undesirable actions, the question ofwhy some people try to impose such punishments on defectors when they expect noreturn from it and sometimes get attacked by the defectors still remains unaddressed.

It is difficult to explain such actions as rational behavior of self-interested individ-uals. To explain such behavior and to deepen our understanding of our communities,we depart from standard game theory that assumes the rationality of self-interestedplayers and seek some answer in an evolutionary game developed by Sethi andSomanathan (1996).

Their argument is based on the tragedy of the commons but can be applied tovarious problems of externalities. We explain their models and results in relation tothe problems of the private provision of public goods.

For simplicity, we assume that the level of the public goods that each individualcan choose is either g∗∗ or g∗, and that g∗∗ is the socially desirable level. In additionto the decision about the public good, each individual is expected to decide whetherthey punish those who choose the selfish action g∗. Such a punishment requires acost γ > 0 per defector and they can impose a cost of δ > 0 on the defector.

Under these assumptions, individuals can be categorized into the following threetypes.

• Cooperator : provides g∗∗, but does not punish defectors.• Defector : provides g∗.• Enforcer : provides g∗∗ and punishes defectors.

We assume that there are n individuals and denote the share of each type in thetotal population by s1, s2 and s3 (s1 + s2 + s3 = 1). The total provision of publicgoods can be written as follows:

G = (1 − s2)ng∗∗ + s2ng∗

By defining V (G, g) ≡ u(G, w − g), the utility of each type can be expressed asfollows:

π1 = V (G, g∗∗)π2 = V (G, g∗) − s3nδ

π3 = π1 − s2nγ

The evolutionary process is assumed to be described by the following replicatordynamics:

si = (πi − π)si (i = 1, 2, 3) (6.1)

where si represents the speed of the change in si and π ≡ s1π1 + s2π2 + s3π3 rep-resents the average payoff of the society. The Eq. (6.1) indicates that the share si oftype-i increases, if its payoff πi is larger than the average π , at the speed according


to the difference (πi − π), and the share si decreases if the payoff of type-i is smallerthan the average, by the natural selection mechanism.

Now, we define the equilibrium of the evolutionary process as follows.

Definition 6.1 An equilibrium s = (s1, s2, s3) of the replicator dynamics is a stateat which s1 = s2 = s3 = 0 holds. We say that s is stable if, for any neighborhood Uof s there exists a neighborhood U1 ⊂ U such that any point in U1 will remain inU . We say that the stable s is asymptotically stable if there exists a neighborhood Usuch that any point in U converges to s.

First, by the definition of equilibrium, for any i with si > 0, we need to haveπi = π at equilibrium. There are only two types of equilibrium.

• D Equilibrium : Everyone is the defector (s1 = s3 = 0 implies π2 = π )• C-E Equilibrium : Everyone is the cooperator or the enforcer (s2 = 0 implies

π1 = π3 = π )

Now, we can present the following propositions about the stability of the twotypes of equilibria (See Sethi and Somanathan (1996, Proposition 2) for the proof).

Proposition 6.3 D equilibrium is asymptotically stable. If δn > V (G, g∗) − V (G,

g∗∗), then there exists a subset S ≡ {(s1, s2, s3) : 0 ≤ s1 < a, s2 = 0, s3 = 1 − s1}in the set of C-E equilibria such that any s ∈ S is stable. There are no other stableequilibria.

Proposition 6.3 can be explained by Fig. 6.3. In the figure, to explain each equilib-rium in the two-dimensional plane, we consider a set {(s1, s2) : s1 + s2 ≤ 1, s1, s2 ≥0} represented by the triangle 0AB. Any point (s1, s2) in the set represents a state(s1, s2, s3) in the evolutionary game, where s3 = 1 − s1 − s2. For example, each ver-tex of the triangle represents the state in which only one type exists. The point 0represents (s1, s2, s3) = (0, 0, 1), the point A represents (s1, s2, s3) = (1, 0, 0), andthe point B represents (s1, s2, s3) = (0, 1, 0). The set S defined in Proposition 6.3is a subset of set 0A in which s2 = 0, and thus can be depicted by the segment 0a inthe segment 0A in Fig. 6.3. The D Equilibrium is depicted by the point B because itis the state in which there is no cooperator or enforcer.

Fig. 6.3 Evolutionarydynamics and equilibrium.Source: Sethi andSomanathan (1996, Fig. 2)

0 s1

s2

1

1

B

A

aS


According to Proposition 6.3,S is a set of stable equilibria. The figure suggeststhat even if there is a departure from any point in the set S , the state converges tothe neighborhood of the point. The result implies that starting from a point in S ,where there is no defector, if some enforcers show up slightly as mutants, they willbe punished by a large number of enforcers as long as the state is near the point0, and as a result would disappear soon. As the enforcers need to bear the cost ofpunishment, their payoff is lower than that of the cooperators who do not punish themutants; thus, the number of enforcers will go down. This means that it does notgo back to the original state before the mutants show up, but goes back to the pointclose to the original state if the number of mutants is sufficiently small.

When the number of enforcers becomes small and the state lies closer to point Ain the set S , even if the number of defectors in the mutants is small, it takes timefor the punishment by the smaller number of enforcers to have effect. In the process,the number of the enforcers goes down and their punishments may not preventthe increase in defectors who can enjoy payoffs higher than average with smallerpunishments. If this happens, the state can never go back to the neighborhood ofthe original state, but converges to point B, where there are no enforcers. At pointB, even if there is a mutation, a small number of enforcers among the mutants canhave little power to defeat the defectors and must die out because of the costs ofmaking ineffective punishments. This indicates that point B, that is, D equilibrium,is asymptotically stable.

Remark 6.8 (Axelrod’s Games) Axelrod (1984) reports the following outcomes ofthe competition of computer programming for the repeated prisoner’s dilemma game.In the first completion, 14 participants from the fields of psychology, economics, pol-itics, mathematics, and sociology, submitted the programs. The prisoners’ dilemmagame was repeated 200 times by these programs in the tournament competition.Among various rules formulated in the programs on how to respond to the actionsof others, the one with the highest average payoff was the tit-for-tat strategy (c.f.Remark 3.10). The common characteristics observed in the rules that earned highaverage payoffs were “being nice,” in the sense that they never betrayed others fromthemselves.

After the first competition, the results and analyses were shared, and the secondcompetition was held. In the second competition, there were 62 participants from sixcountries and the repeated prisoners’ dilemma game was played in the tournament,where the number of repetitions was determined randomly so that each game wouldbe played as if the prisoners’ dilemma game is repeated infinitely.

Among various rules, the tit-for-tat strategy once again earned the highest averagepayoffs. In the second competition, the evolutionary game in which the programswith lower payoffs than the average decreases according to the replicator dynamics inthe tournament. The game is considered to reflect a realistic situation, in that, peoplechange their behavior from those with low payoffs to those with high payoffs. Thequestion is which rules perform better in the long-run in such a situation.

http://dx.doi.org/10.1007/978-4-431-55909-2_3


The strategy that earned the largest share was again the tit-for-tat strategy. “Non-nice” strategies that abused the exploitable strategies were successful and increasedtheir share initially, but they were weeded out as the good-natured strategies becameextinct.

From the analysis of the competition, Axelrod (1984, p.54) pointed out the fol-lowing four characteristics of successful rules: (1) by being nice, don’t get intounnecessary trouble, (2) by being retaliatory, discourage the other player from per-sisting whenever defection is attempted through retaliation, (3) by being forgiving,restore mutual cooperation, and (4) by being clear, make it intelligible to the otherplayer to elicit long-term cooperation. Notice that the tit-for-tat strategy embodiesall those characteristics. It was stressed that “being nice” and “being generous” aretwo important characteristics of the rules that survive in the long-run.

Remark 6.9 (Communities among Animals) Kropotkin (1902) clarified the impor-tance of mutual aid in the animal life by demonstrating many cases. His claim canbe clearly observed in the following statement.

[M]utual aid is as much a law of animal life as mutual struggle, but that, as a factor of evolu-tion, it most probably has a far greater importance, inasmuch as it favours the development ofsuch habits and characters as insure the maintenance and further development of the species,together with the greatest amount of welfare and enjoyment of life for the individual, withthe least waste of energy.

The mutual aid within families in the animals is well-known. The mutual aidbeyond the boundaries of families can be observed among many animals such asants, bees, cranes, elephants, andmonkeys. Kropotkin (1902) argues that suchmutualaid is effective in the harsh struggle for survival. As the discussion in this chapterindicates, it is not so easy to maintain such mutual aid.

There is a possibility that those who provide mutual aid can be weeded out by thefree riders, thus necessitating some punishment on the free riders. The quotation fromKropotkin (1902) in the beginning of this chapter suggests that there are mechanismsto prevent the occurrence of free riders in the animal communities that have surviveduntil today.

Concerning the cooperative behavior in animal communities, there are manyanalyses basedon evolutionary games. For example,Clutton-Brock andParker (1995,p. 214) point out that to induce cooperative behavior in some species, punishmentsfor non-cooperative behavior are imposed by the boss of the community. As they areusually given strong powers and higher chances of mating, there is a benefit of beingthe enforcer of cooperation in the community and a mechanism for such enforcersto have larger shares in the community. Even in animal communities, which havesuccessfully retained the norm of mutual aid, there seem to exist some mechanismsto punish defectors and reward enforcers.



In this chapter, to understand the functions and structure of traditional communities,we started with understanding the voluntary provision of public goods in communi-ties.We show that the private provisionof public goods results in an inefficiently smalllevel of public goods because the public goods produce externalities, thus making itbetter for people to free ride rather than provide them. This problem is an exampleof the “tragedy of the commons” discussed in Sect. 3.3.5. Traditional communities,if they survived, usually have some mechanism to mitigate the problem.

If the legal costs to solve the free rider problem are small, communities will beable to improve the efficiency in the private provision of public goods. However, thecosts are usually not small, thus making it difficult to overcome the problem throughthe legal system. Many communities have mitigated the problem through informalcontracts and punishments such as community rules and ostracisms to prevent the freerider problems. As the market system develops and the mobility of people increases,long-run relationships among members of the community, which are necessary forinformal punishments by communities to work, have been disappearing. Further, thecommunity has begun to face difficulties in maintaining mutual aid because informalpunishment has started having little impact in preventing the free rider problem.

An evolutionary game discussed in this chapter suggests that communities maybe able to sustain the optimal level of public goods if there are many people whofollow the social norms of punishing free riders. However, if free riders show uprepeatedly, the number of people who try to punish them will become smaller andcommunities are likely to end up with a high number of free riders.

In the process of declining voluntary provision of public goods (e.g., mutual aid)in communities, if the government starts providing the public goods, it may crowdout the public goods provided within the communities. The crowding-out theoremsuggests that the expansion of the government can weaken the functions of thecommunity further through crowding-out effects (c.f. Sect. 6.2.3).

It is true that many empirical studies indicate that the crowding-out effects workonly partially, and thus we cannot claim that the expansion or reduction of the gov-ernment have little effect on our society. However, the arguments in this chaptersuggest that it is important to recognize such interactions between the governmentand communities in designing public policies.

References

Abreu, D. (1988). On the theory of infinitely repeated games with discounting. Econometrica, 56,383–396.

Andreoni, J. (1988). Privately provided public-goods in a large economy-the limits of altruism.Journal of Public Economics, 83, 57–73.

Axelrod, R. (1984). The evolution of cooperation. New York: Basic Books.

http://dx.doi.org/10.1007/978-4-431-55909-2_3


Bergstrom, T., Blume, L., & Varian, H. R. (1986). On the private provision of public goods. Journalof Public Economics, 29, 25–49.

Bliss, C., & Nalebuff, B. (1984). Dragon-slaying and ballroom dancing: The private supply of apublic good. Journal of Public Economics, 25(1–2), 1–12.

Clutton-Brock, T.H.,&Parker, G.A. (1995). Punishment in animal societies.Nature, 373, 209–216.Khanna, J., & Sandler, T. (2000). Partners in giving: The crowding-in effects of UK Governmentgrants. European Economic Review, 44(8), 1543–1556.

Heutel, G. (2009). Crowding out and crowding in of private donations and government grants.NBER Working Papers 15004.

Kropotkin, P. (1902). Mutual aid: A factor of evolution. London: Heinemann.Manzoor, S., & Straub, J. (2005). The Robustness of Kingma’s crowd-out estimate: Evidence fromnew data on contributions to public radio. Public Choice, 123, 463–476.

Ostrom, E. (1990). Governing the commons: The evolution of Institutions for Collective Action.Cambridge: Cambridge University Press.

Payne, M. (2001). Measuring the effect of federal research funding on private donations at researchuniversities: Is federal research fundingmore than a substitute for private donations? InternationalTax and Public Finance, 8(5–6), 731–751.

Rege, M. (2004). Social norms and private provision of public goods. Journal of Public EconomicTheory, 6(1), 65–77.

Rose-Ackerman, S. (1986). Do government grants to charity reduce private donations? In S. Rose-Ackerman (Ed.), The Economics of Nonprofit Institutions: Studies in Structure and Policy (pp.313–329). Oxford: Oxford University Press.

Watanabe, Y., & Hojo, H. (1975). Forest Iriai and Village Structure. University of Tokyo Press. (inJapanese).

Sethi, R., & Somanathan, E. (1996). The evolution of social norms in common property resourceuse. The American Economic Review, 86(4), 766–788.

Warr, P. G. (1982). Pareto optimal redistribution and private charity. Journal of Public Economics,19, 131–138.

Warr, P. G. (1983). The private provision of a public good is independent of the distribution ofincome. Economics Letters, 13, 207–211.

Chapter 7New Communities

In democratic countries the science of association is the motherof science; the progress of all the rest depends upon the progressit has made.

Tocqueville, A. (1840; Volume II, Chap.V)

7.1 Introduction

Relationships amongmembers in traditional communities are usually definedvaguely.Hence, it is difficult to define formal rules to maintain cooperation and thus variousinformal punishments based on long-run relationships have been used to preventselfish behavior that harms the community. In traditional communities, in order toimpose social punishments on deviators of the norm of the community, people aremonitored constantly and occasional conflicts with the deviators are solved by thewhole community. In general, the relationships in communities are tight-knit andclosed. For those people who like to have “freedom,” such a relationship can be quiteoppressive.

The relationships in traditional communities have changed a lot in Japan aftereconomic development. As the mobility of people increases in a market economy,relationships in local communities have come to be short-termed, whichmakes socialpunishment less effective. Hence, it became difficult to maintain the level of mutualaid within traditional communities. We have indeed observed the decline in mutualaid, such as the decline in support for the elderly and children, and efforts to preventcrimes in the Japanese communities. Relationships in these communities have grownweaker and weaker, especially in urban areas where people often do not know muchabout their neighbors.

Under such circumstances, it has been expected of the Japanese government toplay roles that traditional communities have played. The Japanese government espe-cially expanded its role in social security. However, the public debt has accumulatedto the extent that public expenditure needs to be reduced. One of the strategies for


135

136 7 New Communities

increasing public goods and mutual aid in the ageing society is encouraging peoplewith relatively high (pure and/or impure) altruism to provide them.1

A French philosopher Tocqueville traveled across America and wrote about itssociety and people in his book Democracy in America. In the book, he mentionedthat associations are the basis of democratic society in America. Unlike traditionalcommunities, which naturally arose in history, associations are communities (i.e.,networks of people) connected by the strong wills to associate with others for certaincommon goals. We refer to such communities as “new communities” and try todeepen our understanding on them as they are highly expected to substitute the rolesplayed by traditional communities and governments.

NPOs can be viewed as a good example of a new community.2 NPOs are expectedto provide public goods that commercial companies, traditional communities, andgovernments fail to provide. In Sect. 7.2, we first analyze the NPOs.

In local communities, if the central and local governments shrink in size, the qual-ity of life will rapidly deteriorate because traditional communities are not workingwell anymore. Under such circumstances, there will be an increase in the private pro-vision of public goods, as suggested by the crowding-out theorem (Proposition6.2).One problem is that the boundaries of local communities in which the public goodsare privately provided are not clear any more because they were weakened by theexpansion of the markets and the government.

However, in the Japanese society, there remain many networks of people whoshare information and trust. These networks can be regarded as the communities.Although they often do not take the form of organizations like NPOs, each networkcan be viewed as an association, and thus as a new community that is tied by the will

1To induce the private provision of public goods, we have to collect the goodwill of many people;as such, we need people with leadership skills. Such a leadership is a kind of public good. Thediscussion in Remark 6.3 may be useful in finding someone who can provide such public goods.2Commercial companies, which seek profits, can be viewed as a community if we consider thenetwork of workers and managers in them. Unlike traditional communities, such a network may beviewed as an association of people bound by strong wills. However, in this book, by defining com-munities as networks of people who are not motivated to seek profits, we have excluded commercialcompanies from our list of communities. This is because with a clear goal of seeking profits, thenetwork of people is usually well defined and disciplined in legal contracts. They can avoid mostof the problems that “communities,” by our definition, face. However, the legal contracts are notcomplete, and the commercial companies sometimes encounter free rider problems and seek profitsas well as welfare of the workers and the society by providing some public goods. In particular,according to the traditional thinking of managers in Japanese companies, there is an idea that acompany should be managed as if it is a family, that is, a traditional community. Even in the periodof high economic growth, managers of Japanese companies have tried treating their workers asfamily members. It may be one of the reasons why the long-term employment and seniority systemhave been adopted by them. It is understandable to some extent because the system of contracts hasnot been well-regarded. It has rather been considered as a kind of guideline, partially because theJapanese society has not for a long time in its history, been a contract-based society, but a relation-based one. However, presently, in modern Japan, the system of contracts has been given greaterregard, which is rapidly changing the nature of Japanese companies. The analysis of Japanese com-panies may provide some insights into the function and structure of communities, which this bookis interested. However, such analyses are beyond the scope of this book.

http://dx.doi.org/10.1007/978-4-431-55909-2_6

http://dx.doi.org/10.1007/978-4-431-55909-2_6


to provide mutual aid among the members of the network. The network of peoplewho share norms and trust is now known as social capital. Recently, the existenceand functions of social capital have been well-analyzed and discussed in economics.In Sect. 7.3, we will show some economic analyses and discussion on social capitalas another example of new communities.

As the market economy expands and individuals are freed from traditional com-munities to have independent lives, people can obtain more freedom, but they willbecome “feeble,” as pointed out byTocqueville.3 Hence, people are expected to coop-erate with their neighbors to provide the public goods that traditional communitieshave provided.

It is a big challenge for the modern societies to find ways to overcome the freerider problems associated with the provision of the public goods and mutual aid.4 Asthe market system expands and enters every aspect of our life, good understandingon new communities will allow us to derive important implications in constructinga sustainable democratic civil society in Japan.

7.2 NPO

As an important example of new communities, we first consider the roles and func-tions of NPOs, which have received substantial attention in Japan, probably becauseof the current social conditions under which the Japanese government cannot expandits role owing to huge public debts.

7.2.1 Concepts of NPO

In Japan, the term NPO is considered to be the abbreviation of “Non-profit organiza-tion” and not “Not-for-profit organization” which seems to be a more popular termin the world. The two expressions are not quite different, but the former refers toa constraint that prevents an NPO from earning profit, while the latter is driven bythe goal or motivation to not seeking profits. Hence, we may say that “Not-for-profitorganization” can earn unexpected profits from its activities.

3“Among democratic nations, on the contrary, all citizens are independent and feeble; they cando hardly anything by themselves, and none of them can oblige his fellow men to lend him theirassistance. They all, therefore, become powerless if they do not learn voluntarily to help one another(Tocqueville 1840; Volume II, Chap.V)”.4 “Associations” in America, which we call new communities, seem to have been shrinking asPutnam (1995) argued. The fact suggests that, just like traditional communities, new communitiescan be easily weakening (c.f. Remark 7.3). New communities, which can be described as the socialcapital, especially fragile because boundaries of such communities cannot be clearly defined (c.f.analysis in Sect. 7.3.3).


In the analysis ofNPOs, the goal of the organization is important, but the constraintthat prevents them from earning any “profit” to be allocated to investors and donorsis also important. Hence, we use the termNPO to define an organization that does notseek and is constrained from making any profits. The no-profit constraint is usuallyclearly defined in the laws of NPOs.5

Now, according to this definition of anNPO, our society has a number of examples.For example, in Japan, almost all private schools, universities, and hospitals can becategorized as NPOs that fall under the no-profit constraint of Japanese laws.

Remark 7.1 (NPOs in Japan) The share of NPOs in the economy is not low inJapan. For example, Salamon et al. (2007) indicated that among eight countries,mostly OECD countries, in which NPOs contributed about 5% on average to theGDP, Japan was ranked third with the share 5.2%. Another estimate by the CabinetOffice of Japan (2009) suggests that the value added by NPOs ranges from 24 to 30trillion yen, which corresponds to a GDP-ratio between 4.8 and 6%, depending onthe definitions of the “value added” by NPOs. Out of the total value added by NPOs,about a half is generated by hospitals; and “social-welfare corporations,” “public-benefit corporations”,6 and incorporated educational institutions that operate privateschools and universities contribute about 16% each. The value-added by the so-called“NPO corporations,” which are set up and operated independently of the government,is about 0.5% of the GDP of Japan.

With regard to this composition of Japanese NPOs, their revenues mostly comefrom the sales of the services sector (57.4%) and the government (41.4%). The shareof “philanthropy,” such as donations and membership, constitutes only about 4.1%of the total revenues of NPOs in Japan. This share is very low as compared to othercountries. For example, according to the research by the Johns Hopkins University,the ratio of philanthropy is in the lowest group among 34 countries.7

These data indicate that Japanese NPOs are not supported by private philanthropy.Most of them are ones that work with or for the government, mainly in social securityand educational sectors, heavily subsidized and protected by the governments. Eventhe share of their value added to GDP has increased from about 3 to 5% in the fifteenyears from 1990 to 2005, it seems to be mainly reflecting the increase in socialsecurity services due to the population aging.

7.2.2 Economic Analyses of NPO

Traditional economics has assumed that the services are provided by producers whotry to maximize profits. However, there are many services provided by NPOs thathave been recognized and analyzed by some economists.

5For studies on NPOs in economics, see Glaeser (2003) and Anheier and Ben-Ner (2003).6These corporations have been set up and operated to support various government activities.7See the report (http://www.ccss.jhu.edu/pdfs/CNP/CNP_table401.pdf).

http://www.ccss.jhu.edu/pdfs/CNP/CNP_table401.pdf

7.2 NPO 139

In economics, a basic proposition is that production in amarket economy becomesefficient because producers are self-interested. Hence, it is natural, especially duringearly stages of economic analyses of NPOs, for many theoretical analyses to empha-size the inefficiency of service provision by NPOs (e.g., Newhouse 1970; Clarkson1972; Pauly and Redisch 1973). There are also many empirical studies that showsuch inefficiency in places such as schools, hospitals, museums, and other not-forprofit enterprises.

Thus, there are many arguments that try to justify the existence and activities ofthe NPOs. The basic perception underlying those arguments is a simple fact thatwith market failure problems the profit-maximizing behavior of producers createsinefficiency. With such problems, governments are expected to play an active role.However, the government causes government failure problems. The basic idea of thenew series of studies is that NPOs canmitigate market failures without getting caughtup in government failures (e.g., Weisbrod 1975; Hansmann 1980, 1981; Easley andO’hara 1983).

For example, it was discussed that NPOs can play a positive role in supplyingservices that tend to be inefficiently small with market failure problems such aspositive externality and incomplete information. The services that NPOs are activelyproducing, such as education and social security, are those in which quality andpositive externality of the services are important and tend to be under-supplied dueto incomplete information on the quality and the free rider problems. NPOs are theones that can mitigate these problems.

One of the difficulties of conducting economic analysis on suchNPOs is formulat-ing the goals of NPOs.We know that these goals do not include profit-maximization.The question is, what are their goals? Property rights of NPOs are not clear becauseinvestors usually have no claim on the properties under the no-profit constraint (c.f.Remark 7.2), and thus the goals are not clear for the analysis of for-profit organization.

This nature of NPOs makes its theoretical analysis a little ad hoc. Many studieshave focused on the decisions of managers who determine the actions of NPOs.It is often assumed that they have a utility function that depends on the rewardsaccruing to them, which can be decomposed into personal monetary rewards andsocial evaluations for their business decisions.

For example, Hansmann (1980) made the following argument. When the problemof incomplete information regarding the quality of the services exists, consumers willnot believe that the commercial companies seek quality improvement, which requiresadditional costs, and reduction in profits, something that investors do not want.

On the other hand, in the case of an NPO, the manager who is interested in socialevaluation can try to incur some costs on his decisions to improve service qualityfor higher evaluation from consumers, because of the no-profit constraint. As suchdecisions are rational, people can believe in them and expect the qualities of servicesprovided by NPOs to be higher than that of for-profit companies. Hence, NPOs canimprove the quality of services provided in markets and expand its services.

Note that the reason that NPOs can mitigate market failure problems and improvemarket efficiency is not that they have altruistic preferences, but that they are notconstrained by the principle of profit maximization.


We show this argument with a simple model developed by Glaeser and Shleifer(2001). The model allows us to consider the managers’ strategic choice betweenbeing an NPO or a for-profit organization. The analysis is interesting because theargument above suggests that NPOs can have some business advantages in marketsunder incomplete information. It also casts some important insights into understand-ing markets for schools and hospitals, in which both for-profit and not-for-profitorganizations coexist.

7.2.3 Basic Model

The basic model of Glaeser and Shleifer (2001) can be described as follows. Eachconsumer purchases one unit of service (e.g., daycare service)with qualityq andpriceP . The quality consumers prefer most is denoted by q∗. Themaximum price that theycan pay for the service with quality q∗ is given by the function P = z − m(q∗ − q),where z and m are parameters of the function.

As for the supply of the service, the cost for providing one unit with quality q isgiven by the cost function c(q) with properties c′ > 0 and c′′ > 0 for the first andsecondderivatives of the function. Themanager needs to pay somenon-monetary costb(q∗ − q) for providing the service with the quality q < q∗ because such a decisionwill not be welcomed by consumers. Such non-monetary costs may be explained bylower self-esteem or by expectation for the reduction in wages as managers of NPOs.

We can calculate the profit by Z = P − c(q), but under the no-profit constraint, Zwill take the form of some perquisite of the manager.8 We assume that the manager’sutility from such a perquisite Z is given by the function V (Z).

Now, if an entrepreneur chooses to be a for-profit company, the utility is givenby Z − b(q∗ − q). If one chooses to be an NPO, we assume that it is given bythe function U (Z , q) ≡ V (Z) − b(q∗ − q). In the following analysis, we assumefor simplicity that V (Z) = dZ with d < 1 because the profit Z is given by someperquisite whose value will be lower than the money given to the manager.

Given the definition Z ≡ P − c(q), it is easy to check that the optimal qualitythat maximizes the utility of the entrepreneur of for-profit companies is given by thefirst order condition c′(q) = b. On the other hand, the optimal condition for the NPOmanager is given by dc′(q) = b. Now, we have the following simple result.

Proposition 7.1 The quality of the service provided by the NPO is better than thatof the for-profit companies, that is, qn > q f .

Proof From the first order conditions, we have c′(qn) = bd > b = c′(q f ). As c′′ > 0,

the condition c′(qn) > c′(q f ) implies qn > q f .

8Examples of such perquisites may include houses, cars with drivers, parties, and overseas travels,which can be accounted for as necessary costs of the organization. In someNPOs, the perquisitemaytake the form of shorter working hours and longer vacations with respect to the high wages offeredto managers, which can be observed in many “public-benefit corporations” where governmentbureaucrats often work after their retirement.

7.2 NPO 141

The result qn > q f implies that the NPO can enjoy a higher price than the for-profit company. However, it does not mean that the entrepreneurs can get higherutility by choosing the form of NPO. It is only when the following condition holds.

d(z − m(q∗ − qn) − c(qn)) − b(q∗ − qn) > z − m(q∗ − q f ) − c(q f ) − b(q∗ − q f )

By rearranging the inequality above, we have the following proposition.

Proposition 7.2 There exists m∗ which satisfies the condition below, and the entre-preneur chooses to form a for-profit company if m > m∗ and an NPO if m < m∗.

m∗ ≡ (1 − d)z + b(q f − qn) − c(q f ) + dc(qn)

(1 − d)q∗ − q f + dqn

7.2.4 Management of NPO

If a higher quality of services can be expected from the NPOs, say, in the field ofeducation and social security, we may be able to expect that the qualities can be evenhigher if the donations are given to the NPOs. The following results support thishypothesis.

In the model, it is assumed that a donor donates D to the NPO and given thedonation D, the manager chooses a quality q and provide it at price P to maximizehis/her utility U (Z , q) ≡ V (Z) − b(q∗ − q) with Z ≡ P − c(q). In the followingpart, we use a general functional form V (Z) with the properties V ′ > 0 and V ′′ < 0,and not the special form V (Z) = dZ .

Suppose, the donor prefers higher quality, but can only change D. We assumethat the donor maximizes the utility (1 − τ)(y − D) + F(q) to choose D, where yis his/her income, τ is the income tax rate, and F(q) represents the donor’s satisfactionlevel from service q with properties F ′ > 0 and F ′′ < 0. To find the optimal decisionof the donor, we solve the problem backwards. Given D, themanager’s decisionmusthave the following property.

Proposition 7.3 The quality q of the service provided by the NPO gets better as thedonation D to the NPO goes up.

Proof Given D, the profit is calculated by Y ≡ D + P − c(q), and the managerchooses q to maximize U (Y, q) = V (Y ) − b(q∗ − q). The first order condition isgiven by

∂U (Y, q)

∂q= −c′(q)V ′(D + z − m(q∗ − q) − c(q)) + b = 0 (7.1)

The second order condition for profit maximization requires that the differentiationof the left-hand side of the Eq. (7.1) with respect to q becomes negative (c.f. Remark3.3), as follows:

http://dx.doi.org/10.1007/978-4-431-55909-2_3


∂2U (Y, q)

(∂q)2= −[c′′(q)V ′(Y ) + c′(q)V ′′(Y )(m − c′(q))] < 0 (7.2)

To observe the effects of the change in D on the quality, we differentiate the firstorder condition and rearrange it to get the following result.

∂q

∂D= −c′(q)V ′′(Y )

c′′(q)V ′(Y ) + c′(q)V ′′(Y )(m − c′(q))

From the assumptions on V and c, we have V ′′(Y ) < 0 and c′(q) > 0; thus, thenumerator becomes positive. On the other hand, the denominator becomes positiveby the second order condition (7.2). Hence, the increase in D increases the optimalquality q for the manger.

Now, given this behavior of the manger, the donor chooses the optimal level of thedonation. The first order condition to maximize the utility (1 − τ)(y − D) + F(q)

is given by

−(1 − τ) + F ′(q)∂q

∂D= 0

and the second order condition is given by

∂[F ′(q)

∂q∂D

]

∂D< 0

Using the results above, we can present the following result.

Proposition 7.4 The donation goes up when the income tax increases, and it goesdown by the same amount as the revenues of the NPO increases.

Proof By differentiating the first order condition to see the effects of the increase inthe tax rate τ , we have:

∂D

∂τ= −1

∂∂D

[F ′(q)

∂q∂D

] > 0

where the last inequality comes from the second order condition. To examine theeffects of the change in the NPO’s revenue z on the donation D, note that the optimalquality is dependent on the amount (D + z) because the effects of the change in Doccur only through the change in the “profit” Y ≡ D + z − m(q∗ − q) − c(q); thus,the increase in z will be entirely offset by the decrease in D by the same amount.9

9See the proof of Proposition 4 of Glaeser-Shleifer (2001) for more details.

7.2 NPO 143

The result above suggests that the donors with a higher tax rate will donate more,and the amount of the donation will be reduced by the same amount with an increasein, say, the public support for the NPO. The latter is the crowding-out effect of publicsubsidy on private donation to the NPO (c.f. Sect. 6.2.4). These results provide usefulinsights in designing public policies for improving social welfare by inducing theactivities of the new communities in the modern society.

Remark 7.2 (Governance of NPO) One of the characteristics of an NPO is thatits property rights are not clearly defined like that of for-profit companies. That is,there is no clear answer to the question “Who owns the NPO”? This nature has animportant meaning in considering the governance of NPOs.

In the case of stock companies as an example of a for-profit company, the structureof corporate governance is clear: The stockholders, as the owners of the company,delegate the decisions to the manager of the company to maximize its profit. As thisclear structure does not exist in NPOs, its governance can be weaker than that offor-profit companies, and this may create inefficiency. Contrary to this conjecture,Glaeser (2003) points out that such inefficiency cannot be generally observed in themanagement of typical NPOs in the U.S., such as hospitals, universities, museums,and churches. Further, he explains that these organizations are under harsh competi-tion with respect to providing services and getting donations and thus the managersneed to make efficient decisions to survive in the markets.

Gleaser (2003) also points out that the unexpected “profit” that the NPO generatesis distributed in the formofwages and perquisites to thosewho have themost preciousresources in the activities of the NPOs because they have high bargaining power inthe redistribution of the “profit” of the organization. For example, in hospitals anduniversities, it is often pointed out that medical doctors and professors have highpowers. It is probably because they have the most precious resources, in the sensethat the organizations will face serious problems in their activities if they lose thesespecialists. In such a circumstance, high bargaining powers will be given to them inthe organization. This perspective seems to have an interesting implication on thegovernance of the NPOs, which is important in considering and designing policiesfor NPOs to improve the quality of our society.

7.3 Social Capital

In social science, ties of people have received the attention of researchers. The conceptof “social capital” has been frequently used to describe and discuss the importanceof the ties of people. As social capital can be found in traditional communities, it isnot really a concept inherent to new communities. However, it is a very importantand useful concept for discussing the functions and role of new communities.

http://dx.doi.org/10.1007/978-4-431-55909-2_6


Fig. 7.1 Images of thesocial capital

(a) (b)

7.3.1 Concepts of Social Capital

The concept and definition of social capital have been discussed for a long time.10

In this book, we follow the definition given by OECD (2001)11:

Definition 7.1 Social capital is the network that facilitates cooperation within oramong others together with shared norms, values, and understanding.

What is important in this definition is that social capital is the network. Althoughthe word “network” is not well defined, it indicates ties of people. Needless to say,what is important is that the network of people be connectedwith and based on sharednorms, values, and understanding, allowing them to cooperate with each other.

In general, we can increase our utility through transactions with others. Hence,whether we can make transactions that raise our utility is very important. If indi-viduals have low transaction costs with other people, they will be able to achievehigh utility under the resource constraintswe face. The “network togetherwith sharednorms, values, and understanding that facilitate cooperation within or among others”will have high values because it will lower the transaction costs.

Figure 7.1 depicts two types of networks composed of six individuals. Using theterms in the graph theory, the network of people can be described by the “node’which represents the individual and by the “edge” which describes the relationshipbetween the individuals. If the “edge” represents the relationship with shared norms,values, and understanding, the network depicted in Fig. 7.1 indeed describes thesocial capital, and we can expect transactions between the two individuals connectedby the edge will be made easily.

10It has been pointed out that the use of “social capital” goes back at least to Hanifan (1916). It hasbeen intensively used and developed after the seminal works by Coleman (1988, 1990), Putnam etal. (1993) and Putnam (1995, 2000).11The definition is slightly different from the one given by Putnam et al. (1993, p. 167), who playedan important role in clarifying the importance and usefulness of the concept of social capital. Hedefined it as “features of social organization, such as trust, norms and networks, that can improvethe efficiency of society by facilitating coordinated actions (Putnam et al. 1993, p. 167)”. Strictlyspeaking, the definition is slightly different from the one given in OECD (2001), but their implica-tions are not very different. By defining social capital as a network with certain properties, we canquantify it to some extent, which is one of the reasons why we adopt the definition given by OECD(2001).

7.3 Social Capital 145

Figure7.1a represents a community with the richest social capital.12 If the peoplehave such close relationships with each other in the community, it must be easy forthem tomake transactions to raise their utility among them and achieve higher utility,which may not be possible otherwise.

On the other hand, the relationship of the people depicted in Fig. 7.1b is not closeand thus the level of social capital is low. In such a community, transactions willbe made among some individuals, but additional costs will be involved in furthertransactions. The social capital will give chances to raise utility by lowering thetransaction costs among the people whom we can trust under the shared norms,values, and understanding. Considering the facts that our life is full of transactionswith others, it is understandable that the level of the social capital must have effectson various economic and social variables.

Now, in the analysis of the social capital, we can analyze it from the “perspective ofthe community” and from the “perspective of individuals”. In the following part, wewill first look at the effects of social capital on the performance of the community, thatis, from the “perspective of the community” (Sect. 3.2). Based on various empiricalstudies, we can observe that social capital has positive effects on the performance ofcommunities. A natural question arising from such studies is how we can encouragethe accumulation of social capital.

To find answers to the question, we look at social capital from the “perspective ofthe individuals” as it is individuals who make decisions about investment in socialcapital. In Sect. 7.3.3, we present a simple theoretical model that clarifies the factorsaffecting individual decisions about investing in social capital. In Sect. 7.3.4, basedon various studies, we continue our investigation of factors that affect individualinvestment in social capital. The analyses will allow us to have some insight toconsider policies that induce the accumulation of social capital.

Remark 7.3 (Classification of Social Capital: Bonding Type and Bridging Type) Inthe discussion about social capital, realizing the differences between bonding-typeand the bridging-type is useful.13 For instance, Putnam (2000, pp. 22–23) explainsas follows. As the example of the former, Putnam (2000) picks up the bible-readinggroup or the same ethnic groups. In the case of Japan, the bonding-type correspondsto the social capital within traditional communities such as families and local com-munities in rural areas.14

On the other hand, as the examples of the bridging-type, groups to promote civilrights and the youth groups that provide social services are chosen. These are the

12Some concerns regarding the network as the “capital” have been expressed by some economists(e.g., Arrow 2000; Solow 2000). As the models below suggest, we can conduct an interestingeconomic analysis by regarding it as the capital. Although it is important to pay attention to thecriticism against such use, we believe that viewing the network of people as the capital of our societyis a useful way of understanding our society.13According to Halpern (2005, pp. 19–22), this classification was said to originate in Gittell andVidal (1998).14We may be able to associate such a bonding-type network with what Toennies (1887) called the“gemeinschaft.” See, for example, Brint (2001) concerning the classification of communities andrelated sociological studies on them.

http://dx.doi.org/10.1007/978-4-431-55909-2_3


communities connected by the strong wills, and thus correspond to what we callthe new communities or associations. The classification is considered to be usefulbecause the two types have quite distinct characteristics empirically. For example,the bridging-type social capital can depreciate more quickly than the bonding-type.

7.3.2 Effects of Social Capital

There are many empirical studies on the effects of social capital on society. In Japan,there is a gradual accumulation of such studies after the Cabinet Office of Japan(2003) estimated the social capital in Japan. For example, Fig. 1.8 about the negativerelationship between the crime rate and social capital is one such example. We willpresent, based on the reports by Halpern (2005) and Inaba (2007), some of theempirical studies on the effects of the social capital on the performances of oursociety and economy.

7.3.2.1 Economic Growth

Based on international comparisons, Knack and Keefer (1997) and Uslaner (2002)claim that social capital has a positive effect on economic growth. However, therelationship found in these studies is just a correlation and not a causal relationshipbetween social capital and the economic growth (c.f. Durlauf 2002). Furthermore,based on time-series data, some studies reveal that social capital and economic growthhave a negative correlation. In Japan, the fact that economic growth is high in urbanareas where social capital is low makes the relationship between social capital andeconomic growth negative (c.f. Inaba 2007, Chap.5).

Concerning the effects of social capital on economic growth, no clear relationshiphas been shown. However, the empirical studies based on international comparisonseem to suggest that certain level of social capital is necessary for economic devel-opment. When economic development reaches some point, the growth rate may beaffected more by other factors. Furthermore, the economic growth may have nega-tive impacts on social capital by weakening traditional communities. Such a causalrelationship between economic growth and social capital may be one of the reasonswhy we cannot have clear evidences about the effects of social capital on the econ-omy. More empirical studies using various factors that affect economic growth andan estimation considering endogeneity problems must be conducted.

7.3.2.2 Employment

There are many empirical studies indicating the importance of social capital onemployment and unemployment.15 In labor markets, there is incomplete information

15See Halpern (2005, p. 45).

http://dx.doi.org/10.1007/978-4-431-55909-2_1


about the opportunity for labor and the abilities of workers. To mitigate the marketfailure problem, the personal network plays an important role. For example, somestudies indicate that many of the workers in some companies were employed by therecommendation of existing workers. It is well-known that many immigrants canfind jobs through the network of immigrants.

In other words, the unemployment problem can be serious among those who havesmall social capital, that is, personal network. To mitigate the youth unemploymentproblem, that is, the difficulty faced by young people in getting full-time regularjobs, encouraging the young to accumulate social capital may be useful.

Furthermore, even in the market for managers, there is some evidence that match-ing is through informal personal networks. In particular, for a woman to get a jobat high ranks, whether she has some connection with the network of “the old boys”with high ranks is identified to be important (e.g., Saloner 1985).16 In Japan, Inaba(2007) found that the ratio of the elderly who can find jobs is higher in prefectureswith higher social capital.17

7.3.2.3 Inequality

There are many studies on the relationship between social capital and economicinequality. For example, Knack and Keefer (1997) reveal that the Gini coefficient,which measures income inequality, tend to be high in countries with low socialcapital. Such relationships can be explained by the fact that when social capitalis low the poor people tend to face more difficulty in improving their conditionsthan people with higher incomes, and thus the income inequality tend to be higher.However, we have to consider the possibility of the inverse relationship that whenincome inequality is high, it is difficult to build trust and share values among people,and thus social capital tends to remain low.

In the empirical studies so far, it is not clear as to what causal relationship toexpect. If both of the causality exist, a policy to pursue economic growth and leavethe problem of the inequality can lower the social capital and worsen the conditionsof the poor people further, which in turn can further lower the social capital. If thedecline in social capital has negative impacts on economic growth, the vicious cyclementioned above can have negative effects on economic growth. The argument maysuggest thatmitigating the problemof economic inequality is important for sustainingeconomic growth.

16In Japan, Brinton (2000) points out that schools and universities have their social capital, that isthe network of the alumni and people who help students find jobs. Existence of such a social capitalmay explain harsh competition to pass the entrance examinations for good schools and universitieswith rich social capital.17The study also indicates that the medical cost for the elderly is lower, per capita, in prefectureswith high ratio of the working elderly.


7.3.2.4 Education

In the early studies of social capital, its effects on education have been discussed.For example, Hanifan (1916) conceptualized the social capital to argue that thesuccess of education in schools depends on the relationship with local communities.Furthermore, Coleman (1988) also developed the concept of social capital through itsrelationship with the performance of education. For example, Coleman (1988) foundthat the drop-out rate fromhigh schools is dependent not only on the characteristics ofthe families, but also on the network among the parents and on whether the networkis of the bonding-type or the bridging-type (c.f. Remark 7.3).

In the recent times, there are many empirical studies indicating the positive effectsof social capital on the drop-out rate and on test scores.18 Such results can be observedin Japan. For example, Inaba (2007) reveals that social capital can have positiveeffects on the rates of the drop-out and bullying.

7.3.2.5 Health

There are many studies on the relationship between social capital and health condi-tions of people. Halpern (2005) points out that we can find the roots of such studiesin Durkeheim (1897), who suggested that suicide rates are strongly affected by thesocial environment that individuals belong to.19 Concerning dementia among theelderly, depression, and other general sickness from the flu to cancer, the disease rateis higher and the survival rate is lower among individuals who tend to be isolateddue to low social capital (Halpern 2005, Chap.3). A plausible reason to explain sucha relationship is that individuals with high social capital have less stress and canreduce the stress, which have positive effects on the disease rates and survival rates.

There are some empirical studies that analyze the relationship between socialcapital and health conditions in Japan. For example, Ichida et al. (2008) showed that,in the questionnaire on subjective perception about health, those who have highersocial capital tend to answer they feel healthy. As there remains a possibility thathealthy people can maintain and enrich the social capital, we cannot say much aboutthe causality. Nevertheless, the result indicates that there is a positive relationshipbetween social capital and health conditions.

7.3.2.6 Government’s Performance

The argument that social capital affects the government’s performance was put for-ward by Putnam et al. (1993) and developed in the field of comparative politics.Putnam et al. (1993) focused on the differences in the performance of the state gov-

18See Halpern (2005, Chap.5).19This statement corresponds to the following sentence byDurkeheim (1897, p. 209): “suicide variesinversely with the degree of integration of the social groups of which the individual forms a part.”


ernments in Italy, and found evidence that these differences can be explained by thedifferences in social capital in each state. Putnam (2000) also claimed that the differ-ence in the performances of the state governments in the United States is affected bythe social capital, and we can observe similar relation in the national governments.Concerning the reasons to explain a positive relationship, Boix and Posner (1998)pointed out that with the accumulation of social capital, consumers become polit-ically more sophisticated, the governments can form cooperative relationship withthe citizens, good policies are more likely to be implemented because the citizensmake better decisions, and cooperative democracy becomes possible.

7.3.2.7 Voluntary Provision of Public Goods

Concerning the effects of social capital on donations and volunteer activities, Apin-unmahakul andDevlin (2008) presents some interesting evidence. It is shown that thevoluntary provision of public goods is higher if bonding-type social capital is higherfor men, while it is higher if the bridging-type social capital is higher for women.It also indicates that religious relationship has strong effects on the voluntary pro-vision of public goods in both men and women. Apinunmahakul and Devlin (2008)argued that these results can be seen as evidence for a hypothesis stating that dona-tions and voluntary activities are encouraged because the value and opportunities ofthe voluntary provision of public goods are high when social capital is high. Theresults indicate the importance of accumulating social capital in encouraging peopleto make private provision of public goods in the forms of donations and voluntaryactivities.20

7.3.3 Accumulation of Social Capital: Theoretical Analysis

As various studies indicate, social capital has important effects on our economy andsociety. If so, questions arise as to howwe can accumulate it andwhether it is possibleto accumulate it by public policies. We have some interesting researches that give ussome answers to these questions.

We first present a simple theoretical model in which individual decisions aboutinvestment on social capital are analyzed in the frameworkof economics.The analysisis based on the perception that it is costly for individuals to form and maintainrelationships with others. If so, individuals choose the optimal investment on socialcapital by comparing the cost and benefit of the investment. Such an analysis is very

20Based on a questionnaire survey, Pargal et al. (2000) explained that the residents of some areasin the capital city of Bangladesh cooperate in improving garbage collection more than the residentsin other areas because of the higher social capital in the former. The results also suggest that thehigher social capital can be a factor that encourages people to provide public goods.


similar to the one for individual investment in human capital, but is slightly differentand is more interesting if we pay attention to the externalities of social capital.

The model below is an economic model developed by Glaeser et al. (2002).21

Let St be the level of social capital22 that each individual has in period-t . Let Sbe the total social capital per person. Given S, each individual can get R(S) for eachunit of individual social capital so that St R(S) is the total return from having theindividual capital St . We assume R′ > 0.

The individual social capital will be accumulated by the investment It and accu-mulation process and can be described as St+1 = δ St + It , where δ (0 < δ < 1) is theratio of the stock that remains after the depreciation of the individual social capital.To make an investment It individuals have to spend time as the cost C(It ), where weassume C ′(It ) > 0 and C ′′(It ) > 0.

Let w be the wage (opportunity cost of the time to accumulate the social capital),T be the length of time, β be the discount factor. If we let θ be the probability that theindividual leaves the community, λ be the ratio of the social capital that remains whenthe individual leaves the community, φ ≡ (1 − θ) + θ λ represents the expected ratioof the remaining social capital from the possibility of leaving the community.

In this model, the optimal investment problem can be given by the followingproblem in which the individual maximizes the discounted life-time utilities for Tperiods.

maxI0,I1,...,IT

T∑t=0

tβ[St R(St ) − wC(It )]

s.t. St+1 = δ φ St + It , t = 0, 1, 2, . . . , T

As each individual will choose the investment, given the total level S, the firstorder condition can be given by, for each period t = 0, 1, 2, . . . , T ,

wC ′(It ) = 1 − (β δ φ)T−t+1

1 − β δ φR(S)

Investment in social capital is low under a low discount factor (i.e., low weightson the future), high wage (i.e., a high opportunity cost of time), high depreciation rate(1 − δ) of social capital, high depreciation rate (1 − λ) from leaving the community,high probability θ of leaving the community, low total social capital S, low return

21The concept of social capital is often criticized because it is difficult to quantify accurately.From the perspective of scientific analysis, the difficulty of proposing testable hypotheses abouteffects of social capital, for example, can be seen as a critical defect of the concept. However, therehave been many attempts to quantify the social capital and test various hypotheses. For example,Knack and Keefer (1997) constructed an index of social capital based on various indexes for socialand economic conditions, and used it to test whether social capital has some effect on economicperformances. See Sect. 7.3.2.1 above for more discussion.22For example, we can think of St as the number of people with whom he/she shares norms, values,and understanding.


R(S) from it, and high age t . These results are easy to understand intuitively, andGlaeser et al. (2002) gives some empirical evidence for some of the theoreticalpredictions above.

Now,we consider the stationary state inwhich the individual social capital remainsconstant. We denote each variable without the subscript t . In the stationary state, bydefinition, we have I = (1 − δ)S and we can show that the effects of changes in thevariables above on the total social capital S is larger.

The reason is that if a change in a factor (e.g., an increase in wage) lowers indi-vidual social capital S, it will reduce the total social capital (S), which in turn furtherdecreases the individual social capital S because the return R(S) from the investmentis now lower. This vicious cycle of lower individual social capital and lower totalsocial capital makes the effects of each factor larger in the stationary state.

Hence, once social capital starts falling, it can decline further and further. Theresult is useful in understanding the rapid decline of social capital, which may beseen in the collapse of traditional communities and weaker human relationships inthe modern society.

To confirm these results in the model, we rewrite the optimal condition in thestationary state as follows:

wC ′((1 − δ)S) = 1

1 − β δ φR(S) (7.3)

Given R(S), the effect of change in the wage w on the total social capital in thestationary state S can be calculated by differentiating the Eq. (7.3) as

C ′((1 − δ)S)dw + [w(1 − δ)wC ′′((1 − δ)S)]dS = 1

1 − β δ φR′(S)dS (7.4)

and by rearranging it under the assumption dS = 0, we have

∂S

∂w= − C ′((1 − δ)S)

(1 − δ)wC ′′((1 − δ)S)(7.5)

On the other hand, the effect of higher wage of each homogeneous individual onthe total social capital S can be obtained by assuming dS = dS in Eq. (7.4) to get

∂ S

∂w= − C ′((1 − δ)S)

(1 − δ)wC ′′((1 − δ)S) − R′(S)/(1 − β δ φ)=

(1

1 − α

)∂S

∂w(7.6)

where α is defined23 as follows:

23α is equal to ∂S∂ S

in the Eq. (7.4). The multiplier effect is larger as α is close to 1, that is, as the

effect of S on S is larger.


α ≡ R′(S)/(1 − β δ φ)

(1 − δ)wC ′′((1 − δ)S)> 0

Under a stability condition, it can be shown that α < 1, and thus the effect of achange in the wage on S is larger than the one on S. As we have argued above, it isbecause of the chain reaction that the change in the total social capital can furtherchange the individual social capital, which in turn changes the total social capital.

Glaeser et al. (2002) call the value 11−α

> 1 the social multiplier. It also suggests,for example, that a small difference between the two regions can result in a fairly bigdifference in the social capital. From the perspective of public policy, the multiplierimplies that the social capital at the stationary state can be raised by a small changein the factor that affects the total social capital via the virtuous circle of individualand total social capital.

7.3.4 Accumulation of Social Capital: Empirical Studies

The theoretical model above gives us a bird’s eye view of how the social capitalaccumulates in our society. They also suggested some ideas about the factors affectingits accumulation. However, the model is too simple to consider all the factors thataffect the accumulation of social capital. Some of the factors have already beensuggested in the studies presented in Sect. 3.2. Here, we would like to present a fewmore empirical findings from Halpern (2005, Part II).

First, as regards the factors affecting individual social capital, some evidenceindicates that individual characteristics such as personality, family, education, job,social class, religion, and way of spending leisure time are important. Some peoplelike to get along with others and some people do not. It is well-known that familyis an important factor that affects the development of our personal characteristics.However, it is not the only factor. We spend lots of time in schools and workplaces,and learn how to build trustworthy relationships with others and in the process buildclose relationships. Hence, education and jobs are also important factors that affectindividual social capital accumulation.

It is important to recognize that the length of time in education and jobs is not theonly factor, but types of education and jobs also affect the accumulation of individualsocial capital. Furthermore, the social class and religion of each individual are knownto have someeffect on their relationshipswith others. It has also been stressed recentlythat howwe spendour leisure time is an important factor. For example, Putnam (2000)pointed out that the length of time for watching TV has a negative effect on individualsocial capital accumulation.

Next, as regards the factors affecting the total social capital accumulation in ourcommunity, the characteristics of schools and community, ethnicity and diversity inthe community, size and physical structure of the buildings and towns, and inequalitywithin the community are proven to have effects. Needless to say, culture, history,

http://dx.doi.org/10.1007/978-4-431-55909-2_3


economy, social structure, and social norms of the society as a whole also affect theaccumulation of social capital in the society and communities.

There are many interesting empirical studies. For example, it is claimed that bychanging the structure of the dormitory from independent room type to the onewith common space, the social capital increased among the students living in thedormitory. This suggests that the size and architectural structure of the communitycan change human relationships, and thus the social capital.

As we have seen in Sect. 7.3.2, there are studies on the negative effects of incomeinequality on the accumulation of social capital. Furthermore, there are many empir-ical studies which show that the mobility of people, expressed in terms of the timefor commutation and the ratio of home-ownership, have impacts on social capital incommunities, as we have suggested in the theoretical model.

Some of the factors indicated to have positive effects on the social capital cannotbe easily changed by the government. However, we can think of some policies tochange various factors affecting social capital. For example, the government willbe able to increase investment in the education for children and the architecturalstructure of public spaces fairly easily, and it may be able to have some effect onfamily relationships, the work environment, the use of leisure time, and the valuesystem of the people to encourage the accumulation of social capital.24


In the modern societies, ties of traditional communities have been weakened andthe relationships among the people have weakened. Under such circumstances, the“new communities,” whichmay be referred to as the “associations” in the democraticsocieties, are expected to play more roles in increasing the private provision of publicgoods which were provided by traditional communities. The governments that facegovernment failure problems and public debt accumulation have also come to payattention to the roles played by these new communities.

In this chapter, we picked up NPO as an example of the new community to havea better understanding on their characteristics and roles. It was indicated that theycan mitigate some of the market failure problems via the no-profit condition and canprovide some of the services that the for-profit companies cannot provide properly.New communities indeed seem to have some potential of improving the society.

NPOs are just an example of new communities that are tied by the strong willsof the people. In our daily lives, we can observe various informal communities con-nected by wills, even though most of them are not organized as NPOs. For example,we have a network of friends or neighbors with whom we can share norms and val-ues. Such a network can be viewed as an example of a new community and is now

24Based on a theoretical model, Apinunmahakul and Devlin (2008) suggest that an increase inthe public provision of public goods such as community centers, playgrounds, and libraries canencourage the accumulation of social capital and the private provision of public goods.


referred to as social capital. In the last half of this chapter, we tried to clarify theeffects of accumulation of social capital and the factors affecting it. The argumentsuggests that social capital and the new communities can have positive effects on oursociety and that there are some public policies that can encourage the accumulationof social capital, and thus the creation of new communities.

The concept of social capital does not exclude the network of people in traditionalcommunities. For example, the so-called “bonding-type social capital” requires closerelationship with what we call traditional communities.

However, recent discussions on social capital mainly focus on the creation of newcommunities where people form networks with other people according to their inter-ests and wills. Considering the importance of the discussion on how the weakeningand collapsing traditional communities can be transformed or integrated into newcommunities, the concept of social capital may be offering an ideal framework todiscuss and analyze traditional communities and new communities in our society.25

References

Anheier, H., & A. Ben-Ner (eds.) (2003). The study of the nonprofit enterprise: Theories andapproaches. New York: Springer.

Apinunmahakul, A., & Devlin, R. (2008). Social networks and private philanthropy. Journal ofPublic Economics, 92, 309–328.

Arrow, K. J. (2000). Observations on social capital. In F. Dasgupta & I. Serageldin (Eds.), Socialcapital: A multifaceted perspective (pp. 3–5). The World Bank.

Boix, C., & Poser, D. N. (1998). Social capital: Explaining its origins and effects and governmentperformance. British Journal of Political Science, 28, 686–693.

Brint, S. (2001). Gemeinschaft revisited: A critique and reconstruction of the community concept.Sociological Theory, 19(1), 1–23.

Brinton, M. (2000). Social capital in the Japanese youth labor market: Labor market policy, schools,and norms. Policy Sciences, 33(3–4), 289–306.

Cabinet Office of Japan. (2003). Social capital: Seeking for virtuous cycles of rich human relation-ship and civic activities (in Japanese).

Cabinet Office of Japan. (2009). Reports on the NPO satellite accounts. National Accounts Quar-terly, 138. (in Japanese).

Christakis, N., & Fowler, J. (2009). Connected: The surprising power of our social networks andhow they shape our lives. Boston: Little, Brown and Company.

Clarkson, K. (1972). Some implications of property rights in hospital management. Journal of Lawand Economics, 15, 363–384.

Coleman, J. S. (1988). Social capital in the creation of human capital.American Journal of SociologySupplement, 94, 95–120.

Coleman, J. S. (1990). Foundation of social theory. Cambridge: Harvard University Press.Durkeheim, E. (1897). Le suicide : Etudes de sociologie. Paris: Presses Universitaires de France.

25In the discussion of communities in modern society, we cannot ignore the so-called cyber-communities and social networks. The analyses on such communities (e.g., Christakis and Fowler2009), are increasing. The scale, scope, depth, and rate of growth of such communities are incom-parable to that of any other type of community. They must have a deep impact on our society. Weneed not deepen our analysis and discussion on these communities now, but we leave them forfuture studies.

References 155

Durlauf, S. N. (2002). On the empirics of social capital. Economic Journal, 112, F437–485.Easley, D., & O’hara, M. (1983). The economic role of nonprofit firms. Bell Journal of Economics,14, 531–538.

Gittell, R., & Vidal, A. (1998). Community organizing: Building social capital as a developmentstrategy. Thousand Oaks.

Glaeser, E. L. (2003). Introduction. In E. L. Glaeser (Ed.), The governance of not-for-profit orga-nizations (pp. 1–43). University of Chicago Press.

Glaeser, E. L., Laibson, D., & Sacerdote, B. (2002). The economic approach to social capital.Economic Journal, 112, F437–458.

Glaeser, E. L., & Shleifer, A. (2001). Not-for-profit entrepreneurs. Journal of Public Economics,81, 99–115.

Halpern, D. (2005). Social capital. Polity Press.Hanifan, L. (1916). New possibilities in education. In E. Ostrom & T. Ahn (Eds.), Foundations ofsocial capital (pp. 22–35). Edward Elgar.

Hansmann, H. (1980). The role of non-profit enterprise. The Yale Law Journal, 89, 835–898.Hansmann, H. (1981). Nonprofit enterprise in the performing arts. Bell Journal of Economics, 12,341–361.

Ichida, Y., Hirai, H., & Kondo K. (2008). Health and social capital. In Y. Inaba (Ed.), Capacity ofsocial capital (Chap. 7) Nippon Hyoron Sha (in Japanese).

Inaba, Y. (2007). Social capital: Challenges in modern economy and society solvable by Ties ofTrust Seisansei Shuppan (in Japanese).

Knack, S., & Keefer, P. (1997). Does social capital have an economic payoff? Cross-country inves-tigations. Quarterly Journal of Economics, 112(4), 1251–1288.

Newhouse, J. (1970). Toward a theory of nonprofit institutions: An economic model of a hospital.American Economic Review, 60, 64–73.

OECD (2001). The well-being of nations: The role of human and social capital. Paris: OECDPublication.

Pargal, S., Huq, M., & Gilligan, D. O. (2000). Private provision of a public good: Social capitaland solid waste management in Dhaka, Bangladesh. World Bank Policy ResearchWorking PaperNo. 2422.

Pauly, M., & Redisch, M. (1973). The not-for-profit hospital as a physicians’ cooperative. AmericanEconomic Review, 49, 253–283.

Putnam, R.D., with Leonardi, R., &Nanetti, R. Y. (1993).Making democracy work: Civic traditionsin the modern Italy. Princeton: Princeton University Press.

Putnam, R. D. (1995). Bowling alone: America’s declining social capital. Journal of Democracy,6(1), 65–78.

Putnam, R. D. (2000). Bowling alone: The collapse and revival of American community. NewYork:Simon & Schuster.

Salamon, L.M.,M. A. Haddock, S.W. Sokolowski, &H. S. Tice (2007)Measuring civil society andvolunteering: Initial findings from implementation of the UN handbook on nonprofit institutions.Working Paper No. 23 (Baltimore: Johns Hopkins Center for Civil Society Studies, 2007). http://www.ccss.jhu.edu/pdfs/ILO/Measuring_Civil_Society.pdf.

Saloner, G. (1985). Old boy networks as screening mechanisms. Journal of Labor Economics, 3,255–267.

Solow, R. M. (2000). Notes on social capital and economic performance. In F. Dasgupta & I.Serageldin (Eds.), Social capital: A multifaceted perspective (pp. 6–10). The World Bank.

Tocqueville, A. (1840). Democracy in America. In P. Bradley (Ed.). Alfred A. Knopf.Toennies, F. (1887). Gemeinschaft und Gesellschaft. Leipzig: Fues.Uslaner, E. M. (2002). The moral foundations of trust. Cambridge: Cambridge University Press.Weisbrod, B. (1975). Toward a theory of voluntary nonprofit sector in a three-sector economy. InE. Phelps (Ed.), Altruism morality and economic theory. New York: Russell Sage Foundation.

http://www.ccss.jhu.edu/pdfs/ILO/Measuring_Civil_Society.pdf

http://www.ccss.jhu.edu/pdfs/ILO/Measuring_Civil_Society.pdf

Part IIISocial Policies in Japan

Chapter 8Population Crisis

The population crisis is only the external aspect of what isreally a crisis in the family as an institution.

Myrdal, A. (1941, p. 4)

8.1 Introduction

The Japanese population started to decline in 2008, and it is forecasted to shrinkfurther at an increasingly rapid rate (c.f. Fig. 1.1). Moreover, the number of elderlypeople (above 65years old) continues to increase. The elderly ratio, that is, the ratioof the elderly people to the total population, will peak at around 38% in 2050 beforeleveling out.

If this demographic trend continues, the population is estimated to be half of whatit was at its peak by 2094, eventually falling to zero around 3300.1 This populationcrisis was basically caused by its very low total fertility rate (TFR). After 1974, theTFR dropped below the population replacement level (approximately 2.07) and hasnever come close to this ratio since then (c.f. Fig. 1.4). As of 2016, it is 1.45.

What could be the reasons for such low fertility rates and the resulting decline inpopulation? In a book written in 1941, Alva Myrdal pointed out that “the populationcrisis is simply a reflection of the crisis in the family as an institution.” Thus, thequestion is, what caused the crisis in the family as an institution? Our answer to thisquestion in Chap.3 was that the expansion of the market economy and social securityreduced individuals’ incentives to have children, thus changing the nature of familiesas an institution.

1The estimate of Japanese population in this book is the one reported by NIPSSR in 2017 exceptfor the estimate for 3300 which is the one in Coulmas et al. (2008).


159

http://dx.doi.org/10.1007/978-4-431-55909-2_1

http://dx.doi.org/10.1007/978-4-431-55909-2_1

http://dx.doi.org/10.1007/978-4-431-55909-2_3

160 8 Population Crisis

In the next section, we describe the problems of declining fertility in Japan andreview basic theories to understand the phenomenon. We argue that the decliningfertility in Japan is consistent with the theories and it can be viewed as a moralhazard problem (c.f. Remark 3.7) or a free rider problem under the generous pay-as-you-go public pension system, which is basically a redistributive system frompeople with children who support their parents to those without them.2 In Sect. 8.3,we discuss social policies for mitigating the problem. We discuss how the Japanesesocial policies have tried to mitigate the problem and explain how they should havebeen designed. Further, we argue that the transformation of social policies fromconsumptive remedy policies to productive empowerment polices is critical for thesustainability of the Japanese society. Section8.4 concludes the chapter.

8.2 Facts and Theories

In this section, we first look at the basic data on the population crisis. Then, we reviewtheories on family formation and the effects of social security on fertility, and arguethat the Japanese experience is consistent with these theories.

8.2.1 Population Crisis in Japan

The population crisis relates to the problem of rapidly declining population withoutany signs of stopping. Figure1.5 depicts the fall in the rate of marriage, which startedin the early 1970s. It is one of the biggest reasons for the decline in fertility. Thefigure also depicts the gradual increase in the rate of divorce. It is estimated that onethird of the marriages will eventually end in a divorce (Raymo et al. 2004).

A change in population can be decomposed into natural and social changes. Nat-ural change is defined by the number of births less the number of deaths. Socialchange is defined by the number of immigrants less the number of emigrants. Cur-rently, the number of deaths exceeds the number of live births in Japan; hence, thenatural change is negative. If it is compensated by social change, the populationmay not decline. However, the inflow of immigrants into Japan has been very small.Figure8.1 indicates that the total number of foreign workers in Japan is very low ascompared to other developed countries.

2Notice that if everyone has children who support their parents, we do not need the pay-as-you-gopension system.

http://dx.doi.org/10.1007/978-4-431-55909-2_3

http://dx.doi.org/10.1007/978-4-431-55909-2_1

8.2 Facts and Theories 161

Fig. 8.1 Ratio of foreign and foreign-born population. Source: United Nations: Migrant Stock2013

8.2.2 Theories for Understanding the Decline in Fertility

8.2.2.1 Family Formation

The economics of family has been gradually developed following the works ofBecker, who received the Nobel prize in 1992. In the existing literature, various phe-nomena surrounding families such as marriage, child-bearing, mutual assistance,inheritance, and divorces have been analyzed and explained through the rationalbehavior of individuals under resource constraints (c.f. Chaps. 4 and 5).

Needless to say, we may not have any children even if we want or we may havemore children than we want. Our inability to control the number of children may bethe reason why we think that our life is mysterious and economic theories cannotbe applied to understand our family lives. Although we often feel that our decisionson family formation are instinctive rather than rational, we also realize throughintrospection that we often make some calculation in making decisions on marriageand the number of children to have. Economics of family tries to explain familyformation from this perspective.

To understand family formation, we need to enquire about themotivation tomarryand bear children. An easy answer may be “for the continuation of our lineage.”However, this is not enough. The question is why we want to have descendants.Figure8.2 depicts representative answers of the Japanese people about having chil-dren, as expressed in the survey of the government in 1972 and 2002.

http://dx.doi.org/10.1007/978-4-431-55909-2_4

http://dx.doi.org/10.1007/978-4-431-55909-2_5


Fig. 8.2 “Opinions about having children (1972)” and “Reasons to have children (2002)”. Source:Data published by NIPSSR

Forming a family means that each individual engages in formal or informal con-tracts to live with someone as a family. There are benefits and costs of engaging insuch contracts. What are the benefits of marrying and having children?

Marriage typically starts with love, which involves the couple’s desire to livetogether. In this case, marriage brings happiness to the individuals just like the con-sumptions of goods and services. After the couples start living together, they canhelp each other to improve the efficiency of their daily lives and prepare for variousrisks such as accidents, illness, and injuries. Hence, marriage brings various benefitsin the long run just like investments.

Likewise, children are a source of happiness and also bring various benefits inthe long run. In particular, children are expected to insure their parents by providingincome and support when they are unable to earn themselves. Figure8.2 indicatesthat in Japan, children are expected to play such a role. As such, in the economicsof family, it is identified that there are mainly two purposes for having children:consumption and investment.

The considerations above suggest that the benefits of family formation are mainlyderived from the fact that families can provide goods and services thatmarkets cannotat adequate prices. For example, family members can provide home services (e.g.,cooking and cleaning), nursery services for children and elderly parents, finances,and insurances. It is well-known that these services are not provided adequately bythe markets mainly because of market failure problems. In particular, markets formany goods and services are missing due to the incomplete information problem(c.f. Sect. 2.2.1). Families, by sharing information, have played important roles inproviding services that markets have failed to provide.

The costs of marriage may not be too big. Bearing some obligations to be goodhusbands and wives and losing some freedom in the marital life may be one of thebiggest costs. The opportunity cost, that is, losing the possibility of getting marriedto someone else, may be seen as a cost of marrying someone, but not a cost of themarriage itself.

On the other hand, the costs of having children are quite high. Raising a child isnot an easy job. They include not only the monetary costs of buying foods, clothes,

http://dx.doi.org/10.1007/978-4-431-55909-2_2


Fig. 8.3 Public and private expenditure for tertiary education institutions in OECD countries.Source: OECD, Education at a Glance 2016, Fig.B2.2

and education for children, but also the cost of time to raise children. In Japan, justlike in other developed countries, women have higher levels of education and work athigher wages now. As the cost of time to raise children is equal to the wage incomethat one must give up to raise them, it is increasing because women’s wages areincreasing. If the benefits exceed the costs, people will choose to form a family. Thenumber of children will be determined by balancing the costs and benefits of havingchildren (c.f. Sects. 4.2.2 and 4.3.2).

It has been pointed out that the costs of getting married and having children ison the rise. First, the cost of education in Japan is quite high as the government’ssubsidies on education is low. This makes private expenditure on education reallyhigh (c.f. Fig. 8.3).

Japanese families also incur costs on “cram school (juku)” for their children topass the entrance exams to schools and universities. Furthermore, as Hirao (2007)stresses, mothers are expected to bear large costs of raising and educating theirchildren, in addition to the opportunity costs of giving up their wage income. Allthese costs discourage people, especially women, from getting married and havingchildren.

In the old days when the markets were underdeveloped, benefits from familyformation were very large. In fact, it was almost indispensable for most of the peoplebecause themarkets often failed to provide essential services. As themarket economydeveloped, many people started earning higher wage income and were able to savemoney for their lives after retirement and purchase insurances for various risks.Hence, the benefits of marriage and having children are getting smaller. If the return

http://dx.doi.org/10.1007/978-4-431-55909-2_4

http://dx.doi.org/10.1007/978-4-431-55909-2_4


on financial assets exceeds that from investing in children, people will not havechildren for investment purpose3 (c.f. Sect. 5.3.2).

The development of the market economy has reduced the net benefit of familyformation, thus contributing to lower marriage and fertility rates. Furthermore, whenthe government starts providing social security, most people need not rely on childrento prepare for risks in their lives (c.f. Sect. 4.3.2, and especially footnote 22 therein).The net benefits of marriage and having children are further reduced. Hence, thedevelopment of social security must have contributed to lower marriage and fertilityrates (c.f. Remark 8.1 for the empirical studies on this claim).

8.2.2.2 Social Security System and Fertility

The negative effects of the development of social security on fertility can be largerwhen the social security system is of the pay-as-you-go type (c.f. Sect. 5.3.3). Underthe pay-as-you-go pension system, young generations as a whole will contributetoward the pension, which is paid to the elderly people. For the system to work, eachindividual is expected to have children who will contribute to the pension system.

However, under the system people can receive pension payment even if they donot have any children. That is, people can free ride on children of other people tomake a living after retirement.4 This makes society inefficient and unstable. Thesocial security system will not be sustainable without people having children.

The problem can be compared to the “tragedy of the commons” (Sect. 3.3.5). Inthe case of the commons (e.g., open fields, forests, sea, lakes), which can be usedby everybody, people tend to exploit the resources till they run out. To mitigate thetragedy of the commons, society must discourage people to exploit the commonresources and encourage them to foster them instead. However, people do not makesuch efforts without adequate incentives.

The pay-as-you-go social security system faces the same problem. Elderly peo-ple receive social security payments from the commons, that is, the pension fundthat the young generations contribute toward. It can be viewed as the commonsbecause elderly people can get the pension payment from it even if they have notcontributed toward raising the young generations.5 In such a case, people will notmake efforts to raise children, which require high costs, and thus, the tragedy ofcommons ensues. Just as the forest disappears unless the new trees are planted, thepay-as-you-go pension will collapse unless people have children to make the pensionsystem sustainable.

3Poor people tend to have more children because they cannot earn enough money to save for theirlives after retirement, and thus, need to rely on their children.4This means that, under the pay-as-you-go system, raising children generates positive externalitiesin the society as the people enjoy its benefits without having to raise their own children.5Needless to say, everyone must have paid pension contribution to get the pension payment. How-ever, the pension system needs young generation for its sustainability. Hence, those who have nochildren are making insufficient contribution to the pension system and free-riding on children ofothers.

http://dx.doi.org/10.1007/978-4-431-55909-2_5

http://dx.doi.org/10.1007/978-4-431-55909-2_4

http://dx.doi.org/10.1007/978-4-431-55909-2_5

http://dx.doi.org/10.1007/978-4-431-55909-2_3


Remark 8.1 (Empirical Studies on Social Security System and Fertility) There aresome empirical studies that explicitly try to find out the relationship between thesocial security system and fertility. For example, Ehrlich and Kim (2007a, b) usedpanel data of 57 countries for 32years to see whether the pay-as-you-go defined-contribution pension system has any negative effects on marriage and fertility. Theyfound that, among the OECD countries, the size of the contribution for the pay-as-you-go pension had negative impacts on both marriage and fertility. Based on theestimation, they ran simulations of policies on marriage and fertility.

Based on panel data of 11 countries from 1750 to 1995, Puhakka and Viren (2006)demonstrated that there is a clear negative relationship between fertility and the sizeof the government, which was used as a proxy for the size of the social securitysystem. They also showed an interesting result that longer life expectancy can havepositive effects on fertility. Such a relationship can be theoretically explained by thefact that the longer we need to depend on our children, the more value will be placedon children, and thus, the greater the demand for having children.

The pay-as-you-go system can be justified to mitigate market failure problemsin the insurance markets (c.f. Sect. 2.2.1). However, it produces the side effect ofreducing fertility, which can endanger the sustainability of the system. This can bemitigated by government support for families in raising children. The necessity ofsuch a supplementary policy is well-identified in economics and it has been analyzedthrough theoretical models. One of the most important contributions, Groezen et al.(2003), claimed that it is desirable, from the perspective of efficiency, to providechild allowance as a subsidy for the positive externality of raising children under thepay-as-you-go pension system (c.f. Remark5.4).

One of its theoretical consequences is that if the government increases the socialexpenditure on the elderly via the pay-as-you-go pension system, it can lead to inef-ficient decline in fertility, unless the social expenditure on families raising childrenincreases. This theoretical relationship between social expenditure and fertility canbe observed in Fig. 8.4.

It depicts the relationship between the total fertility rate and the ratio of socialexpenditure on families to social expenditure on elderly people. The positive corre-lation observed in the figure indicates that when the social expenditure on families islow relative to that on elderly people, the fertility rate is likely to be low as well. Thisis consistent with the theoretical result of Groezen et al. (2003). The relative ratiois important because, given the social expenditure on families, higher social expen-diture on the elderly will lead to lower fertility rates as the benefits from havingchildren decreases. Figure8.4 indicates the existence of such a relationship.

Remark 8.2 (Fertility and Economic Growth) The issues of declining fertility andpopulation aging are dynamic and should be analyzed through dynamic models. Ineconomics, unlike the classical growth models with exogenously fixed populationgrowth rates, a series of dynamic models were developed to explicitly consider thedecisions for child-bearing (e.g., Becker and Barro 1988; Barro and Becker 1989).These models produced various interesting results.

http://dx.doi.org/10.1007/978-4-431-55909-2_2

http://dx.doi.org/10.1007/978-4-431-55909-2_5


Fig. 8.4 Relationship between the total fertility rate and the social expenditures. Sources: Dataon total fertility rates of 2004 in the United Nations, Demographic Yearbooks. The data on socialexpenditures are from OECD Social Expenditure Database 2004. The data on the population ofeach country in 2004 (except France and Greece, for which the 2003 data is used) is from theUnited Nations, Demographic Yearbooks. Notes: To avoid a problem in which social expenditureon families is lesswhen fertility is low,we divide the social expenditure on families by the populationof individuals below 15years of age. Similarly, we divide the social expenditure on elderly peopleby the population of individuals above 65years of age

To examine the impact of fertility on economic growth with respect to its effectson labor supply and technological progress, it is natural to study dynamic modelsin which these variables are endogenously determined. As regards analyses of thesocial security system, many of them are now based on dynamic growth modelswith endogenous fertility rates such that the population growth rates are determinedendogenously (e.g., Groezen et al. 2003). When the models become too complex tobe solved analytically, the calibrations are usually executed to simulate the effectsof policy changes (e.g., Imrohoroglu et al. 1995; Oguro et al. 2011). The modelswith endogenous fertility rates are inevitably more complex. But they are expectedto generate interesting and important results, especially, as most advanced countriesare now suffering from the problems of declining fertility.

8.3 Social Security System and Family Policies in Japan

In this section, we explain the social security system and family policies in Japan.We will see that Japan’s social security system is generous compared with otherdeveloped countries, while its family policies are poor.

8.3 Social Security System and Family Policies in Japan 167

8.3.1 Social Security System

In Japan, the social security system consists of four major components. They arepension, healthcare, long-term care, and public assistance.6 The first three compo-nents are operated as social insurances financed by social security contributions andtax revenue. The last component is fully financed by taxes.7

In this chapter, we will explain the three social insurances as the factors thatchanged the nature of families. Public assistance programs will be explained in thenext chapter where we discuss issues of poverty in Japan.

8.3.1.1 Public Pension

In 1961, the Japanese government introduced the universal public pension and everyJapanese citizen started to participate in them. It has contributed to the reductionof poverty among elderly people with no children who support them. The universalpension system has evolved gradually since then. The current system is as depictedin Fig. 8.5. As indicated by the figure, people are grouped into three categories.People in Category-1 are self-employed, those in Category-2 are employed by privatecompanies or the government, and those inCategory-3 are spouseswho are dependenton Category-2 workers.

The Japanese public pension system is basically a three-tier system.8 The firsttier is the basic pension for everyone. Its contribution rate is fixed and the benefitis dependent on the periods of the contribution. The second tier is the Employees’Pension Insurancewhere contributions and benefits are proportional to eachworker’swage income. The burden of the contribution is shared equally between the employeeand employer. It is only available for Category-2 people. Participation in the first andsecond tier is mandatory for everyone who is applicable. Although the participationin the third tier is voluntary, it is considered to be a part of the public pension asit has some special tax measures for contributions and benefits. It is available forCategory-1 and Category-2 people.

The people in Category-3, who are mostly housewives, do not have to pay anycontribution to receive pension benefits. They can free ride on the contributions ofthe workers in Category-2 because their spouses do not pay extra contributions forthem. The Japanese government has provided a big incentive for women to stay at

6See NIPSSR (2014) for an overview of the social security system of Japan.7About a third of the social security payment, including the public assistance, is financed by the taxrevenues and public deficit.8Recently, Japanese government started to explain the public pension system is a two-tier system andthe third tier is provided as private pension. However, pensions in the third tier are heavily regulatedby the government and rich tax measures are given to them. Hence, we describe the Japanese publicpension system as the three-tier system. The government has been encouraging people to participatein pensions in the third tier to prepare for the reductions in the “public pension.” Category-3 peopleare now able to participate in a defined-contribution pension from 2017.


Fig. 8.5 Public pension system in Japan. Notes: Number in parenthesis represents the number ofpeople (million) in each category in 2014. Source: Ministry of Health, Labor and Welfare

home (c.f. Sect. 11.2 in Epilogue). The big benefit to be a housewife of an employedperson is considered a reason for low female labor participation in Japan (Fig. 2.3).

8.3.1.2 Healthcare

It is also in 1961 that the Japanese government introduced the universal public health-care system. Since then, the systemhas evolved. The healthcare systemcan be dividedinto two parts. One is the health insurance system and the other is the delivery sys-tem of healthcare services. In Japan, universal healthcare means that every Japanesecitizen participates in one of the public health insurances. Hospitals, which providehealthcare services, are mainly private, but heavily regulated by the government.Private hospitals must be non-profit and the fees for medical services are determinedby the national government and not by the hospitals. Patients basically pay 30% ofthe medical costs and the rest is covered by health insurance.9

Public health insurance is divided into three types. Employed workers and theirdependent families belong to the health insurances for employees.10 Elderly peopleabove 75years of age join a special health insurance operated by the prefecture-level local governments. Others participate in the national health insurances operated

9The copayment rate is slightly adjusted depending on age and income.10Big companies can have their own health insurances for their employees and their families.Workers in small and medium sized companies participate in a publicly organized health insurancefor them. Workers in the public sector can participate in the health insurances for them.

http://dx.doi.org/10.1007/978-4-431-55909-2_11

http://dx.doi.org/10.1007/978-4-431-55909-2_2


by the municipal governments.11 As employed workers after retirement belong tothe national health insurances, the average age of the people in the national healthinsurances ismuch higher than the one in the health insurances for employees. Hence,some tax revenues and transfers form the health insurances for employees are givento the national health insurances.

8.3.1.3 Long-Term Care

The Japanese government introduced the public long-term care insurances in 2000.Care providers can be for-profit companies, but there are many regulations of thenational government regarding fees for care services and so on.

People above 40years of age need to participate in long-term care insurances,and they can receive the benefits of the insurance if they are deemed needful oflong-term care. To receive the benefits, based on opinions of specialists includingmedical doctors, the degree of need for care is identified. Depending on the degree,the insurance coverage differs. The copayment rate for long-term care services isbasically 10%.12

Before 2000, people who needed long-term care often stayed in hospitals eventhough they did not need medical services. As costs of hospital services involvingdoctors and nurses are usually high, the government decided to introduce a newinsurance for their long-term care in 2000, and increased the number of care-serviceproviders. It was a wise political decision to save healthcare costs, but some familieswith elderly parents who need long-term care, are now expected to take good careof them as they are now at home and not in the hospitals.

Since the introduction of long-term care insurance, the Japanese government hasbeen expanding the care services for the frail elderly at home, without significantlyincreasing the number of public nursery homes for them. Although such a policymay be good for some families that have elderly people staying at home, it has notrelieved the burden of families with elderly people who need intensive care (e.g.,the elderly with dementia). Figure8.6 indicates that the share of recipients receivinglong-term care at home was high in 2000 in Japan, as compared to other OECDcountries. It continues to increase because of the public policies described above.Currently, some youngworkers are forced to quit their jobswhen their elderly parentswho need intensive care cannot find public nursing homes.

11Insurance premiums for the national health insurances are different depending on where peoplelive.12The copayment rate is dependent on age and income.


Fig. 8.6 Share of recipients receiving long-term care at home. Source: OECD, Health at a Glance2015, Fig. 11.3

8.3.2 Family Policies

A similar attitude of the government can be observed in policies for families withchildren, that is, raising children is the role of families and not of the government.Unless the children or families need special care, the government has not providedmuch support.

As a result, the public support for families that raise children has been small inJapan (c.f. Figs. 8.4 and 8.9). Public support for families can be categorized intotwo types, cash benefits (e.g., child allowances) and in-kind benefits (e.g., daycareservices for children). We explain each type of public support and discuss how theyshould be designed to overcome the population crisis.

8.3.2.1 Cash Benefits

In Japan, cash benefits for raising children have been given under both the socialsecurity system and the tax system. Under the social security system, child allowanceis given to families whose household income is below a certain level. Under the taxsystem, there is an income deduction for each dependent child. Hence, as the numberof children increases, the tax payment is reduced. As the latter involves incomededuction, the benefits are higher for high income households with high marginaltax rate.

However, the system, which allows high-income families to enjoy higher childbenefits, has been criticized. In 2010, after the Democratic Party assumed politicalpower, the new government replaced the old systems of child allowance and incometax deduction for children with a new type of the child allowance system that gives

http://dx.doi.org/10.1007/978-4-431-55909-2_11


every family a fixed amount for each child. However, in 2012, after the LiberalDemocratic Party regained political power, it abolished the new child benefit systemwith the reason that providing child allowances to every family was costly. It restoredthe income-deduction system with some revisions.

Economic theories (e.g., Groezen et al. 2003) in the last section suggest that childallowances for every family is justifiable as a subsidy to internalize the positiveexternalities associated with raising children. However, the fixed amount of childallowance in Japan was so small that it neither helped the families nor provided themwith incentives to have more children.

Needless to say, child allowances are really helpful for low-income families.Hence, when the government’s budget is tight, it may be better to give childallowances to low-income families and not to the high-income families. For fam-ilies with relatively high-income, the other type of family policy, that is, in-kindbenefits, will be more useful in helping and giving families adequate incentives tohave more children.

8.3.2.2 In-Kind Benefits

There has been a social norm in Japan that men should work for bread and womenshould stay at home. Under such a norm, mothers are expected to stay at homeand raise children. Hence, nursery services have been limited and provided only tochildren who cannot have adequate care at home. As such, unless the couples livenearby or with their parents, women had to quit their jobs after they get married orhave children. If they want to continue working then they have to give up havingchildren.13

As women are getting educated more and more, they are trying to remain in thelabor markets. This has led to a decline in fertility rates (c.f. Sect. 4.2.2). Confrontedby such a problem, from the mid-1990s, the government decided to implement theso-called “angel plan” to expand nursery services for working mothers.

Although the government had been expanding nursery services, the supplyremained much smaller than the demand, leading to excess demand for a longtime (c.f. Remark 8.3). Even today, after the government’s attempt to reduce excessdemand since the 1990s, the number of children waiting to get into publicly certifiednursery schools is quite high. Figure8.7 indicates that, as compared to Sweden, thecapacity for childcare is quite limited and most of the children are still raised bytheir mothers at home in Japan services has negative effects on the female laborparticipation in Italy where both the female labor participation rate and the fertilityrate are low as in Japan (c.f. Fig. 8.10).14, 15

13Del Boca and Vuri (2007) showed that the rationing in the use of the childcare.14See Takeda (2005, Chap.6) for the historical development of family policies after the 1990s.15An official estimation of the number of children who cannot get into publicly certified nurs-eries is more than 20,000. However, the government’s estimate of the excess demand depends on the

http://dx.doi.org/10.1007/978-4-431-55909-2_4


Fig. 8.7 Places where children stay in the daytime. Source: Cabinet Office of Japanhttp://www.kantei.go.jp/jp/singi/syakaihosyoukokuminkaigi/kaisai/jizoku/dai06/siryou2.pdf andhttp://www8.cao.go.jp/shoushi/kaigi/ouen/kihon/k_4/pdf/s2-1-2.pdf

Fig. 8.8 Markets forchildcare service

0

Price

Quantity

p∗

Q∗QS QD

p0

p1

Demand Curve

Supply Curve

E

Remark 8.3 (Demand and Shortage in Childcare Services) In urban areas, there hasbeen a shortage in the supply of childcare services. There still is a long waiting list ofpeople trying to get admission into publicly authorized childcare centers. Althoughthe government has tried to increase their numbers, there is no sign of the length ofthe waiting list getting any shorter because there is a big potential demand that is notyet on the list.

The problem can be easily explained by the standard demand-supply figure.Figure8.8 depicts the downward-sloping demand curve and the upward-sloping sup-ply curve for childcare services.

If services are traded in the market, the market equilibrium is expected to be atpoint (p∗, Q∗) where the demand is equal to the supply. In the strictly controlledchildcare system in Japan, the price for the childcare services is determined by thegovernment. It is usually set lower than the market price. At price p0, the demand

(Footnote 15 continued)definition and timingof the estimates. The potential excess demand for nursery services is consideredto be much larger.

http://www.kantei.go.jp/jp/singi/syakaihosyoukokuminkaigi/kaisai/jizoku/dai06/siryou2.pdf

http://www8.cao.go.jp/shoushi/kaigi/ouen/kihon/k_4/pdf/s2-1-2.pdf


(QD) is larger than the supply (QS). The excess demand corresponds to the longwaiting list, including the potential demand, for childcare services.

Once we recognize the essence of the problem, we can find an easy solution tosolve the shortage in the supply of childcare services quickly. It involves raisingthe price of the service to p1, where the demand is equal to the current supply.Furthermore, if the price is set at the market price p∗ or if the market mechanismis allowed to work in the case of childcare services, then there will be an increasein the supply and more people will be able to receive these services. It is importantto recognize that the core of the problem of shortage in the supply of childcareservices is in the determination of the price. If the price is lower than the marketprice, the government should provide enough subsidies until the supply meets thedemand. Otherwise, inefficiency as well as the inequality among those who can usethe service at a low price and those who cannot use the service at all will prevail.


In this chapter, we reviewed economic theories of family formations and tried toexplain the decline in fertility, which caused the population aging problem in Japan.16

We argued that the development of markets and the social security system are thetwo main forces behind the decline in fertility. The expansion of markets allowed theJapanese people to work and save their earned money to prepare for various risks intheir lives. This reduced the benefits of having children. The increase in the women’swages encouraged young women to stay in the labor markets by raising the costs ofhaving children. Thus, the expansion of the market system lowered the net benefitof having children, and thus, the number of children has declined.

The expansion of the social security payment since 1973, the first year of thewelfare state, further reduced the demand for children because, under the pay-as-you-go public pension system in Japan, people can free ride on the contributions ofothers’ children. Hence, the number of children has become inefficiently low.

As of 2016, the TFR in Japan is 1.45, which is far below the replacement level of2.07. In such a circumstance, the most demanded policy is to provide subsidies forraising children to stop the inefficient decline in children. As raising children has pos-itive externalities, especially under the pay-as-you-go pension system, such a subsidycan be viewed as a Pigouvian subsidy (c.f. Sect. 3.3.2) for correcting the undersupplyof activities with positive externalities (i.e., raising children). The subsidy can takethe form of cash benefits and in-kind benefits for families with children.

However, the Japanese government has not yet increased the public expenditureon families very much. Even today, the ratio of social expenditure on families toGDP is one of the lowest among the OECD countries (c.f. Fig. 8.9).

Our analysis, depicted in Fig. 8.10a, suggests that an increase in the social expen-diture on in-kind benefits for families is likely to have positive effects on total fertility

16See Yamashige (2014) for more discussion on the issue in the chapter.

http://dx.doi.org/10.1007/978-4-431-55909-2_3


Fig. 8.9 Social expenditure on families. Source: OECD.stats. Note: Data is for 2013 and those ofcountries with * is for 2010

(a) (b)

Fig. 8.10 Social expenditure, female labor participation rate, and total fertility rate. Sources: Socialexpenditure on in-kind benefits for families per GDP, female labor participation rate (FLP), andtotal fertility rate (TFR), are data of 2013 available at OECD.stat

rate and female labor participation. We can see in Fig. 8.10a that the social expen-diture on in-kind benefits per GDP is low in Japan. The figure suggests that it is animportant reason for the low fertility rate and low female labor participation rate inJapan.


On the other hand, in countries with high social expenditure for in-kind benefits(e.g., Sweden and Iceland),womencanworkmore andhavemore children. It has beenpointed out that the female labor participation rate and the fertility rate came to havea positive relationship as in Fig. 8.10b, while their relationship was negative in thepast (e.g., Kögel 2004). Our analysis suggests that it is because each has a positiverelationship with the social expenditure for in-kind benefits, and some countriesstarted to increase the latter. It also suggests that by increasing the social expenditurefor childcare services, Japan is likely to enjoy higher fertility rate and female laborparticipation rate, which is most demanded in Japan.17 Japanese government has notimplemented such a policy and its social expenditure on families stays at low level.

An alternative policy to mitigate the low fertility rate in Japan is to reduce thelevel of social security which has contributed to the lowering of fertility rates (c.f.Fig. 8.4). However, such a policy is very difficult to implement in a society, where theolder generations who rely on social security or going to do so soon are increasingin numbers.18 From a political perspective, the expansion of social expenditure forfamilies cannot be easily implemented in Japan. The traditional social norm in Japanhas been that of “familialism” (c.f. Esping-Andersen 1990),which places a high valueon families with the expectation of mutual aid within them. This social norm is stillheld by the elderly people, especially the conservative politicians. The governmenthas taken active policies to help traditional families, but the social expenditure onmodern families (e.g., dual-income nuclear families) has been so low in Japan. Thepolitical decision is the essential problem of the population crisis in Japan. We willhave more discussion about it in Chap.11.

So far, we considered policies for raising fertility rates in Japan. There are twoother approaches to overcome the population crisis. The first is a policy to increaseimmigrants rather than babies. The second is to give up on increasing people and relyon capital such as robots and artificial intelligence (AI). When the young generationsare shrinking, both approaches will be necessary. However, each approach comeswith its own difficulties.

First, the Japanese society has been very prudent in increasing immigrants. Theratio of foreign and foreign-born people to the total population of Japan is still verylow, as indicated by Fig. 8.1. The labor shortage is getting more serious, especiallyin the service sectors that depend on growing elderly generations (e.g., medical andcare services). Hence, there will come a time when the government will need torely on foreign workers, even if the people do not want to. This may lead to varioussocial problems such as the immigrants being excluded from society due to a lack ofpreparation by the Japanese to welcome foreigners.

Second, although the reliance on robots and AI seems to be a good policy tocompensate for the decline in labor, these high-tech capital require investmentswhose

17Some countries, such as France and Ireland which have exceptionally high total fertility ratescompared with their female labor participation rates (i.e., large positive deviations from the regres-sion line in Fig. 8.10b), are spending fairly large cash benefits for families. For example, the cashbenefit for families in France is 1.57% and that of Ireland is 2.93%, while that of Japan is 0.88%,as percentage of GDP.18Such a policywould be an inevitable choice if the Japanese government suffers a financial collapse.

http://dx.doi.org/10.1007/978-4-431-55909-2_11


returns go to the investors and not the workers. Those who can invest in and workwith the high-tech capital will earn high incomes, while others will suffer from lowwages and unemployment (c.f. Sachs and Kotlikoff 2012). The government is likelyto face new problems such as inequality, unemployment, and low tax revenue underthe pressure of capital flight and brain drain in a highly globalized economy, whichwe further discuss in our Epilogue (Chap. 11).

References

Barro, R., & Becker, G. (1989). Fertility choice in a model of economic growth. Econometrica,57(2), 481–501.

Becker, G., & Barro, R. J. (1988). A reformulation of the economic theory of fertility. QuarterlyJournal of Economics, 103, 1–25.

Coulmas, F., Conrad, H., Schad-Seifert, A., & Vogt, G. (Eds.). (2008). The demographic challenge:A handbook about Japan. Leiden: Brill Academic Publishers.

Del Boca, D., & Vuri, D. (2007). The mismatch between employment and child care in Italy: Theimpact of rationing. Journal of Population Economics, 20(4), 805–832.

Ehrlich, I., & Kim, J. (2007a). Social security and demographic trends: Theory and evidence fromthe international experience. Review of Economic Dynamics, 10(1), 55–77.

Ehrlich, I., &Kim, J. (2007b). Has social security influenced family formation and fertility inOECDcountries? An economic and econometric analysis. Journal of Pharmaceuticals Policy and Law,9, 99–120.

Esping-Andersen, G. (1990). The three worlds of welfare capitalism. Cambridge: Polity Press.Hirao, K. (2007). The privatized education market and maternal employment in Japan. In F. M.Rosenbluth (Ed.), The political economy of Japan’s low fertility. Stanford: Stanford UniversityPress.

Kögel, T. (2004).Did the association between fertility and female employmentwithOECDcountriesreally change its sign? Journal of Population Economics, 17, 45–65.

Imrohoroglu, A., Imrohoroglu, S., & Joines, D. H. (1995). A life cycle analysis of social security.Economic Theory, 6(1), 83–114.

Myrdal, A. (1941).Nation and family: The Swedish experiment in democratic family and populationpolicy. Cambridge: The MIT Press.

National Institute of Population and Social Security Research. (2014). Social security in Japan.http://www.ipss.go.jp/s-info/e/ssj2014/PDF/ssj2014.pdf.

Oguro, K., Takahata, J., & Shimasawa, M. (2011). Child benefit and fiscal burden: OLGmodel withendogenous fertility.Modern Economy, 2, 602–613.

Puhakka, M., & Viren, M. (2006). The effects of the size of the public sector on fertility. DiscussionPapers No. 8, Aboa Centre for Economics.

Raymo, J., Iwasawa, M., & Bumpass, L. (2004). Marital dissolution in Japan: Recent trends andpatterns. Demographic Research, 11(14), 395–419.

Sachs, J. D., & Kotlikoff, L. J. (2012). Smart machines and long-term misery. NBER WorkingPaper, 8629.

Takeda, H. (2005). The political economy of reproduction in Japan: Between nation-state andeveryday life. RoutledgeCurzon: London; New York.

van Groezen, B., Leers, T., & Meijdam, A. C. (2003). Social security and endogenous fertility:Pensions and child allowances as siamese twins. Journal of Public Economics, 87(2), 233–251.

Yamashige, S. (2014). Population crisis and family policies in Japan. University of Tokyo Journalof Law and Politics, 11, 108–128.

http://dx.doi.org/10.1007/978-4-431-55909-2_11

http://www.ipss.go.jp/s-info/e/ssj2014/PDF/ssj2014.pdf

Chapter 9Poverty

Give a man a fish and you feed him for a day; teach him how tofish and you feed him for a lifetime.

Lao Tzu

9.1 Introduction

In market economies, people fall into poverty under two conditions. First, peoplecannot earn enough money to cover their cost of living. This happens when peopledo not have enough assets and abilities to sell to others (e.g., due to unemployment ordisability) and/or if they are confronted by high costs of living (e.g., due to sickness).Second, the people under the first condition do not receive aid from anybody. Evenif people do not have enough money to make a living, they need not fall into povertyas long as they have good friends, families, or adequate insurances.

The first condition can be referred to as the “negative shock,” and the second oneas the “lack of aid” or the “lack of insurance.” In our lives, we face various “risks” ofnegative shocks. To prepare for such risks, we try to have various “insurances” againstpotential negative shocks. Thus, if poverty is increasing, either risks of negativeshocks are increasing or the system of aid is shrinking. Hence, to reduce poverty, weneed to reduce the probability and magnitude of negative shocks and/or enhance thesystem of aid.

An interesting and important aspect of the aid is how we help those who facenegative shocks. For example, if the social security system gives them fish every day,then people will depend on it and the government has to incur large costs for it. Sucha policy can increase the number of people who would not learn how to fish but tryto rely on the fish provided by the government. That is, public policies can increase,the number of poor people relying on government support.

Lao Tzu, who is usually considered to be around 6th century BCE, is an ancient Chinesephilosopher and the founder of Taoism.


177

178 9 Poverty

This can be described as the moral hazard problem of the generous public assis-tance programs, which essentially involve redistribution of income to the poor fromthe rest. On the other hand, if the government equips people with skills to catchfish, then the people can be independent and the government can save money in thelong-run. Hence, designing good policies for those who face negative shocks is reallyimportant in mitigating poverty.

A big challenge in modern society is related to the replacement of domesticworkers by foreign workers and capital, which has become easier with globaliza-tion and technological development. The risk of losing jobs is much higher now,especially among low-skilled workers. Hence, it has become more important thanever to strengthen policies for mitigating poverty. However, globalization makes theimplementation of redistributive policies more difficult because capital and talentedpeople can easily move to foreign countries if heavier taxes are imposed on themfor redistribution. Many countries that chose to be a part of the globalized societyare now facing challenges in mitigating poverty and conflicts between winners andlosers under the increased difficulty in implementing redistributive policies.

In Sect. 9.2, we first summarize theories to explain poverty and apply themto understand the situation of poverty in Japan. Then, in Sect. 9.3, we discussJapanese policies aimed at mitigating poverty and make suggestions for their reform.Section9.4 provides some concluding remarks.

9.2 Theories and Facts

As indicated in Chap.1 (c.f. Fig. 1.11), poverty in Japan has been increasing. Toexplain the situation in Japan, we start by understanding the theories on poverty.

9.2.1 Theories for Understanding Poverty

As discussed above, the risk of negative shocks and lack of aid during a state ofnegative shock are the main causes of poverty. There are so many risks involved inour lives: risk of getting sick and injured, which require costs of care; risk of beingunemployed and unable to find jobs, which cause negative income shocks; and riskof being born with handicaps, which prevent people from earning enough income tocover their costs of living.

There are many ways to mitigate negative shocks and avoid falling into poverty.We can categorize them into three types as depicted in Fig. 9.1. They include self-help, mutual aid, and social security.

http://dx.doi.org/10.1007/978-4-431-55909-2_1

http://dx.doi.org/10.1007/978-4-431-55909-2_1

9.2 Theories and Facts 179

Fig. 9.1 The three types of aid for mitigating risks

9.2.1.1 Self-Help and Risk-Sharing

Having insurances and savings, if available, is a good example of self-help (self-protection) method that we can employ to mitigate negative shocks. Having a goodeducation will not only help us reduce the risks of negative shocks, but also allow usto mitigate negative shocks, for example, in finding new jobs.

Getting married and having children are also good ways that individuals canchoose to strengthen their self-help capabilities. Having a spouse and children canhelp people when they are sick, injured, or fired. When the social security was notdeveloped, children were important for the subsistence of most elderly people (c.f.Sect. 8.2.2). In fact, it was a prevalent belief in the past that people must get marriedand have children to live longer. In our own lives, we also experience the preciousnessof having good friends in times of negative shocks.

The reason we receive support from our spouse, children, and friends is that theytoo expect to be helped in their time of need. Hence, from a social perspective,having spouses, children, and friends can be viewed as a system of mutual aid.1

Mutual aid can be viewed as a way of sharing risks with people we trust. It isan informal insurance against various risks, using a similar mechanism of formalinsurance contracts in which risks are shared among people whose risks are nothighly correlated.

Remark 9.1 (Risk-Sharing within Communities) In economics, risk-sharing withinfamilies and communities has received much attention and it has been analyzedwell, especially in developing countries where financial markets for risk-sharing areunderdeveloped. For example, Townsend (1994) found that the consumption of eachhousehold in three poor villages in southern India is correlated with the averageconsumption of the village. It means that the consumption of each household does

1The argument suggests that it is very important to have good spouses, children and friends; andhave good relationship with them. If they are not good, they can be a burden for us by increasingthe risks of negative shocks.

http://dx.doi.org/10.1007/978-4-431-55909-2_8

180 9 Poverty

not respond to its income fluctuations, and thus suggests the existence of risk-sharingwithin the community.

Fafchamps and Lund (2003) examined how the households responded to theshocks in income and expenditure in the farms of the Philippines and they found thatit is closely related to the exchange of gifts and borrowing within the network offriends and relatives. Thus, they found evidences of risk-sharing within the personalnetworks, rather than the whole village.

Now, in the developed countries, the social security system has been developed tocover the risks that self-help and mutual aid cannot handle well. It consists of socialinsurance and public assistance. The former consists of insurances for unemploy-ment, pension, healthcare, and long-term care. They are usually financed by socialinsurance contributions and taxes. The public assistance is purely financed by taxrevenues and intends to guarantee the minimum standards of living for all people.

9.2.1.2 Moral Hazard Problems

Economics of uncertainty has clarified that when people have generous insurancesthe problem of moral hazard occurs, wherein people behave carelessly against risksbecause the losses will be compensated by the insurances. Such a rational behaviorcan be viewed asmorally undesirable, and thus is referred to asmoral hazard behavior(c.f. Remark 3.7). For example, when we have full insurance for car accidents wetend to drive less carefully. The degree of the moral hazard problems gets higher asthe insurance gets better in terms of covering negative shocks.

The same problem is likely to occur when the government starts providing goodsocial security to the people. People will be less careful about reducing the risks ofnegative shocks andwillmake fewer efforts in strengthening their self-help capabilityagainst such risks. This type of individual behavior will result in a weaker mutualaid system.

That is, when the social security system is well developed, it becomes rationalfor people to make smaller investments in education and financial assets to preparefor various risks in their lives. The motivation to get married and have children toprepare for risks in our lives will also decrease. People will spend less time with theirfriends to build good friendship for helping each other. Theweakening of families andcommunities, which worsens the poverty among the elderly, youth, single mothers,and children, can be explained as a consequence of the moral hazard behavior underthe generous social security system.2

2We may also describe it as the “crowding-out effect” of government’s provision of social securityon private investment in the formal and informal insurances for various risks (c.f. Sect. 2.3.1). Whenthe government increases its expenditure for social security, private expenditure and time for self-help will be reduced, that is, they will be crowded out (c.f. Costa-Font 2010). Such crowding-outeffects of social security will have deep impacts on society, especially when they lead to decliningfertility that endangers the sustainability of the social security system and of the society. Note thatcrowding-out effect can affect mutual aid when the self-help system is enriched by the development

http://dx.doi.org/10.1007/978-4-431-55909-2_3

http://dx.doi.org/10.1007/978-4-431-55909-2_2


Fig. 9.2 Public assistance and public pension. Source: Ministry of Health, Labor and Welfare

Moral hazard problems can also occur within the social security system. Forexample, when the public assistance programs are generous, people can enjoy aminimum standard of living even if they do not receive good education, work, save,or participate in social insurances. That is, the generous public assistance programscan cause a lot of moral hazard problems. As we explain in Sect. 9.3, the Japanesepublic assistance programs are quite generous, which seems to be the reason whythe ratio of people who pay the pension contributions has fallen to around 60% (c.f.Fig. 9.2)3 among the Category-1 people (e.g., self-employed people). Such a lowratio must be contributing to the increase in the ratio of poor elderly people underpublic assistance programs in recent years, as depicted in Fig. 9.2.

(Footnote 2 continued)of a financial market system (c.f. Sect. 5.3.2). Hence, the development of the market system is alsolikely to have contributed to declining marriage, fertility and friendship in the modern society.3The ratio dropped in 2002 because the collection of pension contributions shifted from the localgovernments to the national social insurance agency.When the local governments were responsible,they collaborated with the neighborhood associations in collecting the contribution. The basicdownward trend since the early 1990s is explained by the decrease in income and the increasingdistrust of the sustainability of the public pension system.When people do not pay the contributions,they do not receive any pension payment, and face troubles in their lives after retirement. However,they will not be in trouble when there is generous public assistance. Thus, the low contributionrate to the public pension system can be explained as a moral hazard problem associated with thegenerous public assistance programs.

http://dx.doi.org/10.1007/978-4-431-55909-2_5

182 9 Poverty

9.2.2 Increasing Poverty in Japan

The typical negative shocks that lead to poverty are different for different age groups.Hence, we distinguish the nature of poverty across different age groups.4

9.2.2.1 Poverty among Elderly People

It was shown by Ohtake (1997) and Ohtake and Saito (1998) that an increase inthe Gini coefficient during the 1980s (c.f. Fig. 1.11) is mainly caused by populationaging and the increase in elderly households. This is because the income inequalitywithin the age-groups is the highest among elderly people. The increase in the elderlypeople contributes to the increase in the average income inequality in Japan.

Based on this finding, one may argue that we need not worry about the increasein the Gini coefficient. However, the fact above suggests that the number of elderlypeople living in poverty may be increasing in Japan, which can be confirmed in theincreasing number of elderly people under the public assistance program, as depictedin Fig. 1.13.

If the elderly people were living with their children, they would usually receivegood care and the Gini coefficient would not be high. However, the ratio of elderlypeople living with their children has been declining (c.f. Fig. 1.6). Hence, a part ofthe problem of the poor elderly can be explained by weakening families, that is, theweakening mutual aid system for the elderly who face low income after retirement.

9.2.2.2 Poverty among Children and Single Mothers

In Japan, it is well-known that about one out of six children lives in poverty.5 Theratio of children in poverty is among the highest in the OECD countries. There aremany Japanese children who are suffering from poor living conditions (c.f. Fig. 9.3).

Figure9.4 indicates that poverty among children is high in households with singleparents. The number of divorces and the ratio of divorces to the marriage have beenrising in Japan (c.f. Fig. 1.5), contributing to rising poverty among children.

It is especially high among single-mother households because single mothers onan average earn less than other family groups. The poverty rate of the single-motherhouseholds in 2004was about 66%, which is the second highest among the 24OECDcountries (Abe 2008, pp. 52–54). On the other hand, the poverty rate of single-fatherhouseholds is about 19%, which is higher than the average, but much lower than thatof single-mother households.

Hence, the low wage income among women is one of the most essential problemsfor poverty among children. Even though the average wage for women is lower than

4See Tachibanaki (2005) for a comprehensive study on the income inequality in Japan.5The definition of “poverty” here is relative poverty; hence, it does not indicate that these childrenare in absolute poverty.

http://dx.doi.org/10.1007/978-4-431-55909-2_1

http://dx.doi.org/10.1007/978-4-431-55909-2_1

http://dx.doi.org/10.1007/978-4-431-55909-2_1

http://dx.doi.org/10.1007/978-4-431-55909-2_1


Fig. 9.3 Poverty among children. Source: Cabinet Office of Japan (2014, Sect. 3.1)

Fig. 9.4 Poverty among children. Source: Cabinet Office of Japan (2014, Sect. 3, Chap.3)

that for men, it is found that most of the wage gap can be explained by gender produc-tivity gap not by the sexual discrimination (e.g., Kawaguchi 2007). The difference inthe wages comes from the difference in the jobs and productivity of women andmen.More men than women are engaged in jobs that pay high salaries. Such differences

184 9 Poverty

in jobs is likely to be the result of sexual division of labor under prevalent socialnorms in Japan.

There are two effects of such social norms on women’s wages. First, more womenwill quit their jobs after they get married or pregnant. The discontinuation of workwill reduce the wage, especially under the wage system based on seniority and life-time employment in Japan. Second, considering that women quit their jobs after theyget married or pregnant, the rate of return from education will be very low; hence,women tend to have lower levels of education as compared tomen. Even if women goto university, it is observed that they tend to choose majors that have low correlationswith wage income after graduation (e.g., Yamada 2004, Chap.7). Hence, lower levelsof investment on women’s education and skill development for employment underthe social norm of sexual division of labor, is one of the most important reasons forwomen’s wages generally being low and for single mothers to fall into poverty.

9.2.2.3 Poverty among Young People

Issues of poverty among young people in Japan are well-known from the discussionon the “working poor.” Themajority of theworking poor comprise of non-regular andpart-time workers. This problem is not unique to Japan and is prevalent worldwide(c.f. UNESCAP 2007).

The reason behind the increasing number of low-wage workers in developedcountries can be attributable to globalization (c.f. Förester-Pearson 2002). There areabundant low-skilledworkers in developing countries, and in the globalized economy,more production operations are carried out in these countries at low-wages. Hence,in developed countries, the demand for low-skilled workers decreased, while that forhigh-skilled workers increased. Income inequality among working people expandedand the low-skilled workers, who faced low wages, made up the working poor.

Furthermore, as the young population, and then the total population, started todecline in Japan, Japanese firms can no longer rely on steady growth of domesticdemands. They need to grab demands in the global markets where competition isharsh. Productions of the Japanese firms are affected more by the changes in demandand supply of other countries.When firms face higher uncertainty in their production,they prefer flexible employment. Hence, the number of workers that the Japanesefirms can guarantee life-time employment to has been declining. More and moreworkers are being employed as non-regular and part-time workers.

In other words, in the globalized economy, workers have started to face higherjob uncertainty as firms are facing higher competition and thus uncertainty in globalmarkets. Hence, the risk of negative shocks among young people has gone upwith theglobalization of our economy. If young people have a good system of aid to mitigatesuch risks, they need not have fallen into poverty. However, the transformation ofthe society has weakened the mutual aid system.


Their parents no longer need to rely on them because of the good social securitysystem; hence, they have lower incentives to provide their children with good edu-cation and aid. The ties of the communities have also been weakening, making itdifficult for young people to get enough support from their friends and neighbors.The government too has not provided adequate support for the young people because,unlike elderly people, they are the ones who can and should be able to assist them-selves. The globalization and the weakening ties of families and communities seemto be the basic reason for the increase in the population of the working poor in theworld.6

9.3 Public Assistance and Support for the Poor

The current public assistance program was introduced in Japan after the SecondWorld War and it has played an important role in guaranteeing the Japanese a mini-mum standard of living.7 Thanks to this program, we rarely see poor people on thestreets or in the rural areas of Japan. As the economy developed and society matured,the minimum standard of living set in the public assistance program has been raisedand with it the program started to face many problems.

First, the number of recipients has increased recently, leading to an increase in theexpenditure on the program. Second, the stigma of applying for the program has beendeclining. As it guarantees a minimum standard of living, it has been discouragingpeople fromworking, saving, and even paying the pension contributions for life afterretirement. Hence, it may tap the poor people in poverty, which is often referred toas “welfare trap.”

To mitigate such moral hazard problem under the generous public assistanceprogram and meet the challenges of increasing poverty in Japan, the governmenthas tried to enhance public support for those who are suffering from poor livingconditions. However, the tight budget constraint with public debt accumulation andpopulation aging has left little room for the Japanese government to expand publicassistance and support for the increasing population of poor people in Japan.

6The youth unemployment rate was increasing in the 1990s after the burst of the bubbles (c.f. Mitani2008). It has been recently declining in Japan and is around 5% as of 2016. This is in sharp contrastwith the high rates in most of other developed countries. Due to the declining young populationin Japan since 1995, the demand for young workers has become stronger. Even if young peoplecannot find regular jobs, they tend to chose to work at low wages partly because they cannot getmuch support from the government and families if they are unemployed. See Arai et al. (2015) forexplanation of some changes in the labor market for the young Japanese.7Garon (2002) provides an interesting and convincing argument about how the Japanese governmenthistorically has tried to discourage people to depend on the public assistance program, and encouragethem to depend on their families and communities.

186 9 Poverty

Table 9.1 Minimum standard of living in public assistance program

Area type Household type Single elderly Elderly couple Single mother

Composition(age)

Single (68) Husband (65)Wife (65)

Mother (30)Children (4, 2)

Area 1-1 (Urbanarea)

Basic life 80,870 120,730 189,870

Housing(Maximum)

53,700 64,000 69,800

Total 134,570 184,730 259,670

Area 3-2 (Ruralarea)

Basic life 65,560 97,860 159,900

Housing(Maximum)

32,000 38,000 42.000

Total 97,560 135,860 201,900

Source: Ministry of Health, Labour and Welfare (Outline of Public Assistance Program, 2016)

9.3.1 Public Assistance Program

The newConstitution of Japan, promulgated in 1946 after the SecondWorldWar, hasan explicit provision that requires the government to guarantee a minimum standardof living to its subjects. Article 25 states the following:

All people shall have the right tomaintain theminimum standards ofwholesome and culturedliving. In all spheres of life, the state shall use its endeavors for the promotion and extensionof social welfare and security, and of public health.

Although it was difficult for the government to play such a role right after thewar, it finally introduced the public assistance program in 1951. It provides publicassistance to those who have no ability, no wealth, no one with obligation to supportthem, and no other legal means to maintain a minimum standard of living.8

9.3.1.1 Basic Structure

To avail public assistance, one has to have nothing. If the poor have their own houses,they usually have to sell them. In order for the poor people to receive the welfarepayment, they have to prove that they have no parents, children, or siblings to sup-port them. The conditions for receiving public assistance are quite difficult. However,once the people manage to, the minimum standard of living guaranteed by the publicassistance program is fairly high. Table9.1 depicts three types of household to illus-trate the amount of payment provided under the Japanese public assistance program.

Three types of household inTable9.1 are single elderly household, householdwithelderly couple, and single-mother family. The payment is dependent on household

8A good reference for the public assistance program can be found in NIPSSR (2014, Chap.6).

9.3 Public Assistance and Support for the Poor 187

composition, age, and cost of living. Areas are divided into six types of areas. InTable9.1, cases of the most expensive area (Area 1-1) and the least expensive area(Area 3-2) are illustrated. For example, if a man at age 68 lives alone in the mostexpensive area (Area 1-1), he is entitled to receive �80,870 for his basic life and�53,700 at maximum for rents of an apartment.

Note that the basic pension payment per month is now about �65,000 if one paidthe pension contribution for forty years. It is much less than the public assistanceprogram guarantees for the elderly people. Even if the man has not paid pensioncontribution at all in his life, he can receive �134,570 every month once he isadmitted to be in the public assistance program.9

Table9.1 also shows that a single mother with two children at age 2 and 4 canreceive a monthly payment at �201,900 in a most rural area and �259,670 if shelives in urban area. Many Japanese people often get surprised when they find outhow generous the Japanese public assistance program is. Hence, it is no wonder thatthe number of recipients, especially elderly recipients, is increasing.

The high minimum standard of living set in the public assistance program maybe good from the viewpoint of equity. However, economic theories suggest that itwill discourage people from working because they will receive a constant flow ofincome regardless of the time they work for as long as their own income is less thanthe minimum standard of living.

It is also foreseeable that many poor people, who expect to receive welfare pay-ment, stop paying the basic pension contributions because, even if they pay, whatthey can get under the public assistance program is the same, and thus, paying thecontribution in their youth is deemed a waste of money. Hence, the generous publicassistance program has negative side effects on labor supply, savings, and participa-tion in the public pension program. It is important to design a system that mitigatessuch moral hazard behavior (c.f. Sect. 9.2.1.2).

The Japanese public assistance program now has a mechanism to encourage therecipients to work. The recipients can now keep about 15% of their wage income,thus giving them an incentive to earn higher incomes. Despite this improvement, themarginal tax rate on wage income under the public assistance program is as high as85%. For example, even if a single mother worked hard to earn�100,000 per month,the total income will increase by only about �15,000. It will provide recipients withvery little incentive to work.

9.3.1.2 Challenges and Reforms

To provide the recipients with greater incentives to work, a negative income tax orthe so-called earned income tax credit (EITC) can be more effective (c.f. Remark9.1). It is because the tax system guarantees no or little fixed income and increases

9If an elderly person receives the pension payment, the difference between theminimum standard ofliving (�134,570) and the pension income (�65,000) will be paid to him/her in the public assistanceprogram.

188 9 Poverty

the welfare payment as the labor income of the poor increases. Under the system,poor people need to work hard to achieve a minimum standard of living. Such astrong incentive to work will encourage the poor people to be independent. It willalso allow the government to pay less to protect every poor person, and thus helpmore people in need.

Although the introduction of such a system has been discussed in Japan, theyhave not been implemented yet because of the high costs of implementation. Toimplement such a tax policy, the Ministry of Finance needs to acquire accurateinformation about the income of the poor people. However, the tax authority ofJapan has not yet collected information on the incomes of poor households. Undersuch circumstances, it is very costly to introduce and operate the EITC program. Theargument is convincing if we try to embed the system in the tax system.

However, it is reasonable to introduce a system similar to the EITC program in thewelfare system because the governments, especially the local governments, alreadyhave fairly precise information on the incomes of the poor households. The abilityto help more poor people is an attractive feature of the EITC-type public protectionsystem. Although it may be difficult to introduce negative income tax or the EITCin the tax system on a large scale, an efficient mechanism for providing incentivesto work can be introduced to improve the public protection system.

9.3.2 Support for the Poor

9.3.2.1 Tax and Social Security System

In public policies that support the poor, redistribution through the tax system andthe social security system play important roles. Historically, strong emphasis hasbeen placed on the tax system as the redistributive system in Japan. “Direct taxes”such as income tax and wealth tax have been preferred to the “indirect taxes” suchas consumption tax (VAT) as the former can adjust the individual tax burden withrespect to their economic conditions. The share of revenue from direct taxes has beenlarge in Japan as compared to the European countries (c.f. Ishi 2001, Chap.1). The“progressiveness” of income tax and other direct taxes has been discussed extensivelyfrom the viewpoint of redistribution.

However, in the globalized economy, capital and labor can move easily; theirchoice of location is greatly influenced by the tax rates in each country. The high taxrates for redistribution can induce capital flight and brain drain, which can harm theeconomy. It is also important to recognize that the tax system itself cannot improvethe conditions of the poor people who pay little to no tax. A change in the tax systemcan have little effect on the poor.

Considering that globalization endangers the sustainability of progressive directtaxes and the condition of the poor people, it seems wise to use the social securitysystem as the main tool to support the poor. To mitigate poverty, it is especiallyimportant to empower the low skilled workers who face higher risks of falling into


poverty. The tax system should be designed to collect revenues efficiently for suchredistributive policies. It is well-known that indirect taxes such as the VAT are moreefficient than direct taxes (c.f. Stiglitz and Rosengard 2015, Chap.20).

Such a strategy to mitigate poverty can be summarized as follows:

(1) In a tax system,more emphasis should be placed on efficiency than redistribution.(2) In a social security system, more emphasis should be placed on empowerment

(e.g., teaching how to fish) than relief of the poor people (e.g., giving fish).(3) For the efficient improvement in equality, the tax system and the social security

system should be reformed integrally.

Remark 9.2 (Earned Income Tax Credit) In some countries, such as the U.K. andthe U.S., a system to support low income people, which is referred to as “earnedincome tax credit (EITC),” is introduced in the tax system. Although the name of thesystem is different from country to country, the basic mechanism is well-known as“negative income tax” in the theory of taxation (c.f. Stiglitz and Rosengard 2015).The negative income tax system levies a negative tax, that is, a subsidy, on the incomeof low-income people.

To understand the basic idea of negative income tax system, consider a sim-plest income tax in which the tax payment is calculated by (tax rate)× (income−“minimum income”). Suppose the tax rate is 50%. If an individual’s income is greaterthan the “minimum income,” then she/he pays half of the difference as the tax. Ifthe income is less than the “minimum income,” then she/he can receive half of thedifference as a negative income tax (i.e., subsidy). As low-income people can workat higher wages under the negative tax system, they will have a greater incentiveto work. This structure is in sharp contrast with the public assistance program thatdiscourages poor people from working.

The actual EITC is more complex in terms of the conditions and rules of taxcredit. For example, it is often designed to be dependent on the number of children,to support low-income families with children. The EITC suggests that the tax systemcan also provide support for the poor as well as incentives to work. This can beviewed as an empowerment policy. However, for the system to work well the taxauthorities need to have good information about the income and condition of eachhousehold.

In Japan, the tax authority is not sharing information with local governmentswhich have fairly good information about the low-income households to administervarious social security policies. This separation of administrative bodies makes theintroduction of theEITC in the Japanese tax systemdifficult. In the long-run, itmaybeuseful to integrate the administrative bodies of tax system and social security systemso that the government can help the poor people more efficiently with minimumadministrative costs.

190 9 Poverty

9.3.2.2 In-Kind Policy and Lending

We have been emphasizing the importance of empowering people with high risks offalling into poverty, especially in a globalized economy. Such an emphasis resonateswith our basic claim for social policies to shift from consumptive remedy to produc-tive empowerment. Among empowerment policies, providing people with educationand opportunities to work is especially important. Such a policy to provide in-kindbenefits will be referred to as an in-kind policy.10

As regards education policy, recent studies in economics (e.g., Heckman 2006)have indicated the importance of early childhood education amongpoor people. Thus,education in schools and preschools is important. As we have argued in Chap.8,public support for schools and preschools is small in Japan as compared to otherdeveloped countries. This means that if children are born in poor families, it is likelythat they remain poor and their children may remain in poverty (c.f. Fig. 9.4). Thelow public expenditure on education seems to be one of the reasons why some youngpeople are having difficulty in getting regular jobs, and are working as part-timeworkers who have higher risks in falling into poverty.

The expansion of good childcare and preschools is important not only for provid-ing high-quality education to children, but also for providing mothers with opportu-nities to earn wage income. Without such opportunities, women who want to havechildren may have to quit their jobs and become housewives, which can raise therisk of women falling into poverty. If women with children can keep working in thelabor markets, even after getting divorced or losing their husbands, the risk of fallinginto poverty will be minimized, as suggested by the low poverty rate among singlefathers (c.f. Sect. 9.2.2). We have seen that the probability of divorce in Japan hasincreased to about one third among the married couples (c.f. Fig. 1.5).

Providing the elderly people with more opportunities to work will also be usefulin decreasing poverty among elderly people. Due to the huge public debt, strongpressure to cut public expenditure will eventually force the level of public pension togo down. The age at which people start receiving public pension will also be raisedfrom the current one (65years). The retirement age should be raised to give elderlypeople more opportunities to work.

From the viewpoint of empowerment, lending rather than giving support may besomething that policy makers should consider. If the public support can increase thecapabilities of people, chances of their paying back the costs of the support will goup. Under the government’s tight budget constraint, this means that lending can bebetter than giving. For example, expanding opportunities of borrowing money foreducation and training can be a good policy in general to empower the people.11

10The Japanese government started to emphasize the importance of social inclusion and the supportfor poor people to be independent. A new law was implemented in 2015 to enhance such policies.11 It is necessary to emphasize that adequate subsidies should be provided for education to internalizethe external effects of education and training.

http://dx.doi.org/10.1007/978-4-431-55909-2_8

http://dx.doi.org/10.1007/978-4-431-55909-2_1


Expanding chances of borrowing for daily lives may be useful for single mothersand unemployed people, who are often under strict liquidity constraints.12

9.3.2.3 Private Support

Social inclusion is really important for the provision of support to those who arefacing negative shocks. It implies that public as well as private support, are bothindispensable. As we have mentioned repeatedly, in the past, traditional families andlocal communities provided mutual aid to those in need of support and care. Theweakening of traditional communities in Japan is viewed as one of the reasons forthe increase in the suffering of the poor in recent times.

The Japanese local governments have been trying to cooperate with traditionalcommunities to help the poor. For example, the minsei-iin (voluntary communalwelfare workers) were commissioned by each municipal government to assist thepeoplewho need support and care. They are voluntary social workers chosen from thelocal communities and they were expected to work with the municipal governmentsto support the people in need (c.f. Thränhardt 1996; Garon 1997, 2002). The word“voluntary” simply implies that there is nomonetary reward for their work except thatthey can receive a very small amount of money to cover the costs of their activities.

It is considered to be an “honorary post” in a traditional local community inJapan. As of today, there are about 230,000 minsei-iin in Japan. However, it hasbeen pointed out that finding candidates for theminsei-iin is becoming more difficultas local communities are weakened, young people become individualistic, and thenumber of elderly people who need support increases. For most of the people, jobsand responsibilities expected by the municipal governments are too big to be doneas “voluntary” work.

On the other hand, the number of NPOs providing aid to poor people is increasingin Japan. Some examples include food banks for poor households, providing supple-mentary after-school lessons to children belonging to low-income households andproviding adequate shelters for poor people. A recent example is of the activitiesknown as the “dining room for children,” in which volunteers provide dinners tochildren in trouble at very low prices. They are providing not only food at low prices,but also places (communities) where children and parents who are in trouble canbelong to.

The traditional local communities, in which everyone is forced to participate incommunal activities, have weakened. However, we can observe an increase in thewillingness to contribute to the society among the Japanese people (c.f. Fig. 9.5).The number of volunteers in the activities of NPOs to support the poor seems tobe increasing. Some local governments are already trying to coordinate with these

12The expansion of lending for elderly people may be useful and not just from in terms of savingpublic funds. If the money is given to the elderly people, say, in the form of public pension, it mayeventually find its way to the pockets of their children and grandchildren in the forms of gifts andinheritances. If the money is lent to the elderly people, the government may be able to recover somespending from the elderly person’s inheritances.

192 9 Poverty

Fig. 9.5 Willingness to contribute to the society. Source: CabinetOffice (2007, Chap.2, Fig. 2-1-30)

“new communities” that are held together by the will of the people (c.f. Chap. 7).From the perspective of public policies, it will be challenging, but promising for thelocal governments to find, foster, and bring together the goodwill of the local people,that is, to enrich the social capital (c.f. Sect. 1.2.2) in each area, to expand the socialinclusion of low-income households for mitigating poverty in Japan.


People fall into poverty when they face negative shocks and when they have insuffi-cient systems of aid to overcome these shocks. The increase in poverty in Japan canbe partly explained by the increasing risk of negative shocks under globalization andthe shrinking system of aid in society.

Mutual aid received through building families and communities, which has histor-ically been the main system for mitigating risks besides self-help, is reducing due tothe development of markets and a social security system. It is because financial assetsand the government began providing coverage for various risks in our lives; thereby,decreasing the need for mutual aid. Hence, in our theoretical argument, it is mainlythe development of markets and social security system that weakened families andcommunities in the modern society.

From the perspective of public policies, it is very important to recognize that theexpansion of social security has negative side effects on the formation of familiesand communities. One of the serious consequences of such side effects is the declinein fertility and thus population. If the government does not implement any policyto mitigate the problem, it will face a decline in tax revenues and social securitycontributions. When the government is unable to provide adequate social security,people who have been dependent on it will face difficulty in making a living andmay fall into poverty. From a long-run perspective, we may be able to claim that thesocial security system can create poverty.

http://dx.doi.org/10.1007/978-4-431-55909-2_7

http://dx.doi.org/10.1007/978-4-431-55909-2_1


To avoid such problems, the social security system and public policies to helppoor peoplemust be strategically designed. In designing them, we need to realize thatformulating good public polices is becomingmore difficult in the era of globalizationwhere people and governments of foreign countries respond to public policies of eachcountry.

Enhancing the social security system and redistributive tax system to help poorpeople may sound good, but we need to realize that it will have negative side effectsof inducing outflow of smart individuals and companies (i.e., brain drain and capitalflight) to other countries. Public polices to empower individual ability for reducingrisks and mitigating shocks are more effective than redistributive policies.

We may be able to argue that reducing the size of the social security systemwill contribute toward enhancing self-help and mutual aid for a sustainable society.However, such attempts are politically difficult to fulfill in Japan because the elderlypeople are becoming the political majority due to population aging. This argumentsuggests that it is important for the government to correctly foresee the side effectsof the social security system and social policies. The Japanese government has failedto consider these side effects, and thus has been facing difficulties in designinga sustainable system to mitigate poverty in Japanese society. One of the effectivepolicies to mitigate poverty in Japan is to expand investment in human capital andsocial capital to empower individuals and communities in dealing various risks thatcan push us into poverty (c.f. Chap.11).

References

Abe, A. (2008). Poverty among Children: Considering Inequality in Japan. Iwanami Shoten (inJapanese).

Arai, Y., Ichimura, H., & Kawaguchi, D. (2015). The educational upgrading of Japanese youth,1982-2007: Are all Japanese youth ready for structural reforms? IZA Discussion Paper SeriesNo. 8870.

Cabinet Office of Japan. (2007). White Paper on National Lifestyle 2007. Summary in Englishavailable at http://www5.cao.go.jp/seikatsu/whitepaper/h19/06_eng/index.html.

Cabinet Office of Japan. (2014). White Paper on Children and Young People 2014 (in Japanese).Costa-Font, J. (2010). Family ties and the crowding out of long-term care insurance.Oxford Reviewof Economic Policy, 26(4), 691–712.

Fafchamps, M., & Lund, S. (2003). Risk sharing networks in rural Philippines. Journal of Devel-opment Economics, 71, 233–632.

Förester, M., & Pearson, M. (2002). Income distribution and poverty in the OECD area; trends anddriving forces. OECD Economic Studies, No. 34.

Garon, S. (1997). Molding Japanese minds: The state in everyday life. Princeton: Princeton Uni-versity Press.

Garon, S. (2002). Japanese policies toward poverty and public assistance: A historical perspective.World Bank Institute Working Paper, Stock No. 37200.

Heckman, J. J. (2006). Skill formation and the economics of investing in disadvantaged children.Science, 312(5782), 1900–1902.

Ishi, H. (2001). The Japanese tax system (3rd ed.). Clarendon Press.

http://dx.doi.org/10.1007/978-4-431-55909-2_11

http://www5.cao.go.jp/seikatsu/whitepaper/h19/06_eng/index.html

194 9 Poverty

Kawaguchi, D. (2007). A market test for sex discrimination: Evidence from Japanese firm-levelpanel data. International Journal of Industrial Organization, 25, 441–460.

Mitani, N. (2008). Youth employment in Japan after the 1990s bubble burst. In Young workers inthe global economy: Job challenges in North America, Europe and Japan (Chap. 5). Cheltenham,U.K.: Elgar.

National Institute of Population and Social Security Research. (2014). Social security in Japan.http://www.ipss.go.jp/s-info/e/ssj2014/PDF/ssj2014.pdf.

Ohtake, F. (1997).Distribution of incomeandwealth in the 1980s.Theoretical EconomicsQuarterly,13(76), 385–402. (in Japanese).

Ohtake, F., & Saito, M. (1998). Population aging and consumption inequality in Japan. Review ofIncome Wealth, 44(3), 361–381.

Stiglitz, J. E., & Rosengard, J. (2015). Economics of the public sector. New York: WW Norton &Company.

Tachibanaki, T. (2005). Confronting income inequality in Japan: A comparative analysis of causes,consequences, and reform. MIT Press.

Thränhardt, A. M. (1996). Institutional continuity and functional change: The development of thesystem of minsei-iin in the history of social welfare in Japan. In S. Metzger-Court & W. Pascha(Eds.), Japan’s socio-economic evolution: Continuity and change (Chap. 6). Japan Library.

Townsend, R. (1994). Risk and insurance in village India. Econometrica, 62, 539–591.UNESCAP. (2007). Economic and social survey of Asia and the Pacific 2007. United NationsEconomic and Social Commission for Asia and the Pacific.

Yamada, M. (2004). Society with disparity in hopes: Loss of hope of “Losers” can divide Japan.Chukka Shobo (in Japanese).

http://www.ipss.go.jp/s-info/e/ssj2014/PDF/ssj2014.pdf

Chapter 10Regional Disparity

Even the prosperous inevitably decay.Humane King Sutra

10.1 Introduction

The population of rural Japan has been aging and declining at a rapid rate. Thereare many rural areas where the elderly ratio exceeds 50% and their sustainability isunder doubt. The local governments in such areas have poor fiscal conditions becausethey cannot collect much tax revenues, but need to spend a lot on the elderly (c.f.Fig. 10.4). On the other hand, in many urban areas, due to the migration of youngpeople from rural areas and a large number of women at their childbearing age, thepopulation continues to grow. The regional disparity in terms of population has beenexpanding.

Such regional disparities have existed for a long time in Japan, but have not beenwell recognized because of large fiscal transfers to the rural areas (c.f. Sect. 1.3.1).With the huge public debt accumulated in Japan (c.f. Fig. 1.3) and the slowdownof economic growth in Japan (c.f. Fig. 1.2), it is getting more difficult to continuesuch generous fiscal transfers to rural areas. Furthermore, the expected decline inthe urban population in the near future will accelerate the decay of rural areas thatwere prosperous in the past. To make things worse, most Japanese people seem tobe accepting the decay of the rural areas as being inevitable. Is there a way to makethe rural areas sustainable?

Such a view is in sharp contrast to the optimistic views observed in the followingquotes from the OECD Territorial Reviews on Japan’s strategy to revitalize ruralareas (OECD 2016, p. 3).

Humane King Sutra is one of the East Asian Buddhist scriptures.


195

http://dx.doi.org/10.1007/978-4-431-55909-2_1

http://dx.doi.org/10.1007/978-4-431-55909-2_1

http://dx.doi.org/10.1007/978-4-431-55909-2_1

196 10 Regional Disparity

While it is too early to judge the long-term impact of this strategy, we do find groundsfor optimism: the economic analysis confirms that Japanese cities function well, generatingcomparatively strong agglomeration benefits, though better horizontal co-operation amongthe municipalities in metropolitan areas could enhance these benefits substantially. Japan’srural areas, despite their difficulties, actually outperform most of their OECD peers and arein many cases pioneering revitalisation strategies based on local assets and potentials.

This book does not share such optimism. Although we acknowledge that currentrevitalization strategies designed by the government can be effective and pioneering,they will not work well just like the past revitalization initiatives that the governmenttried to implement for several decades since the 1970s (OECD 2016, p. 194).

To explain such a view, we need to clarify the reasons for regional disparityin Japan and the nature of it. Like the last two chapters, we will pay attention tothe transformation of Japanese families and communities and claim that weakenedfamilies and communities are important factors for the decay of many rural areas.

To mitigate the depopulation problem, the national government has been tryingto provide generous subsidies to local governments in rural areas. This is the funda-mental reason why “Japan’s rural areas, despite their difficulties, actually outperformmost of their OECD peers.” Such generous subsidies have contributed not only tothe accumulation of national public debt, but also to the persistence and expansionof regional disparities by allowing poor local regions to depend on redistributivetransfers from the national government. This problem can be described as a moralhazard problem in local regions under the generous redistributive system.

Such policies can be compared to a medical prescription for mitigating symptomswithout curing the underlying disease. The Japanese government should prescribeempowerment policies to regenerate rural regions. The current revitalization strate-gies indeed have some elements of empowerment policies. However, as long as thegenerous subsidies are given to the rural areas, the local governments and peopletherein will not feel a sense of crisis and make efforts to revitalize their respectiveareas. Hence, we cannot expect the current revitalization strategies to work in thelong-run as long as there are generous fiscal transfers to the local governments.

One of the fundamental problems concerning regional disparities is the designof a fiscal transfer system from the national government to the local governments.As the argument above suggests, the design of the fiscal transfer system to reducedisparities can increase the disparities in the long-run. We show how the Japanesefiscal transfer system is designed and how it has affected regional disparities.

In this book, by “regional disparity,” we basically refer to the disparities in thedemographic conditions such as changes in population and the elderly ratio of eachregion. Alternatively, it may be defined in terms of economic disparities measured,say, by the average per-capita income of each region. As we are mainly interestedin the sustainability of the society, we will pay more attention to the demographicfactors over the economic ones.

We do not think that regional disparity is always bad. There can be reasonable andjustifiable regional disparities, when we examine the reasons behind such disparities.However, there are certain unjustifiable disparities that are created by the poorlydesigned fiscal system and social policies. Regional disparity is a signal that suggests


the presence of problems in public policies. It is important to redesign themby payingattention to such signals, to make society sustainable.

In the next section, we describe problems of regional disparities in Japan, andexplain some theories to understand them. In Sect. 10.3, we discuss how Japanesepublic policies have been trying to mitigate the problems and how these policiesshould have been designed.Wewill argue that a shift from consumptive relief policiesto productive empowerment policies is once again critical for the sustainability ofthe Japanese society. Section10.4 concludes the chapter.

10.2 Facts and Theories

In this section, we first describe the problem of regional disparities in Japan and thenreview economic theories to understand it.

10.2.1 Changes in Population Distribution

The Japanese government can be decomposed into three levels (c.f. Fig. 10.1).1 Atthe first level is the national government. Then, there are 47 prefectural governments.The third level consists of the governments of cities, towns, and villages, which arereferred to as municipal governments. The national government has been trying toreduce the number of municipal governments. As of 2016, there are 1,718 municipalgovernments.

10.2.1.1 Regional Disparity in Japan

The population distribution across prefectures is depicted in Fig. 10.2. The columnsin the figure indicate the population distribution of 2015. The curve in the figureindicates the population distribution of 1955 ranging from the highest to the lowest(as it goes to the right).2 The figure clearly indicates that the shift in populationconcentration from rural areas to urban areas has occurred in the past 60years. Thepopulation movement that brought such a change was depicted in Fig. 1.10.

Population distribution in Fig. 10.2 roughly corresponds to the wage distributionacross prefectures depicted in Fig. 10.3. These two figures suggest that population in

1For more information about the current system and issues of local governments in Japan, see Coun-cil of Local Authorities for International Relations (2015). For its history and structural problems,see Steiner (1965).2The names of prefectures corresponding to the curve for 1955 are not the same as those of thecolumns for 2015. However, there are little changes in the population ranking among prefectures.

http://dx.doi.org/10.1007/978-4-431-55909-2_1


Fig. 10.1 Government system in Japan

Fig. 10.2 Population distribution in 1955 and 2015. Source: National census, statistics bureau(http://www.stat.go.jp/index.htm)

Fig. 10.3 Wage and net inflow of population in each prefecture. Source: National census and basicstatistics for wage structure 2014 (in Japanese). Notes: The population inflow represents the netinflow of the young people moved to the prefecture from 1995 to 2010, which is calculated bysubtracting population below 15years old in 1995 from the one for 15–30 years old in 2010. Thewage represents the monthly basic salary excluding bonuses etc.

http://www.stat.go.jp/index.htm


the urban prefectures with high wages are large, because of the population migrationto those urban prefectures.

10.2.1.2 Disappearing Rural Areas

The national government has identified the “depopulating areas” and provided subsi-dies to those areas.3 According to Ministry of Internal Affairs and Communications(2015), the areas cover about 59% of the total land, with about 46% of the municipalgovernments’ districts. Many fiscal measures have been undertaken to save thesepoor regions along with fiscal transfers under the local allocation tax system, as isexplained in Sect. 10.3.1. Such measures allowed many local governments to sur-vive, but created moral hazard problems (c.f. Remark 3.7) and soft budget problems(c.f. Remark 3.9) by which poor local governments could continue inefficient fiscaloperations and accumulate public deficits.4

Figure 10.4 indicates that there are big gaps between per-capita expenditure andper-capita tax revenue especially in the local governments of rural areas with highelderly ratios. The gap has been filled by the transfers from the national government.However, as the whole society is aging, each point in Fig. 10.4, which representsper-capita expenditure or per-capita tax revenue of a local government, will move tothe right5 and fiscal gap will continue to expand further. Under the limited budget ofthe national government with huge national debts, the current fiscal transfer systemwill not be sustainable. The local governments that have been protected by the systemmay have to go bankrupt.

3In administrative terms, the “depopulating area” is defined as an area that satisfies one of thefollowing conditions: (1) the rate of population decline from 1960 to 1995 exceeds 30%; (2) therate of population decline from 1960 to 1995 exceeds 25%, and either (i) the ratio of people above65years of age exceeds 24% or (ii) the ratio of people from 15 to 30years of age is less than 15%;(3) the rate of population decline from 1970 to 1990 exceeds 19%; and (4) the fiscal capability indexis less than 0.42.4A moral hazard problem occurs when there are generous systems of redistribution or insurance. Asoft budget problem occurs when the principal (e.g., national government) rescues the agents (e.g.,local governments).5In Fig. 10.4, per-capita expenditure and per-capita revenue of each local government is depicted bya point. For example, if the elderly ratio of amunicipal government is 20%, its per-capita expenditureand per-capita revenue is respectively depicted by a point locating at 20% on the horizontal axis.The gap between the two points (i.e., fiscal gap) will not be so large when the elderly ratio is 20%.However, in the process of the population ageing, the elderly ratio of the municipal governmentwill go up and its fiscal gap is likely to expand as Fig. 10.4 suggests. As most local governmentswill be ageing, points in Fig. 10.4 will move to the right and the total fiscal gap will be expandedfurther.

http://dx.doi.org/10.1007/978-4-431-55909-2_3

http://dx.doi.org/10.1007/978-4-431-55909-2_3


Fig. 10.4 Population aging, local tax revenue, and expenditure. Source: Statistics bureau (http://www.stat.go.jp/index.htm)

10.2.2 Theories for Understanding Regional Disparity

10.2.2.1 Migration

The regional disparities are caused by the migration of the people. We provide threetheories to understand migration behavior.

First, people search for jobs with high wages and tend to migrate to areas wherewages are high. In general, economicgrowth stimulates populationmigration to urbanareas because the growth usually starts in urban areas (e.g., Harris and Todaro 1970).It explains why people move from rural areas to urban areas in the period of higheconomic growth (c.f. Fig. 1.10). The resulting population concentration in urbanareas produces economies of scale, in the sense that per-capita income increases withthe increase in population. This further induces the inflow of population to the urbanareas (e.g., Krugman 1991). The economies of scale in cities also suggest that thedepopulation of rural areas can reduce their attractiveness and accelerate the processfurther.

The economic theory of migration behavior basically explains the large inflow ofpeople to urban areas in the process of economic growth and the positive net inflowof young people to areas where wages are high (c.f. Fig. 10.3). As the large inflowof people increases the labor supply in these urban areas, the wage gaps amongregions must reduce. This fact can partly explain why the “[s]patial concentration ofeconomic activity is increasing, but spatial disparities are not” in terms of per-capitaGDP in Japan (OECD 2016, p. 16). The disparity of the per-capita GDP, measuredby the Gini coefficient, is indeed smaller in Japan than most other OECD countries(OECD 2016, Fig. 1.15).



http://dx.doi.org/10.1007/978-4-431-55909-2_1


Second, it is well know that public policies and competition among local govern-ments can induce migration. Among various studies on this phenomenon, the ideaof “voting with feet,” coined by Tiebout (1956), is probably the most well-knownone. The idea is that our residential choices are dependent on the policies of the localgovernment, which can be compared to our voting for local public policies with ourfeet. Tiebout (1956) argued that competition among local governments under votingwith feet will induce the government to make efforts to improve the quality of publicgoods and lower the tax burden, which will enhance the efficiency of public policies.

Finally, changes in families and communities can change migration behavior. Inour residential choice, “attachment” to our families and hometowns (communities)plays a role (e.g., Mansoorian and Myers (1993), Dahl and Sorenson (2010) andKawano et al. (2004)). If people have a strong attachment, they will choose to liveclose to their families and stay in their hometowns. On the other hand,when the ties offamilies and communities are weakened, for instance, under generous social security,young people can easily move out of their hometowns and leave their parents. Thistheory can explain the relationship between the development of the social securitysystem and the increase in regional disparity.

Under the Japanese social custom of self-employed people, such as farmers, fish-ermen, and owners of small andmedium-sized enterprises, bequeathing their land andcapital to their children, the young people leaving their hometowns often implies theshutdown of local businesses. The loss of jobs in rural areas will accelerate depopula-tion in these areas, even though the shutdown of inefficient businessesmay contributeto the increase in local productivity. Furthermore, based on the theory of social cap-ital (c.f. Sect. 7.3.3), we can claim that an increase in mobility can reduce the socialcapital of communities, and thus, accelerate the depopulation of rural areas throughmigrations to urban areas.6

10.2.2.2 Social Consequences of Migration

As regards the social consequences of migration, we have several theoretical argu-ments. First, with respect to economic growth and migration, we can claim that theinflow of people to growing regions will increase the efficiency of labor markets andhave positive effects on the national economy. However, there are various externali-ties in the residential choices of people (Sect. 1.3.1). Therefore, we cannot claim thatmigration always enhances the efficiency of society. We will face many inefficiencyproblems in overpopulated as well as underpopulated areas.

Second, as regards to public policies and migration, Tiebout (1956) argued thatmigration enhances the efficiency of society. However, there is a counterargument

6Lee et al. (2016) has analyzed the migration behavior of the people suffered from the Great EastJapan Earthquake, and found that attachment to home and hometowns play an important role inkeeping them stay at the damaged area. However, if people lose their personal network, they showhigher intension to leave the damaged area. Shimada (2015) also studied the migration behaviorafter disasters and obtained similar results. These findings suggest the importance of social capitalin the residential choice although the analyses are made for unusual disaster cases.

http://dx.doi.org/10.1007/978-4-431-55909-2_7

http://dx.doi.org/10.1007/978-4-431-55909-2_1


that the competition among selfish local governments can deteriorate the efficiencyand equity of the society (e.g., Wilson (1999) and OECD (1998)). For example, localgovernments can be involved in a “tax competition” in which each local governmenttries to lower the tax rate for the high-income people to attract the rich, who arethe main contributors to the income tax revenue. Such a tax competition not onlydeteriorates the income distribution, but also lowers the tax revenue to provide ade-quate public goods vis-à-vis the case in which there is no such competition. Hence,migration can have both positive and negative effects on society.

Finally, with regard to the relationship between migration and families and com-munities, we have already suggested above that migration reduces social capital,and thus have negative effects, especially in the short-run (c.f. Sect. 7.3.2). However,other theories suggest that migration may increase the social capital in new commu-nities, and thus can enhance the efficiency of the society, especially in the long-run(c.f. Remark 7.3).

As the above argument indicates, economic theories suggest that migration canhave both costs andbenefits for society.Hence, the government should design policiesand systems to increase the benefits of migration, while minimizing its costs.

10.3 Fiscal Decentralization and Regional Policies

Since theEdo period, starting in 1603, the Japanese society has beenmanaged by onenational government. There were no major civil wars despite the Edo governmentbeing replaced by the new Meiji government. Japan’s history has paved the way fora unitary state and not a federal one.

After the Second World War, the new constitution of Japan, established in 1947under the influence of the U.S., explicitly defined local self-governments. Althoughlocal governments are called “local public entities” in Japan, they have characteristicsof autonomous democratic governments, as per the provisions of the Constitution(Chapter VIII). Unlike the national government in Japan, which takes the form of aparliamentary system, the local governments in Japan represent a presidential sys-tem in which people directly choose the mayors of municipal governments and thegovernors of prefectural governments. Despite this political framework, Japanesegovernment has maintained a centralized system by various regulations and fiscaltransfers from the national government to local governments.

Recently, local governments have been given greater autonomy in determiningtheir policies. In 1995, the decentralization promotion law was enacted for five yearsto promote further decentralization to local governments. However, the regionaldisparity discussed above may suggest that there are many local governments thatare heavily dependent on the national government. Analysis of the fiscal systemand regional policies in this section suggests that to mitigate the disparity betweenrural and urban areas the direction of the current fiscal decentralization to municipalgovernments need to be reconsidered. A greater role of the national government

http://dx.doi.org/10.1007/978-4-431-55909-2_7

http://dx.doi.org/10.1007/978-4-431-55909-2_7

10.3 Fiscal Decentralization and Regional Policies 203

Fig. 10.5 Structure of intergovernmental fiscal relationship. Source: Based on data provided bythe Ministry of Finance (http://www.mof.go.jp/budget/fiscal_condition/related_data/sy014_2409.pdf) in Japanese

may be necessary in some regional polices if we consider selfish behavior of localgovernments to make their own regions sustainable.

10.3.1 Intergovernmental Relationship

In Japan, the total expenditure of the local governments now exceeds that of thenational government although their total tax revenue is smaller than that of thenational government (c.f. Fig. 10.5). It means that there is an intergovernmentaltransfer system between the national government and the local governments.7 Weexamine the structure of the system because it affects the behavior of the local gov-ernments and in turn leads to regional disparity in Japan.

10.3.1.1 Local Governments

In Japan, there are 47 prefectures and 1718 municipal governments as of 2016. Twolevels of the local governments, prefectural level and municipal level, have similarstructure with a direct relationship with the national government, as depicted in Fig.10.1. The number of municipal governments was 15,859 in 1889 and 3100 in 2007.The national government has been trying to reduce this number by encouraging themerger of small cities and towns so that eachmunicipal government can have a higher

7See Akai (2015) for more information and discussion on the Japanese local government system.

http://www.mof.go.jp/budget/fiscal_condition/related_data/sy014_2409.pdf

http://www.mof.go.jp/budget/fiscal_condition/related_data/sy014_2409.pdf


management capability to play larger roles given in the process of decentralizationto municipal governments (c.f. Sect. 10.3.2).

In Japan, the national government has had strong powers in determining policiesand local governments had been implementing policies it determined.However,whenthe economic growth slowed down, there occurred certain frustrations among localgovernments that they want to have more powers to determine their policies in thespirits of “local self-governments” defined in the Constitution. Hence, the processof the fiscal decentralization in Japan has been the process in which the nationalgovernment provides more powers to local governments, especially to municipalgovernments, so that they can play larger roles in designing and implanting policies.

The structure of intergovernmental fiscal relationship is illustrated in Fig. 10.5.As the figure indicates, more revenues are collected at the national level than at thelocal level, but the total expenditure at the local level is larger than that at the nationallevel. Hence, in terms of public expenditure, the local governments play larger roles.However, the national government has played an important role in allocating moneyto local governments, and thus has the power to control the local governments.

There are basically two ways for the national government to transfer its revenueto the local governments. One way is to provide fixed-rate subsidies earmarked oncertain expenditure by the local governments. Another way is to use the so-calledlocal allocation tax, which is explained below.

10.3.1.2 Local Allocation Tax

The local allocation tax is not an ordinary tax. It is simply the name of the revenueallocated to the local governments. The revenue is collected from five national taxesat fixed rates: 33.1% of individual income tax, 33.1% of corporate income tax, 22.3%of consumption tax, 50% of liquor tax, and 100% of local corporate tax. Each of themis collected by the national government.

The revenue is then allocated to the local governments as follows. The nationalgovernment first calculates the levels of expected expenditure (E) and revenue (R)of each local government.8 If the revenue of the local government is larger than itsexpenditure, that is, E − R ≤ 0, then the national government does not allocate anyrevenue to it. If E − R > 0, then the national government tries to cover the gap byproviding local allocation tax as a general-purpose subsidy to the local government.

As the expected expenditure and revenues are both calculated theoretically, sucha subsidy will not discourage the efforts to collect tax revenue, unlike the case inwhich actual revenues are used in the calculation. However, it does discourage localgovernments from maintaining and expanding the local tax base. For example, theexpected revenue is calculated based on the number of tax payers and the corporationsin each region; hence, even if their numbers are declining, the local governmentsneed not worry too much about it because the local allocation tax will be increased

8The expected expenditure is called “standard fiscal needs” and the expected revenue is called“standard fiscal revenue.”


Fig. 10.6 Revenue structure of local governments. Source: Ministry of internal affairs and commu-nications (http://www.soumu.go.jp/iken/zaisei/28data/chihouzaisei_2016_en.pdf). Notes: Townsand villages are referred to as “big towns and billages” if they have population of 10,00 and moreand as “small towns and villages” otherwise

in response to the decline in the expected tax revenue. As the local allocation taxincreaseswith the increase in the number of elderly people, which is one of the factorsconsidered in the calculationof the expected expenditure, the local governments againdo not have to worry about it much.

Hence, the local allocation tax provides insurance to the local governments againstthe fall in tax revenues as well as rise in expenditures in the long-run. As the localallocation tax is financed by the national tax revenue, it can be viewed as a redistrib-ution mechanism among the local governments of Japan. Although it is a reasonableredistributive system, it can inducemoral hazard behavior on the part of local govern-ments. The local governments need not make efforts, say, to retain young taxpayersfrom leaving their towns or to finance the increasing expenditures on the elderlypeople. Hence, it worked as a mechanism to accelerate depopulation and aging inrural areas.

Figure 10.6 indicates that small towns, mostly those that are located in rural areas,are heavily dependent on the local allocation tax, which can be understood as theconsequences of the redistributive system. The structure is very similar to the publicassistance program, which discouraged poor people from working hard to becomeindependent (c.f. Sect. 9.3.1).

To mitigate such a moral hazard problem, the local allocation tax system is nowmodified. The amount of local allocation tax is now set equal to E − 0.7 × R if itwere positive. Under such a rule, if a local government makes efforts to raise theexpected revenue (R), the local allocation tax will be reduced only by 70%, not by100%, and there will be a 30% increase in local revenue. It provides some incentiveto make efforts. However, the effect is likely to be small, because it is similar to a70% income tax rate, which strongly discourages people from working.

http://www.soumu.go.jp/iken/zaisei/28data/chihouzaisei_2016_en.pdf

http://dx.doi.org/10.1007/978-4-431-55909-2_9


10.3.2 Regional Policies

We argued that the Japanese local allocation tax system has probably acceleratedthe depopulation process in rural areas. Furthermore, in Japan, the generous socialsecurity system described in the last two chapters also played a role in acceleratingthe depopulation process as the government now provides care to the elderly people,thus allowing the young people in rural areas to leave their parents and migrate tourban areas.

We have argued that the attachment to hometowns is one of the factors that preventpeople from leaving rural areas. One of the important sources of attachment, whichis well-documented in the literature, is the existence of families and friends in theirhometowns (e.g., Dahl and Sorenson 2010).When social security was not developed,mutual aid within families and communities was very important. Hence, for manypeople, there were certain risks involved in leaving their hometowns, which workedto prevent people fromgoing out their hometowns. However, with the development ofsocial security such risks were reduced, thus weakening the attachment to hometowsand increasing the mobility of people.

The higher mobility had a positive effect on the economic development of Japanin the short-run by mitigating the labor shortage in urban areas and helped themto continue growing. However, from the perspective of efficiency it had negativeeffect by congesting the urban areas and leading to the under-use of resources inthe rural areas. As we have already argued in Sect. 10.2.2, population distributionhas externalities that cannot be well internalized in the markets. Hence, free mobilitydoes not guarantee the efficiency of society in general. This justifies the government’sactive regional policies.

Recognizing this problem, the national government has started implementingactive regional policies since the 1970s. As the problem involves nation-wide pop-ulation mobility, the national government should take the responsibility of regionalpolicies. However, population mobility is highly affected by the system and thebehavior of local governments. In the following, we first provide an overview of thenational government’s regional policies and discuss how they should be formulatedfor reducing regional disparities. Then, we consider the local government’s systemfor making the Japanese society sustainable.

10.3.2.1 Strategies of the National Government

The national government has tried to mitigate regional disparities, especially in the1970s when the big urban areas witnessed rapid economic growth and faced variousproblems of the overpopulation. It increased public investment to less developedareas (c.f. Sect. 1.3.1). Due to these efforts, several regional cities have grown to bethe centers of different regions. These centers are referred to as regional urban cities.

As the national government started to accumulate a public debt it was criticized forspending a lot of money on public investment. The criticisms were convincing as the

http://dx.doi.org/10.1007/978-4-431-55909-2_1


rural economies were heavily dependent on the public investment for job creation.For employment in non-urban areas, public investment had to be maintained forever.Such a policy did not seem to be healthy at all.

The expenditure on public investment was gradually reduced. The national gov-ernment formulated and implemented various policies to revitalize the rural areas.However, due to the generous redistributive fiscal system such as the local alloca-tion tax and social security, the incentives for the local governments as well as thepeople to revitalize their respective regions were weak. Most of the regional policiesformulated by the national government did not produce the desired effects.

Due to such failure in regional policies, there seems to be a public sentiment that itis impossible to revitalize rural areas, and thus spending money on regional policiesis useless. The population mobility from rural areas to big urban centers, especiallyTokyo and its neighboring areas, is still continuing (c.f. Fig. 1.10). Although thenational government has been trying to revitalize the rural areas as summarizednicely in OECD (2016), it seems that people in the extremely rural areas are givingup on the sustainability of their hometowns.

The fundamental problem of the revitalization policies is that they require effortsof the local people to revitalize their areas. As suggested by the theory of voluntarypublic goods provision (c.f. Sect. 6.2.2), efforts to revitalize rural areas will not bemade without the right incentives. As long as the national government providesresources for the revitalization policies, they may seem to be successful. However,the superficial success will soon become a failure once the budget for these policiesruns out, just like other Japanese regional policies that have failed since the 1970s.

Majority of the people in rural areas are elderly people who have little incentiveto revitalize the areas because they are living well under the generous social securityand fiscal transfer systems, and their children are living well in urban areas. Theirsentiments seem to be described by the phrase well-known in Japan: “even the pros-perous inevitably decay.” As long as the lives of the people are guaranteed until theirnatural death, it is fine if their hometowns decay. If this process continues, most of therural areas will vanish and people will move to the big urban cities where the fertilityrates are low; eventually leading to the disappearance of the Japanese society.9

Is there a sustainable regional policy? The only hope seems to lie in the redesignof the redistribution system to provide adequate incentives to achieve the desirableland use. As population decline is inevitable in Japan, unless we welcome a verylarge number of foreigners, it would be unreasonable to retain all the rural areas.Hence, Japan should give up its extreme rural areas. Further, it is not a good idea toconcentrate people in a few big urban cities, considering the risks of natural disasters(e.g., earthquakes) and the extremely low fertility rates in these areas.

Hence, the goal is to change the fiscal systems and policies to lead people to livein regional core cities with good infrastructure, which has gradually been developedthrough public investment since the 1970s. They are good places for young people tolive, work, and form families. They are also good places for the elderly people to have

9For example, the total fertility rate in Tokyo in 2015 was estimated at 1.17, which is far below thenational average 1.46, and the population sustaining level at 2.07.

http://dx.doi.org/10.1007/978-4-431-55909-2_1

http://dx.doi.org/10.1007/978-4-431-55909-2_6


comfortable lives. The inflow of the elderly people from the extreme rural areas willcreate new jobs for the young people in the service sectors dedicated to the elderlypeople in the regional core cities in each prefecture. Unlike public investment, jobsin the service sectors for the elderly are secure as the demand for services for theelderly will continue to grow. Such job security will provide good circumstances forthe young couples to raise children.

The basic idea here is that, instead of letting the young people move to the bigurban areas, such as Tokyo, Osaka, and Nagoya, leading young and elderly peopletoward regional core cities is a better strategy. Such a policy will stop the inflow ofJapanese people to the urban “black holes” and maintain the diversity of cities tomake the Japanese society sustainable. To attract young and elderly people to theseregional core cities, the government should improve their conditions. Elderly peoplewill be attracted to these cities if there are good houses with adequate accessibility tomedical and care services. Young people will find them attractive if they offer goodeducation, jobs, and nursery services to raise children.

Therefore, the national government should not just revitalize these areas, butenrich them with well-designed social system to attract young and elderly peoplealike. What is important for the sustainability of the region is not encouraging thecontinuous efforts of the people in the “revitalization policy,” but achieving a naturalbalance of the region by creating circumstances in which self-interested behavior ofthe people (e.g., to have good life) can natually result in continuous activities (e.g.,trades of goods and services including labor, in markets, families, and communities)in the region, which can be referred to as the “wellness policy” to achieve healthyconditions of the region.10

10.3.2.2 Reforms of the Local Government System

To find a good balance for the sustainability of rural regions under the limited bud-get of the national government, a society should select the areas that can be madesustainable at low costs. To make a wise selection, good information about variousconditions in each area is indispensable. Hence, the selection and investment shouldbe made by the local governments and not the national government. The municipalgovernments are too small to make such decisions. The prefectural governmentsshould play an important role in such decisions. However, the areas that the prefec-tural governments can cover may still be too small to make good decisions on theselections of the sustainable areas.

In Japan, there has been a discussion about integrating the 47 prefectures to abouta dozen “states” (c.f. Fig. 10.7). Such an administrative reform of the local govern-ment system can reduce the differences among the prefectural governments. It willallow the national government to downsize the redistributive system, which is themain source of the moral hazard problem. If the municipal governments have smaller

10If the healthy conditions are achieved in the rigion, even the small towns and villages in extremelyrural areas will be sustainable.


Fig. 10.7 Administrative reform of local governments. Source: Wikipedia (https://upload.wikimedia.org/wikipedia/commons/9/9e/Region_system11.png)

roles and the states have greater roles, the necessity for redistribution can be furtherreduced. The big urban cities and several regional urban cities in Japan roughly coin-cide with the centers of a dozen states. If this administrative reform is implemented,regional policies to strengthen the regional urban cities can be implemented moreeasily because such policies will not harm the states but rather strengthen them.11

In Japan, decentralization has endowed the municipal governments with morepowers. For the municipal governments to bear greater roles, larger redistributionsare necessary. To mitigate this problem, the national government has encouragedmergers of municipal governments so that they can play greater roles. Now, thereare only about 1700 municipal governments. An inevitable consequence of such areform is that the local governments are not really close to the people any more.

11Such regional policies to develop the regional urban city will face difficulty when there are 47prefectures because only prefectures with regional urban cities will be strengthened and thus otherprefectures will go against such policies.

https://upload.wikimedia.org/wikipedia/commons/9/9e/Region_system11.png

https://upload.wikimedia.org/wikipedia/commons/9/9e/Region_system11.png


The decentralization could have been done to give more powers to the prefecturallevel governments in Japan. Fiscal differences among prefectures are much smallerthan those amongmunicipalities, and thus the need for redistribution will be reduced.The mergers of the prefectures into a dozen states to reduce fiscal differences furtherwill be even better. In the new system, the redistribution among local governmentscan be reduced. If the prefectures or states play larger roles in providing basic publicgoods, roles of municipal government will be reduced. Each municipal governmentshould be given the power to determine its expenditure within the revenue that itcollects by itself.

The argument for promoting decentralization in Japan is akin to saying that morepower should be given to children who are heavily dependent on their parents’income. This type of decentralization is not healthy because their children will con-tinue to be dependent and try to overuse the family income. If powers of makingchoices are given to children who earn their own revenues, those who want to makebetter choices will work hard to expand their revenue. Decentralization should bedesigned considering the rational behavior of the local governments to mitigate themoral hazard problems.


Regional disparities have expanded in Japan and its population continues to flow intourban areas, especially the Tokyo region. There are a few serious concerns about thegrowing regional disparities. First, with rapidly shrinking population and aging inrural areas, the fiscal sustainability of rural local governments is being questioned.Second, the Japanese are absorbed into the metropolitan area where the total fertilityrates are low and big earthquakes are expected to occur in the near future (c.f.Preface). If this trend continues, the Japanese economy and society will face seriousdifficulties, especially when a big earthquake hits the center of the Tokyo region.

The inflow of population from rural areas to urban areas has been observed sincethe Japanese economy started to recover from the Second World War. To mitigatedepopulation in the rural areas and overpopulation in urban areas, the Japanese gov-ernment started to enrich the infrastructure in rural areas to prevent population out-flow. Such regional policies worked well in the 1970s to mitigate the inflow of thepeople into urban areas (c.f. Fig. 1.10).When basic infrastructures were developed inrural areas and public investments in these areas were slowed down, the populationstarted to flow into Tokyo area again in the 1980s where the property bubble wasgrowing. After bubble burst in 1989, there was a net outflow of population from theTokyo region. However, as the urban economies recovered and started to reap thebenefits of globalization in the 1990s, the population flowed back into the Tokyoregion again.

Despite the depopulation in rural areas, the local governments have maintainedfiscal balance because there have been generous intergovernmental transfers fromthe national government to the local governments. The system can be compared

http://dx.doi.org/10.1007/978-4-431-55909-2_1


to the Japanese public assistance program for the growing number of poor people,which we discussed in Chap.9. The fiscal transfer system has helped the rural localgovernments, but they have become dependent on the redistributive system, whichhas discouraged their efforts to be independent.

Such a protection system created a moral hazard problem. Just like the publicassistance program possibly increasing poverty by keeping poor people in a welfaretrap, the system may have trapped the rural governments in poverty and worsenedthe depopulation problem in these areas, which can be referred to as “depopulationtrap”.

Tomitigate the regional disparities, enhancing the redistributive protection systemis not awise policy. The national government should enhance the policies to empowerlocal governments so that they can maintain the number of local residents. Suchpolicies have recently been implemented to revitalize the rural areas. Although theOECD (2016), for instance, has indicated some optimism about the success of suchpolicies, we have deep doubts about it.

For such a revitalization policy to work, people in the rural areas must put in alot of effort. However, due to the generous redistributive policies, Japan’s rural areas“outperform most of their OECD peers” with regard to the per-capita GDP and thepeople there do not have the incentives to make efforts for survival. Ironically, due tothe relatively good conditions of rural areas in Japan,we believe that the revitalizationstrategies are not likely to work well. As the room for the national government toprovide generous subsidies to these areas shrinks, rural areas will fade out. Youngpeople in rural areas are fleeing into urban areas, which will result in rural areasdisappearing with only a few surviving urban areas.

After facing a risk of decay, some regions may make real efforts for survival; butpeople in most rural areas will give up on the revitalization with the belief “even theprosperous inevitably decay.” The worst possible outcome of the flow of the peoplefrom rural to urban areas is the concentration of the Japanese in Tokyo area, knownas the Japan’s “black hole” (c.f. Masuda 2014), where fertility rates are very low anda big earthquake is expected to occur in the near future. The only realistic policyto avoid this is not to revitalize rural areas but to diversify the flow of people intoseveral regional core cities in each prefecture and several regional urban cities where,for instance, major national universities and relatively big companies are located. Byinvesting in infrastructure and higher education, the regional urban cities will be ableto attract people, especially young talented people, from the neighboring areas, thuspreventing them from flowing into big urban areas.

When the population in Japan is declining, “selection and concentration” is neces-sary and useful in regional development policy. However, just like ordinary businessdecisions, excessive concentration can be risky. Selecting several regional urbancities and making intensive investment in those areas seem to be a better decision.Some rural areas will find ways to utilize regional resources for their sustainability.However, many of them will inevitably decay. The government will have to providepeople in these areas with minimum support or support to change their residences.

Such a regional restructuring will proceed slowly and quietly. What seems impor-tant for policymakers is not to use resources to prevent natural death of some regions

http://dx.doi.org/10.1007/978-4-431-55909-2_9


in an attempt to revitalize them but to use resources to empower other regions care-fully so that the entire body, that is, Japanese society can live a healthy life for a longtime.

As regards the revitalization policies in Japan, OECD (2016, p. 3) states that“[m]any of these are likely to be of great interest and relevance to the many OECDcountries facing similar challenges.” As these policies may end up in a waste ofresources, it may be better for other countries to wait and see if they really work inJapan.

References

Akai, N. (2015). Fiscal consolidation and local public finances in Japan: Incentive problems asso-ciated with intergovernmental transfers and their political roles. In T. Ihori & K. Terai (Eds.), Thepolitical economy of fiscal consolidation in Japan. (Chapter 4), Springer.

Council of Local Authorities for International Relations. (2015). Local government in Japan 2012(2015 Rev. Ed.). http://www.clair.or.jp/j/forum/pub/jichi-en_3.pdf.

Dahl, M. S., & Sorenson, O. (2010). The social attachment to place. Social Forces, 89(2), 633–658.Harris, J., & Todaro,M. (1970).Migration, unemployment and development: A two-sector analysis.American Economic Review, 60(1), 126–42.

Kawano,M., Yoshitake, T., Tatsumi, H., &Kajita, Y. (2004). Analysis of residential choice behaviorat community scale. In Proceedings of European regional science association. http://www-sre.wu-wien.ac.at/ersa/ersaconfs/ersa04/PDF/40.pdf.

Krugman, P. (1991). Increasing returns and economic geography. Journal of Political Economy, 99,483–499.

Lee, Y.-J., Sugiura, H., & Geciené, I. (2016). Stay or relocate: The roles of networks after theGreat East Japan earthquake. In E. Jones & A. J. Faas (Eds.), Social network analysis of disasterresponse, recovery, and adaptation. (Chapter 15) Butterworth-Heinemann.

Mansoorian, A., & Myers, G. (1993). Attachment to home and efficient purchases of population ina fiscal externality economy. Journal of Public Economics, 52, 117–132.

Masuda, H. (2014). Disappearing local regions: Rapid population decline caused by MonopolarConcentration to Tokyo. Chuo Kouon Shinsha. (in Japanese).

Ministry of Internal Affairs and Communications. (2015). Present conditions of policies for depop-ulating areas. (in Japanese).

OECD. (1998). Harmful tax competition: An emerging global issue. OECD Publishing.OECD. (2016). OECD territorial reviews.Shimada, G. (2015). The role of social capital after disasters: An empirical study of Japan basedon time-series-cross-section (TSCS) data from 1981 to 2012. International Journal of DisasterRisk Reduction, 14, 388–394.

Steiner, K. (1965). Local government in Japan. Stanford, Calif.: Stanford University Press.Tiebout, C. (1956). A pure theory of local expenditure. Journal of Political Economy, 64, 416–424.Wilson, J. D. (1999). Theories of tax competition. National Tax Journal, 52(2), 269–304.

http://www.clair.or.jp/j/forum/pub/jichi-en_3.pdf

http://www-sre.wu-wien.ac.at/ersa/ersaconfs/ersa04/PDF/40.pdf

http://www-sre.wu-wien.ac.at/ersa/ersaconfs/ersa04/PDF/40.pdf

Chapter 11Epilogue: Social Transformation and PublicPolicies

Such prophylactic policy must naturally and necessarily bedirected at the family and at the children who are the people ofthe future. Such a policy is in a high degree an “investment,” inthe personal capital of the country.

Myrdal, G. (1940, p. 206)

11.1 Social Transformation, Public Policies, and SocialNorms

Families, and communities in general, change as the society surrounding themchanges, causing a social transformation that further transforms them.

In Japan, the rapid development of the market economy after the Second WorldWar freed the Japanese people from the obligation of being a part of families andlocal communities, thereby, changing the structure and functions of these institutionsin society. Such changes demanded the expansion of the social security system by thegovernment. When the Japanese government responded to the voices of its peopleand expanded the social security system in the 1960s and early 1970s, it broughtabout a total change in the nature of families and communities.

We attempted to explain this transformation of the Japanese society (Chap. 1)based on various sociological studies (Chap. 2) and economic analyses (Chaps. 3–7),and applied the argument to understand andfind solutions to three serious social prob-lems in Japan (Chaps. 8–10). The three problems we picked up are population crisis,increasing poverty, and increasing regional disparities. These are serious problemsas they indicate the unsustainability of the Japanese society. In particular, the rapidpopulation decline with population aging, which is depicted in Fig. 1.1, will threatennot only the long-run sustainability of the Japanese society, but also its short-runsustainability when we consider the huge public debt accumulated by the Japanesegovernment (Fig. 1.3).

Although we are trying to explain the relationship between social transformationand public policies based on the Japanese experience, economic theories and socio-


213

http://dx.doi.org/10.1007/978-4-431-55909-2_1

http://dx.doi.org/10.1007/978-4-431-55909-2_2

http://dx.doi.org/10.1007/978-4-431-55909-2_3

http://dx.doi.org/10.1007/978-4-431-55909-2_7

http://dx.doi.org/10.1007/978-4-431-55909-2_8

http://dx.doi.org/10.1007/978-4-431-55909-2_10

http://dx.doi.org/10.1007/978-4-431-55909-2_1

http://dx.doi.org/10.1007/978-4-431-55909-2_1

214 11 Epilogue: Social Transformation and Public Policies

logical studies suggest that such an experience can occur in any democratic society.Most of the developed countries indeed have experienced low fertility, increasingpoverty, and depopulation in rural areas. Therefore, these experiences suggest thatproblems can occur in many developing countries in the near future if the govern-ments expanded the social security systemwithout careful consideration of its impacton families and local communities.

However, there are some differences between the Japanese experience and those inother developed countries. As suggested by various figures in this book, the Japanesegovernment is still trying to rely heavily on mutual aids within families while for-mulating public policies (c.f. Figs. 2.2, 8.6, and 8.7). In particular, little support hasbeen provided to families for raising children, while the support for the elderly hasbeen quite generous. (c.f. Figs. 8.4 and 8.9)

In this epilogue, to explain such characteristics of Japanese policies, we touch onthe close relationship between public policies and the social norms. Public policieschosen by the people are greatly affected by the existing social norms, which arein turn affected by the public policies. This interactive and cyclical relationshipbetween public policies and social norms seems to have molded the path of socialtransformation in Japan, making the sustainability problem of Japan more seriousthan that of other developed countries.

In Sect. 11.2 below, we try to identify the factors that led to this situation in Japan.Then, in Sect. 11.3, we draw lessons for developing countries that may face similarproblems in the near future. In the final section, to discuss social policies in the 21stcentury, we look at the future where further globalization and technological progressare expected.

11.2 What Went Wrong in Japan?

Japan is facingmany problems now because of its generous redistributive systems forits citizens aswell as different regions. These systemswere introduced to protect thosewho were suffering from poor living conditions. However, they produced multipleside effects that can be described as themoral hazard problems or free rider problems.The side effects should have beenmitigated by adequate public policies.However, thegovernment failed to do so, and as such, was confronted by more serious problems.

For example, the pay-as-you-go pension system in Japan is a system that redistrib-utes income from families with children who support their parents to those withoutthem. In the absence of such a system, most elderly people need to have childrenand rely on them during their life after retirement. On the other hand, if there isa generous social security system, they need not make efforts to raise children ormaintain a good relationship with them. The consequences are lower fertility ratesand separation of the elderly people from the young people, which has three furtherimpacts.

First, nursery services provided by grandparents are no longer available for youngcouples. Hence, the young couples need to purchase nursery services if they keep

http://dx.doi.org/10.1007/978-4-431-55909-2_2

http://dx.doi.org/10.1007/978-4-431-55909-2_8

http://dx.doi.org/10.1007/978-4-431-55909-2_8

http://dx.doi.org/10.1007/978-4-431-55909-2_8

http://dx.doi.org/10.1007/978-4-431-55909-2_8

11.2 What Went Wrong in Japan? 215

0% 20% 40% 60% 80% 100%

1998

2003

2008

Strongly Agree Somewhat Agree

Somewhat Disagree Strongly Disagree

0% 20% 40% 60% 80% 100%

1993

1998

2003

2008

Fig. 11.1 Changing Social Norms of Families. Source: Nishioka et al. (2010, Data in Fig. 6-1)

working after having children. However, nursery services have not been well pro-vided, thus leading to further decline in fertility.

Second, the independence of elderly people contributes to the increase in povertyamong them. Despite elderly people believing they can live independently of theirchildren, life is uncertain and the risk of falling into poverty remains. If parents andchildren live together, these risks can be minimized. With the separation of families,elderly parents fall into poverty more easily.1

Finally, the migration of the young people from towns to urban cities for work hasincreased parents’ risk of falling into poverty and induced the depopulation problemin rural areas. Now, young people feel less guilty of leaving their parents at home inrural areas because, as indicated by Fig. 11.1, more people have come to believe thatit is not the responsibility of children to provide income support and care for elderlyparents, but something that should be provided by the social security system. If thesocial security system were not generous, elderly parents would have tried to keeptheir children in their hometowns as they would not make an adequate living afterretirement otherwise.

The social security system freed young people, especially women from takingcare of their elderly parents. It has probably contributed to the economic growthof urban areas. However, its side effects have presented themselves in the form ofdeclining fertility, increasing poverty among elderly people, and expanding regionaldisparities. Having failed to recognize these side effects and formulate adequatepolicies to mitigate them are the reasons for Japan facing such serious problems.Now, the question arises, why has the Japanese government failed in this respect?

First, the Japanese government has a greater preference for policies that the peoplewant than social policies to mitigate long-run problems. The expansion of the socialsecurity system and achieving high economic growth have been popular public poli-cies. The government implemented these policies without taking their adverse sideeffects into serious consideration. To mitigate the adverse side effects, the Japanesegovernment should have increased social expenditure on families to support them

1It has been pointed out that child poverty is also increasing in Japan. See Sect. 9.2.2.

http://dx.doi.org/10.1007/978-4-431-55909-2_9


0% 20% 40% 60% 80% 100%

Japan 2014

Sweden

France

USA

Korea

Japan Agree

Somewhat Agree

Somewhat Disagree

Disagree

Don't Know

"Husband should work outside home and wife should stay at home."

Fig. 11.2 Social norms of sexual division of labor. Source: Cabinet Office of Japan (2006, 2014)

in raising their children (c.f. Sect. 8.2.2.2). Such policies also have the potential ofbeing popular, but has not been implemented by the government in Japan.

One of the reasons is the social norm that “husband should work outside thehome and wife should stay at home.” Figure 11.2 shows that the ratio of those whoanswered “agree” or “somewhat agree” with the norm exceeds 50% in Japan and ishigher than those of other four countries in 2004.2

This norm has sustained in Japan for a long time probably because it has beenpolitically favored by leaders of the Liberal Democratic Party (LDP), which has ba-sically been the ruling party from 1955 till date except for a short period in between.Rosenbluth (2007, p. 12) points out that “early party platforms of the LDP statedexplicitly that women should stay at home to help their husbands be good work-ers.” Rosenbluth (2007, p. 212) also states the following: “In 2004, a special LDPcommittee on constitutional revision denounced the 1947 constitution’s “women’srights” clauses for nurturing “egoism in postwar Japan, leading to the collapse offamily and community.”

Under this prevailing social norm, the Japanese government did not expand child-care services very much. Such a public policy contributed to the sustainability of thesocial nom, because following the social norm becomes the right choice for coupleswho want to have children. If the government expands nursery services, women neednot stay at home because they can retain their jobs and raise children (c.f. Fig. 8.10).

Even if the Japanese government did not want to increase the childcare services, itcould have expanded the child allowances to encourage young couples to have morechildren. Although such a policy is in agreement with the prevailing social norm,the government has not expanded child allowances much. As an explanation for thisattitude of the government, it has been claimed that providing incentives for havingmore children reminds people of the government’s policy to increase the populationof children and young people during the wartime.3

2The ratio was 60.1% in 1992. As Fig. 11.2 indicates, it is 44.6% in 2014. The social norm has beenchanging gradually but speed of the change is low.3See Takeda (2005) for comprehensive studies on reproductive and population policies of Japan.

http://dx.doi.org/10.1007/978-4-431-55909-2_8

http://dx.doi.org/10.1007/978-4-431-55909-2_8


This may be only a partial explanation for the government’s policy of not expand-ing social expenditure on raising children. The real reasonmay lie in the government’sunwillingness to spend money for supporting child-rearing because such a policy isin conflict with the ideology of the pro-business LDP as it “would have requiredthe government to pay potentially huge amounts of money for childcare support,which violated the party’s pledge to its business supporters of keeping governmentexpenditures to a minimum” (Rosenbluth 2007, p. 12).

Garon (1997) convincingly argued that the Japanese government has beenmoldingminds of the people by actively involving in their daily life and by not providingsupport to individuals so that their families and communities provide it to avoidlarge public expenditure. Even though it may look like a laissez-faire policy, Garon(1997, p. 38) argues that “‘laissez-faire’ would have been the last word to describe themany instances in which the state intervened to force families and neighbors to aidthe poor.4” The Japanese government’s not increasing the expenditure for childcareservices and child allowances but providing subsidies for housewives and youngcouples’ living with elderly parents may be seen as an example of such strategy ofthe Japanese government.

Japanese government started to implement expansionary population policies afteracknowledging the population crisis facing Japan in 1990 when the total fertility ratewas found to be 1.57 in 1989,5 the lowest after the war since 1966, the year of thefire horse (hinoeuma)6 (c.f. Fig. 1.4). Although the Japanese government has madeefforts in promoting gender equality and fertility since then, after 25years, the gendergap in Japan is still among the worst in the world7 and the total fertility rate is farbelow the replacement level. The efforts and expenditure have been too small tosucceed in overcoming shocks and crises (c.f. Fig. 8.4).

The slow response of the Japanese government was due to the government’sunwillingness or inability to consider child support policies as a political agenda.In countries where high in-kind benefits (e.g., childcare services) are provided tofamilies, such benefits have been increased not usually because the governmentswant to encourage women to work and have more children, but mainly because theyare trying to provide women with equal opportunities to work.8

4See also Yamashige (2002) for more discussion on the “Japanese-style welfare state” and relation-ship between the arguments by Garon (1997) and those in this book. The decline in the traditionalsocial norm among young people in the past twenty years is quite large. “Molding Japanese minds”by the traditional government’s strategy does not seem to be working well to make Japanese societysustainable.5The shock was called the “1.57 shock.”.6The year is known as the bad year for girls to be born, according to Chinese astrology.7Even today, there is a big gender gap in Japan. For example, in the “Global Gender Gap Index”published by the World Economic Forum, Japan ranks at 111th among 144 countries in 2016. TheJapanese government’s policies for gender equality have been too small to change the structure ofthe Japanese society and economy.8Rosenbluth (2007, p.19), for instance, observed the following: “Public provision of childcare,which is generous and nearly universal in Sweden, appears to have been more of a response to theneed of working mothers .... than a cause of their initial foray into the market.”.

http://dx.doi.org/10.1007/978-4-431-55909-2_1

http://dx.doi.org/10.1007/978-4-431-55909-2_8


Furthermore, the reason behind providing high cash benefits (e.g., child al-lowances) to families in some countries is usually not to increase the number ofchildren, but to give children good and equal opportunities to grow in safe andhealthy conditions. Hence, social expenditure on families, which have positive ef-fects on fertility (c.f. Figs. 8.4 and 8.10), have been increased in many countries (c.f.Fig. 8.9) because of the social norm (or value) of guaranteeing basic human rightsto children and women, and not because of the motivation to increase fertility.

Ifwomen’s political voicewas stronger in Japan, the social expenditure on familiesmight have been higher. The traditional social norm and women’s low economic andpolitical participation has prevented Japanese women from raising their demands.Hence, for a long time now, the social norm has contributed to the low level of socialexpenditure on families, and in turn, to the low fertility rate.9

The Japanese government has been silent and hesitant about implementing ef-fective public policies to mitigate the fertility and population problems. Economicpolicies have been engineered to achieve long-run economic growth in Japan, proba-bly because big businesses in Japan have historically supported the LDP. On the otherhand, the social policies have been reactive, protective, and consumptive in nature.10

The idea of “social engineering” to achieve a desirable society in the long-run hasbeen missing in Japanese policymaking.

As Garon (1997) clarified, Japanese government has been providing various sup-port to “traditional families (e.g., 3-generation families and single-earner families)”and traditional local communities (neighborhood associations) so that they can pro-vide supports to their members (c.f. Fig. 11.3). As the market economy develops,traditional families and communities have beenweakend; and the government startedto increase the direct support to individuals (e.g., public pension and long-term careservices), which further weakened them (c.f. Fig. 2.1 and Sect. 2.3 in Chap.2). Theredistributive fiscal system to provide support to individuals caused moral hazardproblems and free rider problems. Japanese society now face various social prob-lems such as declining fertility, increasing poverty, and regional disparities, becauseof such a social transformation.

To avoid and overcome these problems, the Japanese government should enrichsupport and empowerment policies for the “nuclear families (e.g., dual-earner fam-ilies and single-parent families)” which increased as the result of the weakening ofthe “traditional families”. It also should enrich support and empowerment policiesfor new communities (e.g., NPOs and social capital) to compensate for the loss ofpublic goods supplied by the traditional communities and the government with hugepublic debt.

Fig. 11.3 also suggests taht the government need to enrich policies to empowerindividuals to encourage their self-help so that they can mitigate various risks bythemselves. We have repeatedly argued that policies to provide prophylactic in-

9See, for example, Rosenbluth (2007, Chap.1) for this argument.10In countries with high social expenditure on families, women in politics and business seem tohave been the driving political force behind the enhancement of social policies, something that ismissing in Japan.

http://dx.doi.org/10.1007/978-4-431-55909-2_8

http://dx.doi.org/10.1007/978-4-431-55909-2_8

http://dx.doi.org/10.1007/978-4-431-55909-2_8

http://dx.doi.org/10.1007/978-4-431-55909-2_2

http://dx.doi.org/10.1007/978-4-431-55909-2_2


Individuals

Markets

Traditional Communities (Neighborhood Associations)

New Communities (NPOs, Social Capital)

Traditional Families (3-generation, Single-earner)

Nuclear Families (Dual-earner, Single-parent)support

empowerment

supportempowerment

supportempowerment

empowermentsupport

empowermentsupport

Government

Fig. 11.3 Social Transformation and Public Policies

kind benefits (e.g., education and daycare services) are more effective to empowerindividuals, families, and communities than policies to provide curative cash benefits(e.g., pension and subsidies) that can be easily consumptive.

11.3 Lessons for Developing Countries

There seem to be many lessons that other countries, especially the developing coun-tries can learn from the experiences of Japan. First, it is very important to recognizethat demography is a crucial factor for economic growth and social sustainability.For example, the demand for cars, houses, and high-tech products that often leadto economic growth is high among the young. Furthermore, technological progressin production and the provision of services are often due to the young generation.Hence, population aging can have negative effects on the growth of the domesticmarkets and economic growth.

The demography also affects social sustainability. When the total fertility rate isless than two, the population is doomed to decline forever. The consequence is theextinction of society. Even before extinction, the society will suffer from problemsassociated with population aging, especially if it has a good social security system.The decrease in the number of young people who pay taxes and social securitycontributions will result in the reduction of social security payment. One of theconsequences is the increase in poverty among the elderly people.

Hence, each country should pay close attention to its future demography to preventthe occurrence of serious social problems that Japan is now facing. In this book, weargued repeatedly that the social security systemhas deep impacts on the demographyby changing the fertility behavior and social norms regarding families (e.g., seeFig. 11.1). Hence, the design of the social security system is the key to a desirabledemographic pattern, and thus, for economic growth and social sustainability.

There seem to be two options for avoiding a population crisis. The first option isto set the social security level at a minimum so that people continue to depend ontheir families for making a living after retirement. The second option is to expand


the social security for the elderly people and simultaneously increase support forfamilies with children so that the fertility rate does not fall below two.

The Japanese government attempted to execute the first option by emphasizing theimportance of family ties. It provided some subsidies for young couples’ living withtheir elderly parents and women’s being the housewives. However, the level of socialsecurity continued to be generous, which made it irrelevant for the elderly peopleto live with their children or receive income support from them. It also changed theprevailing social norm of Japan. More Japanese people have come to believe that itis the government’s responsibility to provide support and care for the elderly people.The government had no choice but to expand the social security for the elderly,especially when the voters are aging.

The government should have chosen the second optionmuch earlier in the processof expanding the social security for the elderly. In other developed countrieswith goodsocial security systems, the government took such policies and succeeded in raisingthe fertility rate with some help from the increasing number of immigrants. TheJapanese government has not so far taken such policies to increase social expenditureon raising children as well as the number of foreign workers. The current problemis a result of such inadequate public policies of the Japanese government.

However, we may not be able to blame the politicians or bureaucrats of theJapanese government as it is the Japanese who did not choose public polices fora sustainable society. With the myopic preference of the ordinary people, the demo-cratic government may not be able to implement public policies that are desirable inthe long-run and may have to give up on making the society sustainable.

This political problem may be one of the most serious challenges for democraticsociety. However, there are some democratic countries that have succeeded in main-taining the population for instance. A good example of a country based on the firstoption is the U.S., while countries based on the second option include some Nordiccountries. The political decisions about public policies are determined by prefer-ences of the people influenced by the social norm on which public policies havegreat impacts.

The interactive relationship between public policies and prevailing social normssuggests that the path for public policies should be carefully designed so that intendedfuture policies are in agreement with the social norms that have been modified bypublic policies in the past. Although such public policies are not easy to formulate,specialists who are concerned with the sustainability of the society should do theirbest to find the optimal path of public policies.11

11The basic problem is that it is difficult to implement policies that entail short-run costs and long-run benefits. For example, environmental policies present such a problem. One way of solvingthis problem is to use public debt, which turns short-run costs into long-run costs. Another wayis to make the current generation to acknowledge the future benefits as benefits accruing to them.Changing the social norm to change people’s perspective via education, for instance, can be a publicpolicy. If it is difficult, then supporting and empowering those who have such views and politicalinfluences can be another public policy to make society sustainable.

11.4 Public Policies in the 21st Century 221

11.4 Public Policies in the 21st Century

In the face of rapid decline of the young working generation in Japan, it is sometimesargued that the decline is not a big problem because high-tech capital such as robotsand artificial intelligence (AI) will naturally substitute for the loss of workforcein Japan. One may even argue that Japan has made a right choice in an era whenhigh-tech capital replaces most human jobs as unemployment would have increasedotherwise. The argument above suggests that public policies to sustain the populationmay not be desirable in this century.

However, what the argument implies is that people of a country can shrink foreverand robots and AIs can dominate society. The rapid progress of AIs may force us toaccept such a doomed society, but if we still have a room to choose the future of oursociety, the better choice is to have a sustainable society in which the total fertilityrate is close to two for the population to remain constant. What are the desirablepublic policies to realize such a society in the 21st century?

Some optimistic people may argue that, when high-tech capital starts producingmost goods and services, people neednotwork asmuch as before anduse the availabletime and income to have children and enjoy family lives. Hence, most families willnaturally choose to have more than two children. Unfortunately, economics tells usthat it will not be our future (e.g., Sachs and Kotlikoff 2012).12 The problem is thatthe profits that the capital generates would be returned to the investors and not theworkers because there would be no investment otherwise. The wage of each workerwill be just equal to the value produced by her/him (i.e., value of themarginal productof labor). Hence, it will be just an illusion that workerswill earnwage incomewithouthaving to work much.

Others may argue that the government can impose heavy taxes (e.g., “robot tax”)on the capital and redistribute them to workers. Unfortunately, such a policy willnot work well either because if a country imposes heavy taxes on the capital, theninvestment will be made in other countries. Unless countries can engage in tax coor-dination to stop lowering the capital income tax, the capital income tax will go downand stay at a low level.

Therefore, the only way to guarantee adequate income to workers is to educatepeople so that they can utilize robots and AIs and work with them, especially infields where they cannot fully replace human workers. Each country should seek toexpand output rather than reducing the number of workers to produce the same levelof output. When the capital starts replacing human jobs, workers may demand thegovernment to protect them. Protective policies are not useful in the 21st century asthey lead to rampant poverty with little tax revenue from capital income. Thus, thebest policy seems to be “empowerment policies” (e.g., good education policies) thatenable people to work along with the high-tech capital.

12There is a growing number of studies on effects (e.g., employment, wages, and inequality) andcauses (e.g. population ageing and low fertility) of robotization. See, for example, Acemoglu andAutor (2010), Benzell (2015), Sachs (2015) and Abeliansky and Prettner (2017).


With the development of communication and transportation technologies, theeconomy will be further globalized in this century. The weakness of the Japaneseeconomy and its society in this era of globalization lies in its very low dependence onforeign workers (c.f. Fig. 8.1). One of the key public policies to mitigate the damagesof population decline should be a sound immigration policy.

Japan should attempt to increase the population of high-skilled foreign workers toproduce and utilize the high-tech capital and other foreign workers to produce goodsand services that those high-tech capital fail to provide at low costs. If Japan startsincreasing the number of foreign workers more actively and carefully, future of theJapanese economy and society will become brighter.

We believe that whether Japan can change its immigration policy and decide tobuild a new society is the key to its sustainability. The foreign peoplewill alsomitigatethe population decline by increasing the rate of marriages and child birth in Japan.It is observed in most developed countries that increasing immigrants usually havepositive effects on fertility, especially when they come from developing countrieswith higher fertility rates.13 They also bring new culture and contribute to makinginnovations in economic and social activities.

Although foreign workers can bring benefits, they will require the society to incursome costs.14 Differences in languages, cultures, and social customs, can be thesources of conflicts between foreigners and the native people. If the society wantsto minimize such conflicts, it needs to incur the costs of assimilation policies. If itfails to implement good assimilation policies, it will suffer from an increase in theconflicts as its social cost.

The Japanese society has been a homogenous society for a long time and has beenapprehensive about increasing the number of foreignworkers, because people believethat the costs will be higher than the benefits. The perception that Japan should notincrease the number foreigners has almost taken the form of a social norm, and assuch, the government has not touched upon policies related to it.

With the population declining fast in Japan and its economy being globalizedfurther, the benefits of increasing foreign workers seem to have increased beyond itscosts. Experiences of other countries suggest that the assimilation policy is the keyto minimize the costs of increasing the number of foreign workers. For foreigners, itis well known that the network of the people especially from their home countries,that is, social capital is important for making adequate life and thus for assimilatinginto new society (e.g., Zetter et al. (2006) and Bankston (2014)). If the governmentinvests well in foreign workers and their social capital to reduce the differencesin languages, cultures, and social customs, the costs will be minimized, while thebenefits are maximized.

13See, for example, Adserà and Ferrer (2014) for a comprehensive surve on the demographic effectsof immigrants.14There is an accumulation of researches on the benefits and costs of immigrants. See, for example,West (2011) for a general survey. In relation to the public policies, fiscal impacts of increasingimmigrants have been studied well. In general, the fiscal impacts are estimated to be relativelysmall. See, for example, Rowthorn (2008).

http://dx.doi.org/10.1007/978-4-431-55909-2_8

11.4 Public Policies in the 21st Century 223

The quote in the beginning of this epilogue, written by Gunnar Myrdal morethan seventy years ago, still seems to be valid for any society trying to achievesustainability. Even in the 21st century, it is still a good policy to invest in the“personal capital,” which we now call human capital, so that families and childrencan survive the highly globalized competitive world.

The development of communication technologies will change the nature of fam-ilies and communities, while new medical technologies will change the nature ofbirths and deaths of human beings. Globalization, especially the higher mobility ofthe people across the world, will change the composition of people in the societyand lead to new challenges in families and communities. Public policies in the 21stcentury will be required to respond adequately to technological development andglobalization and make the best use of them, considering their effects on families,communities, and demography.

Japan is still struggling to recover from the damages of the Great East JapanEarthquake and the resulting nuclear melt down, while two big earthquakes areexpected occur in urban regions with 70% probability within the next 30years. Itspublic debt has exceeded 200% of GDP, the highest among the OECD countries.Population started to decline since 2008 and the elderly ratio is estimated to be around38% in 2050. Rural areas are suffering from serious depopulation and poverty hasbeen and will be increasing under the population ageing.

In facing these challenges, Japan needs to find good policies for minimizingdamages and recovering from the damages. We have argued that it is important toimplement policies to empower nuclear families and new communities in additionto those for traditional families and local communities. Based on the analysis ofthe migration behavior after big disasters in Japan, Shimada (2015) concluded that“this shows the necessity of putting social capital at the center of a recovery.” Publicpolicies to accumulate human capital and social capital seem to be effective not onlyin Japan but also in every society in the 21st century when robots and AIs will bespreading rapidly over the society.

Such a policymaking will not be easy. We need to develop economics of familiesand communities further to meet new challenges in the era of new technologies andglobalization. We believe that good public polices, as pointed out by Myrdal (1940),“must naturally and necessarily be directed at the family and at the children who arethe people of the future” for our society to be sustainable, even in the 21st century.

References

Abeliansky, A., & Prettner, K. (2017). Automation and demographic change. Discussion Papers,No. 310. Center for European, Governance and Economic Development Research

Acemoglu, D., & Autor, D. (2010). Skills, tasks and technologies: Implications for employmentand earnings. Handbook of Labor Economics, 4(B), 1043–1171.

Adserà, A., & Ferrer, A. (2014). Immigrants and demography: Marriage, divorce, and fertility. IZADiscussion Paper, No. 7982.

Bankston, C. L. (2014). Immigrant networks and social capital. Cambridge: Polity.


Benzell, S. G., L. J. Kotlikoff, G. LaGarda, & J. D. Sachs (2015) Robots are us: Some economicsof human replacement. NBER Working Paper No. 20941.

Cabinet Office of Japan. (2006). Attitude survey on the low fertility society in five countries. (inJapanese)

Cabinet Office of Japan (2014) Survey of public opinions. (in Japanese)Garon, S. (1997). Molding Japanese minds: The state in everyday life. Princeton: PrincetonUniversity Press.

Lee, Y-J., H. Sugiura, I. Geciené (2016) Stay or relocate: The roles of networks after the Great EastJapan earthquake. In E. Jones and A. J. Faas [eds.] Social network analysis of disaster response,recovery, and adaptation (Chapter 15). Oxford: Butterworth-Heinemann.

Myrdal, G. (1940). Population: A problem for democracy. Cambridge: Harvard University Press.Nishioka, H., Yamaguchi, M., Koyama, Y., Chitose, Y., Kamano, S., Suga, K., et al. (2010). Thefourth national survey on family in Japan 2008. Journal of Population Problems, 66–2, 48–75.(in Japanese).

Rosenbluth, F. M. (2007). The political economy of Japan’s low fertility. Stanford: StanfordUniversity Press.

Rowthorn, R. (2008). The fiscal impact of immigration on the advanced economies. Oxford Reviewof Economic Policy, 24(3), 560–580.

Sachs, J. D. & L. J. Kotlikoff (2012) Smart machines and long-term misery. NBERWorking Paper,8629.

Sachs, J. D., Benzell, S. G., & G. LaGardia (2015) Robots: Curse or blessing? A basic framework.NBER Working Paper 21091.

Shimada, G. (2015). The role of social capital after disasters: An empirical study of Japan basedon time-series-cross-section (TSCS) data from 1981 to 2012. International Journal of DisasterRisk Reduction, 14, 388–394.

Takeda, H. (2005). The political economy of reproduction in Japan: Between nation-state andeveryday life. RoutledgeCurzon: London; New York.

West, D. M. (2011). The costs and benefits of immigration. Political Science Quarterly, 126(3),427–443.

Yamashige, S. (2002) Families, communities and governments in improving socio-economic per-formances: Japanese experiences. World Bank Institute Working Paper # 37204.

Zetter, R., with D. Griffiths, N. Sigona, D. Flynn, T. Pasha, & R. Beynon (2006). Immigration,social cohesion and social capital: What are the links?. York: Joseph Rowntree Foundation.

Index

AAdverse selection problem, 26, 32, 54Altruism, 93–97, 101, 103, 104, 111, 122,

136, 139impure, 123

Artificial Intelligence (AI), 175, 221Association, 10, 11, 30, 116, 136, 146, 153Asymmetric information, 26, see also

incomplete informationAttachment,201, 206

BBackward induction, 58, 59Bargaining cost, 35, 86, 89Biological, 71, 72, 93, 111Black hole, 208, 211

CCapital, 175, 178, 221, see also social capi-

talhuman, 84, 223market, 104public, 10

Cash benefit, 170, 218Child allowance, 74, 77, 111, 165, 216Childcare, 9, 171, 172, 190, 216, 217Choice set, 28, 52Coase theorem, 35Commitment, 60, 99Commons, 39, 127, see also tragedy of the

commonsCommunity, 10, 30, 34, 36,37, 115, 117

definition, 7failure, 33local, 32, 126, 127new, 116, 136, 143, 146, 153

traditional, 116, 135Consumptive, 43, 160, 190, 197, 218, 219Cooperation, 12, 35, 37, 39, 43, 65, 115, 135

difficulty, 25, 34in repeated games, 61social capital, 144

Crowding-in, 123Crowding-out, 36, 119, 120, 122, 123, 180Curative, 43, 219

DDecentralization, 202, 209Democracy, 149, 213Democratic society, 29, 43, 136, 153, 214,

220impossibility, 30

Depopulation, 40, 200, 214, 215economic growth, 13fiscal transfer, 65, 196, 199, 205, 210

Derivative, 48Differentiation, 49Discount factor, 62Divorce, 8, 17, 82, 85–87, 89

EEarned income tax credit, 187, 189Economic growth, 12, 206, 215, 219

effects, 14, 40, 146, 147, 165, 200, 204Japanese, 5, 7

Education, 25, 66, 94, 141, 208, 222female, 75, 76, 163, 184risk, 179, 180, 190social capital, 148, 152, 153

Efficiency, 24Elderly ratio, 3

© Springer Japan KK 2017S. Yamashige, Economic Analysis of Families and Society, Advances in JapaneseBusiness and Economics 16, https://doi.org/10.1007/978-4-431-55909-2

225

226 Index

Empowerment, 160, 189, 196, 197, 222in-kind policy, 190

Envy-free, 28, 52Equal opportunity, 28, 52, 218Equity, 15, 27, 28, 33, 187Evolution, 104, 115, see also gmeExternality, 25, 33, 124, 129, 139, 173

children, 88, 109–111, 164, 165, 171migration, 201, 206social capital, 150solution, 35, 65, 66, 76

FFamily, 8, 31, 71, 94, 213

community, 115economics, 73, 94, 223migration, 196, 202, 206policy, 166, 170population crisis, 159, 217poverty, 17, 182social expenditure, 173, 218social norm, 35, 216, 220

Female labor, 44, 168, 171First order condition, 49, 51Foreign worker, 160, 175, 220, 222Forgiveness, 63Free rider problem, 115, 132, 133, 139, 214

community, 33moral hazard, 55social security system, 164, 167solution, 116, 124, 136

GGame

evolutionary, 129, 130, 132repeated, 61, 125, 126sequential, 58simultaneous-Move, 56

Generous, 61, 132, 166, 180, 185, 214fiscal system, 195, 207, 210

Gift, 67, 93, 94, 101, 180, 191Gini coefficient, 15, 16, 147, 182Globalization, 210, 214, 222

effects, 12, 178, 184, 188, 192Governance, 29, 143Government, 28, 34, see also local govern-

mentfailure, 29, 121, 139, 153municipal, 169, 197, 202, 203, 208national, 5, 14, 168, 196, 197, 199, 202–204, 206

prefectural, 197, 202, 208

HHealth insurance, 41, 168Healthcare, 9, 26, 41, 168, 180Housewife, 44, 84, 167, 168, 190, 220Human capital, 223Human rights, 27, 28, 218

IImmigrant, 147, 160, 175, 220, 222Incentive, 14, 29, 58, 66, 95, 164, 185, 207

fiscal transfer, 205, 207free ride, 33, 64, 65moral hazard, 55revitalization, 207, 211to have children, 159, 171, 216to work, 167, 187, 189

Income effect, 51Income inequality, 15, 147, 182, 184Incomplete contract, 27, 30, 33, 66, 82, 85Incomplete information, 26, 100, 146, 162

community, 32, 33financial market, 31government failure, 29insurance, 54, 55quality, 123, 139, 140

Indexfree ride, 126

Inequality, 27, 28, 30, 147, 152Inheritance, 93, 191In-kind

benefit, 171, 190, 217policy, 190

Institution, 23, 28, 34, 88, 108, 159Insurance, 26, 31, 32, 53, 54, 179Intergovernmental, 65, 203, 204, 210Investment, 12, 88, 175, 213, 222

children, 72education, 76, 153, 180, 184marriage, 83, 84social capital, 145, 149

LLabor, 44, 48, 73, 76, 146, 184Labor union, 31, 39Lagrange, 49Local allocation tax, 14, 15, 199, 204, 207Local government, 188, 191, 197, 202, 203,

208, 209competition, 200, 201incentive, 14, 60, 65subsidy, 196, 199, 205

Long-term care, 9, 169, 180

Index 227

Love, 31, 45, 93, 111, 125, 162

MMarket, 24, 34

failure, 24–27Marriage, 8, 17, 31, 42, 71, 72, 82, 87, 89,

162Migration, 13, 14, 197, 199, 201, 215Missing market, 25, 67Mobility, 206, 207

capital, 176effects, 34, 116, 128labor, 12, 39, 66social capital, 153, 201

Monitoring, 29, 55, 116, 128Moral hazard problem, 26, 32, 55, 180, 214

declining fertility, 160fiscal transfer system, 196, 199, 205, 211public assistance, 185Samaritan’s dilemma, 61, 103

Morals, 66, 125, 128Murahachibu, 32, 39, 125, 127, see also

punishiment, Japanese societyMu-tual aid, 7, 19, 115, 175, 178

difficulty, 64, 135, 192social norm, 124, 125

Myrdal, Alva, 42, 160Myrdal, Gunner, 43, 213, 223, 224

NNash bargaining, 63, 83, 89Nash equilibrium, 57, 59Negative income tax, 187, 189Neighborhood association, 10, 11, 30, 32,

181Network, 7, 10, 30, 144, 147Neutrality, 37, 121New community, see also communitydefin-

ition, 116NGO, xi, 11, 116, 121Non-profit, 137, 168Normal goods, 51Not-for-profit, xi, 137NPO, xi, 11, 30, 116, 121, 136, 138

definition, 137in Japan, 138, 191models, 140, 141

OOld-age security hypothesis, 72, 108Opportunity cost, 49

Ostracism, 125Out-of-wedlock, 88Overpopulation, 13, 206

PPay-as-you-go, 65, 109, 110, 112, 164, 165Pension, 40, 64, 104, 109, 180

Japanese, 7, 167, 187side effects, 110, 164universal, 26

Pigouvian subsidy, 66, 173Population distribution, 197, 206Poverty

causes, 52, 177, 178children and single mothers, 182elderly people, 42, 182, 215, 220globalization, 178, 184, 188, 222in Japan, 17, 19, 182private support, 10, 180, 191public support, 18, 167, 185, 190welfare trap, 185, 211young people, 184, 189

Prisoner’s dilemma, 64, 131Productive, 45, 160, 190, 197Property rights, 25, 35, 40, 67, 139, 143Prophylactic, 43, 45, 213, 219Protective, 15, 42, 218Public assistance, 167, 180, 185, 205

Japanese, 17, 182, 186moral hazard problem, 181, 189, 210

Public debt, 5, 30, 213, 221Public goods

community, 32definition, 36donation, 56local governments, 201private provision, 117–119, 122, 129,136

social capital, 149, 153Punishment, 33, 63, 107, 129, 135

free riders, 132ineffective, 34Japanese society, 32, 39, 126mechanism, 29, 32, 65psychological, 66repeated game, 128social, 66, 124, 129, 135social norm, 125, 126

QQuality, 26, 77, 123, 139–141

228 Index

RReciprocity, 7, 33Redistributive system, 160, 188, 205

fiscal transfer system, 214moral hazard problem, 55, 196, 211tragedy of the commons, 65

Regional disparity, 12, 19, 195, 197, 203Religion, 66, 125, 128, 152Replicator dynamics, 129–131Revitalization, 196, 207, 211Risk, 18, 26, 32, 47, 52, 54, 108, 162

aversion, 52marriage, 88pooling, 54poverty, 177, 178sharing, 54, 179

Robot, 175, 221Rotten-Kid Theorem, 95, 96, 103, 111

SSamaritan’s dilemma, 60, 103Second order condition, 49, 51Selfish, 62, 123, 135

children, 72, 95, 97, 100, 103, 105local governments, 65, 201parents, 96social norm, 128tragedy of the commons, 64, 65

Siblings, 93Side effect, 88

economic growth, 13, 15fiscal transfer system, 14moral hazard problem, 214public assistance program, 187social security system, 104, 110, 165,215

Signal, 15, 123, 196Single mother, 87, 180, 182, 187, 191Social capital, 10, 39, 137, 143, 146, 201,

202bonding-type, 145, 148, 149bridging-type, 145, 148, 149definition, 144poverty, 192, 223

Social engineering, 218Social expenditure, 165, 173, 216Social inclusion, 191Social insurance, 3, 35, 41, 167, 180Social norm, 213, 223

care, 9, 104, 105, 108, 111, 175, 215, 220cooperation, 66, 124, 128disappearing, 66

marriage, 88mutual aid, 35sexual division of labor, 19, 44, 171, 184

Social policy, 160, 190, 196, 214definition, 43

Social security system, 7, 166, 178economic growth, 166, 215fertility, 25, 64, 112, 164, 220Japanese, 5, 167migration, 201, 206moral hazard problem, 181, 192, 214reform, 188, 189social norm, 215, 220

Soft budget, 60, 199Strategic bequest, 97Strategy

dominant, 57tit-for-tat, 61, 63, 131trigger, 61

S.t. (subject to), 49Subgameperfect equilibrium, 59, 61, 98, 126Substitution effect, 51

TTax competition, 201Threat

credible, 60empty, 60, 99

Total fertility rate, 159, 166, 174, 207, 217definition, 6replacement level, 159

Traditional community, 115, 116, see alsocommunitydefinition, 116

Tragedy of the commons, 32, 64, 127, 129social security system, 65, 164solutions, 65

Transaction cost, 36, see also bargainingcostTrap

depopulation, 211welfare, 185, 211

Trust, 12, 31, 35, 37, 136, 137, 144, 147, 179

UUnemployment, 185, 221

insurance, 32, 180social capital, 146

Utility, 48expected, 53

VVoting with feet, 200

Index 229

WWelfare state, 5, 42

first year of the, 5, 6, 173Japanese-style, 217

Welfare theoremfirst, 24, 67second, 27

Working poor, 184

Economic Analysis of Families and Society: The Transformation of Japanese Society and Public

Documents

Transcript of Economic Analysis of Families and Society: The Transformation of Japanese Society and Public