Volume LII No. 2 - PIDS Admin December... · 2015. 12. 30. · Volume LII No. 2 ISSN 1655-1516...

TH

E PH

ILIPPINE

RE

VIE

W O

F EC

ON

OM

ICS

Decem

ber 2015Volum

e LII No. 2

Volume LII No. 2 ISSN 1655-1516 December 2015

The Philippine Review of Economics

Editor-in-ChiefEMMANUEL S. DE DIOS

Editorial Advisory BoardRAUL V. FABELLA

HAL C. HILL

CHARLES Y. HORIOKA

KIAN GUAN LIM

ROBERTO S. MARIANO

JOHN VINCENT C. NYE

GERARDO P. SICAT

JEFFREY G. WILLIAMSON

Associate EditorsDANTE B. CANLAS

RAMON L. CLARETE

LAWRENCE B. DACUYCUY

FRANCISCO G. DAKILA JR.

CRISTINA C. DAVID

JONNA P. ESTUDILLO

MARIA S. FLORO

GILBERTO M. LLANTO

ANICETO C. ORBETA

ERNESTO M. PERNIA

STELLA LUZ A. QUIMBO

Managing EditorHONLANI RUTH J. RABE

A joint publication of theUniversity of the Philippines

School of Economicsand the

Philippine Economic Society

Urban-rural income and wage gaps in the Philippines

American private direct investment in the Philippines after independence

Singapore’s five decades of development

The Philippines: on the road to being an emerging economy

Development progeria

Malaysia-China trade

Class participation in politics in Southeast Asia

The effect of trade policy on firm productivity in Thai manufacturing

Households’ Access to Financial Services

The ASEAN business cycle and China’s slowdown

Goods trade liberalization under the ASEAN Economic Community

Just how good is unemployment as a measure of welfare? A note

Karl Kendrick Chua, Louie Limkin, John Nye, and

Jeffrey G. Williamson

Gerardo P. Sicat

Chia Siow Yue

Dante B. Canlas

Sarah Lynne S. Daway and Raul V. Fabella

Tham Siew Yean and Andrew Kam Jia Yi

Joseph J. Capuno

Archanun Kohpaiboon and Juthathip Jongwanich

Gilberto M. Llanto

Bhanupong Nidhiprabha

Ramon L. Clarete and Philip Arnold P. Tuaño

Emmanuel S. de Dios and Katrina Dinglasan

FESTSCHRIFT FOR HAL C. HILL

Maria Socorro Gochoco-BautistaArsenio M. Balisacan

Guest Co-Editors

Aims and Scope: The Philippine Review of Economics (pre) invites theoretical and empirical articles on economics and economic development. Papers on the Philippines, Asian and other developing economies are especially welcome. Book reviews will also be considered.

The pre is published jointly by the up School of Economics and the Philippine Economic Society. Its contents are indexed in the Journal of Economic Literature, EconLit, and RePEc. pre’s readership includes economists and other social scientists in academe, business, government, and development research institutions. Publication Information: The pre (issn 1655-1516) is a peer-reviewed journal published every June and December of each year. A searchable database of published articles and their abstracts is available at the pre website (http://pre.econ.upd.edu.ph). Subscription Information: Subscription correspondence may be sent to the following addresses:

• [email protected] • pssc Central Subscription Service,

psscenter, Commonwealth Avenue, 1101, Diliman, Quezon City, Philippines. P.O. Box 205, up Post Office, Diliman, Quezon City, Philippines 1101PHONE: 922-9627, FAX: 924-4178/926-5179

Submissions: Authors may submit their manuscripts to addresses below:

• [email protected] or [email protected]• The Editor, The Philippine Review of Economics, Rm 237,

School of Economics, University of the Philippines, Diliman, Quezon City, 1101.

Manuscripts must be written in English and in ms Word format. All graphs and tables must be in Excel format. Submission of a manuscript shall be understood by the pre as indicating that the manuscript is not under consideration for publication in other journals. All submissions must include the title of the paper, author information, an abstract of no more than 150 words, and a list of 3–4 keywords. Complete guidelines can be viewed in pre’s website. Copyright: The Philippine Review of Economics is protected by Philippine copyright laws. Articles appearing herein may be reproduced for personal use but not for mass circulation. To reprint an article from pre, permission from the editor must be sought.

EDITOR-IN-CHIEFEmmanuel S. de DiosUP SCHOOL OF ECONOMICS

EDITORIAL ADVISORY BOARDRaul V. FabellaUP SCHOOL OF ECONOMICS

Hal C. HillAUSTRALIAN NATIONAL UNIVERSITY

Charles Y. HoriokaASIAN GROWTH RESEARCH INSTITUTE

(KITAKYUSHU)

Kian Guan LimSINGAPORE MANAGEMENT UNIVERSITY

Roberto S. MarianoUNIVERSITY OF PENNSYLVANIA

John Vincent C. NyeGEORGE MASON UNIVERSITY

Gerardo P. SicatUP SCHOOL OF ECONOMICS

Jeffrey G. WilliamsonHARVARD UNIVERSITY

ASSOCIATE EDITORSDante B. CanlasUP SCHOOL OF ECONOMICS

Ramon L. ClareteUP SCHOOL OF ECONOMICS

Lawrence B. DacuycuyDE LA SALLE UNIVERSITY

Francisco G. Dakila Jr.BANGKO SENTRAL NG PILIPINAS

Cristina C. DavidPHILIPPINE INSTITUTE

FOR DEVELOPMENT STUDIES

Jonna P. EstudilloNATIONAL GRADUATE INSTITUTE

FOR POLICY STUDIES (TOKYO)

Maria S. FloroAMERICAN UNIVERSITY (WASHINGTON D.C.)

Gilberto M. LlantoPHILIPPINE INSTITUTE


Aniceto C. OrbetaPHILIPPINE INSTITUTE


Ernesto M. PerniaUP SCHOOL OF ECONOMICS

Stella Luz A. QuimboUP SCHOOL OF ECONOMICS

MANAGING EDITORHonlani Ruth J. RabeUP SCHOOL OF ECONOMICS

COPY EDITORDinna Dayao

The Philippine Review of EconomicsA joint publication of the UP School of Economics (UPSE)and the Philippine Economic Society (PES)

PRE

2015 OFFICERS AND BOARD OF DIRECTORS

PRESIDENTVictor AbolaUNIVERSITY OF ASIA AND THE PACIFIC

VICE PRESIDENT Rosemarie EdillonNATIONAL ECONOMIC AND DEVELOPMENT AGENCY

SECRETARY Majah-Leah RavagoUNIVERSITY OF THE PHILIPPINES DILIMAN

TREASURER Ronald MendozaASIAN INSTITUTE OF MANAGEMENT

MEMBERSAnthony AbadTA TRADE ADVISORY GROUP

Gil BeltranDEPARTMENT OF FINANCE

Roehlano BrionesPHILIPPINE INSTITUTE FOR DEVELOPMENT STUDIES

Jose CamachoUNIVERSITY OF THE PHILIPPINES LOS BAÑOS

Lawrence DacuycuyDE LA SALLE UNIVERSITY

Diwa GuinigundoBANGKO SENTRAL NG PILIPINAS

Peter Lee UUNIVERSITY OF ASIA AND THE PACIFIC

Emilio NeriBANK OF THE PHILIPPINE ISLANDS

IMMEDIATE PAST PRESIDENTStella Luz QuimboUNIVERSITY OF THE PHILIPPINES

PHILIPPINE REVIEW OF ECONOMICSEmmanuel de DiosEDITOR

COUNCIL OF ADVISERS Romeo BernardoLAZARO, BERNARDO, TIU AND ASSOCIATES

Raul FabellaUNIVERSITY OF THE PHILIPPINES DILIMAN

Cielito HabitoATENEO DE MANILA UNIVERSITY

Ernest Leung

Solita Collas-MonsodUNIVERSITY OF THE PHILIPPINES DILIMAN

Cesar VirataC. VIRATA AND ASSOCIATES

The Philippine Economic SocietyFounded 1961

The Philippine Economic Society (PES) was established in August 1962 as a nonstock, nonprofit professional organization of economists.

Over the years, the PES has served as one of the strongest networks of economists in the academe, government, and business sector.

Recognized in the international community of professional economic associations and a founding member of the Federation of ASEAN Economic Associations (FAEA), the PES continuously provides a venue for open and free discussions of a wide range of policy issues through its conferences and symposia.

Through its journal, the Philippine Review of Economics (PRE), which is jointly published with the UP School of Economics, the Society performs a major role in improving the standard of economic research in the country and in disseminating new research findings.

At present the society enjoys the membership of some 800 economists and professionals from the academe, government, and private sector.

• Lifetime Membership –Any regular memberwho pays the lifetime membership dues shall be granted lifetime membership and shall have the rights, privileges, and responsibilities of a regular member, except for the payment of the annual dues.

• RegularMembership–Limitedtoindividuals21years of age or older, who have obtained at least a bachelor’s degree in economics, or who, in the opinion of the Board of Directors, have shown sufficient familiarity and understanding of the science of economics to warrant admission to the Society. Candidates who have been accepted shall become members of the Society only upon payment of annual dues for the current year.

• Junior Membership –This is reserved for full-time college or graduate students majoring in economics. Affiliation for junior membership is coursed through the Junior Philippine Economic Society (JPES).

For more information, visit: www.phileconsociety.org.

The Philippine Review of Economics

Volume LII No. 2 December 2015 ISSN 1655-1516

iii Foreword

1 Urban-rural income and wage gaps in the Philippines: measurement error, unequal endowments, or factor market failure?Karl Kendrick Chua, Louie Limkin, John Nye, and Jeffrey G. Williamson

22 American private direct investment in the Philippines after independence

Gerardo P. Sicat

38 Singapore’s five decades of development: lessons and future directionsChia Siow Yue

65 The Philippines: on the road to being an emerging economyDante B. Canlas

84 Development progeria: the role of institutions and the exchange rateSarah Lynne S. Daway and Raul V. Fabella

100 Malaysia-China trade: old and new routes Tham Siew Yean and Andrew Kam Jia Yi

118 Class participation in politics in Southeast Asia Joseph J. Capuno

ii Contents

143 The effect of trade policy on firm productivity in Thai manufacturingArchanun Kohpaiboon and Juthathip Jongwanich

170 Households’ Access to Financial Services: Some Evidence from Survey DataGilberto M. Llanto

192 The ASEAN business cycle and China’s slowdownBhanupong Nidhiprabha

210 Goods trade liberalization under the ASEAN Economic Community: effects on the Philippine economy

Ramon L. Clarete and Philip Arnold P. Tuaño

234 Just how good is unemployment as a measure of welfare? A note

Emmanuel S. de Dios and Katrina Dinglasan

Foreword

Hal C. Hill retired from the Australian National University in 2014, where he was H.W. Arndt Professor of Southeast Asian Economies, capping a long and distinguished academic career. This Festschrift is a small tribute to a distinguished colleague and friend. The papers were contributed by an impressive and select group of scholars, some of whom are his former students but all of whom, in any case, are his close friends.

Hal has devoted his entire career to understanding and promoting economic development in Southeast Asia. He stands in the front ranks of experts in the economics and political economy of many countries in the region, especially

Indonesia and the Philippines, the subjects of his research over many decades. He has mentored scores of the region’s scholars, and policy makers, including many who currently hold important positions in government. He has guided and generously promoted the careers of young researchers.

Hal’s work is especially valued in the Philippines—his first overseas foray after graduate school was as a Visiting Assistant Professor of Economics at the University of the Philippines School of Economics from May 1980 until May 1982—where his critical analysis of difficult issues and unprejudiced attitude have made him a scholar trusted by all sides.

Both co-editors of this Festschrift have had the honor and pleasure of working closely with Hal over the years. As part of his wide-ranging work, Hal co-edited two volumes on the Philippines with Arsi Balisacan, the first being a well-received survey of long-term trends and major sectors in the Philippine economy [Balisacan and Hill 2003]. This volume was significant in bringing together the work and wisdom of both an earlier and a more recent generation of development scholars, providing an updated analysis until the threshold of the current century. An important sequel to that venture in “transnational scholarship” was Balisacan and Hill [2007], which tackled subnational issues of regional development and

iv Foreword

decentralization. Typical of Hal’s influence here, as elsewhere, was his insistence on a comparative perspective, i.e., on the need to examine local experience in the light of what other countries have undergone, thus avoiding the solipsism that tempts all country specialists. In 2013, he organized and co-edited a volume with Corina Gochoco-Bautista [Hill and Gochoco-Bautista 2013] which discusses the challenges Asia faces in sustaining growth by examining Asia’s growth performance relative to its growth dynamics, the development of key sectors and institutions, savings and investment behavior, and aspects of country-level idiosyncrasies and regional cooperation and integration. Ever active and relevant, Hal is even now among those helping the Philippine government craft a long-term vision for Philippine growth and development dubbed Filipino 2040. An initiative of the National Economic and Development Authority in collaboration with the country’s multilateral development partners, the business sector, and the academe, Filipino 2040 aims to identify the policy and institutional imperatives necessary to realize the vision of a high quality of life for all Filipinos.

On a personal level, Hal has ever been the gracious host whenever any of us would visit Canberra. He would open a nice bottle (or maybe two or three) from his wine cellar. On his visits to the Philippines, he has generously plied the dean’s cellar with some of the best bottles in his collection. Corina once gave his name as a reference when she applied for a visa to Australia. The embassy tried to verify whether she in fact knew him by asking her to list his birth date! Hal very kindly provided the information.

Hal is certain to remain active and to continue his distinguished scholarship for many more years to come. In the meantime, this is one way to thank him for his friendship and to convey to him the high regard and appreciation of the economics profession.

References

Balisacan, A. and H. Hill, eds. [2003] The Philippine economy: development, policies, and challenges. Oxford: Oxford University Press.

Balisacan, A. and H. Hill, eds. [2007] The dynamics of regional development: the Philippines in East Asia. Cheltenham, U.K.: Edward Elgar and Asian Development Bank.

Hill, H. and M.S. Gochoco-Bautista [2013] Asia rising-growth and resilience in an uncertain global economy. Cheltenham, U.K.: Edward Elgar and Asian Development Bank.

The Philippine Review of EconomicsVol. LII No. 2, December 2015 pp. 1-21

PRE

Urban-rural income and wage gaps in the Philippines: measurement error, unequal endowments, or factor

market failure?

Karl Kendrick Chua*, Louie Limkin*, John Nye**, and Jeffrey G. Williamson***

Income inequality is higher in the Philippines than in most of its Asian neighbors, and spatial inequality accounts for a fairly large share of it. There is little evidence of labor market failure in the Philippines since, when properly measured, wage gaps by skill level are modest. Unequal endowments account for most of the urban-rural income gaps. That is, individual attributes of workers and households explain the majority of the urban-rural gaps, and schooling, skill, and experience are the three individual characteristics that matter most. Provincial variables, like typhoon incidence, government corruption, school crowding, and access to health facilities, matter far less. Workers born in the cities and immigrants to the cities invest more in human capital than do rural workers. However, this paper cannot tell us how much of that is due to better human-capital-building infrastructure supply in the cities and how much is due to higher urban demand for that infrastructure.

JEL classification: D3, J3, O3, R1Keywords: regional inequality, wage gaps, migration, the Philippines

1. Introduction

Large urban-rural wage and income gaps have been one of the most-studied phenomena in the development literature.1 At the same time, high rates of labor migration from rural to urban employment and from agriculture to industry and modern service employment have been notable characteristics of the transition

1 Weber [1899], Hatton and Williamson [1992], Yankow [2000], Gould [2007], Sicular et al. [2007], and Young [2013].

2 Chua, Limkin, Nye, and Williamson: Urban-rural income and wage gaps in the Philippines

from pre-industrial to modern development. The coexistence of rapid emigration from poor and low-wage rural areas with persistent urban-rural income gaps might seem a puzzle, until one realizes that rapid semi-skilled and skilled labor demand growth and job creation in urban areas might explain both.

The Philippines, despite starting above its Southeast Asian neighbors’ development levels in the first half of the twentieth century and even in the 1950s, has underperformed since the 1970s, especially in industry (de Dios and Williamson [2014]; World Bank [2013]). It also has relatively high income inequality, rural-urban income gaps, and regional inequality.2 Policy is said to have favored Luzon, and, in particular, Metro Manila and its surrounding regions. Although labor market forces should have led to a closing of the urban-rural gap and a diminution in regional inequality, they have not done so, perhaps, some argue, due to impediments to migration, such as minimum wage legislation, regularization policy, labor unions, urban immigration restrictions, land zoning rules, and other market interventions.

This paper does find some evidence of labor market failure, but it is very modest. In general equilibrium theory, land, labor, and capital markets all matter in accounting for spatial income gaps. But what about the spatial labor market itself? Are there big urban-rural wage gaps? If so, are they explained by policy intervention, weak institutions, or rather by rural behavior, lack of language skills, poor rural schools, fear of religious persecution (a poor Muslim south and a rich Christian north), or some overwhelming demographic glut in rural areas [Williamson 2013]?

Understanding the determinants of income inequality both within and between regions has been and will be essential to identifying the sources of poor Philippine performance and to designing policy that will promote more rapid equitable development. The last major study of spatial income inequality in the Philippines was undertaken by Balisacan and Fuwa [2004]. In their study, the authors concluded that income inequality had undergone little change from 1985 to 2000. They also noted that income inequality was higher in the Philippines than in most of its Asian neighbors, although it is not as high as in Latin America, with which the Philippines is often compared. Furthermore, they concluded that while spatial inequality accounted for a large share of overall income inequality in the Philippines, that share seemed to be diminishing. Finally, they reported that income per capita was slowly converging across provinces, in contrast with many other developing economies.

2 For example, see Berry [1978] and Balisacan and Fuwa [2004].

The Philippine Review of Economics, Volume LII No. 2, December 2015 33

This paper extends the Balisacan and Fuwa assessment forwards one decade. It also improves the data base and estimates the sources of the urban-rural gap using Fields and Oaxaca decomposition techniques. We explore the roles of measurement error, unequal endowments, and labor market failure, plus the likely impact of capital and land market failure. Thus, we ask: How much of the nominal wage gap disappears when proper site-specific prices—especially those of non-tradable services—are used for rural areas? How much of the measured average wage gap disappears when it is computed for the same occupations and skills, rather than for some average wage? Are the gaps different for skilled and unskilled workers, perhaps reflecting different responses to wage differentials?3

These are the questions that should be answered in making a judgment about labor market failure. If, instead, the focus is on the determinants of spatial income inequality, then different questions are relevant. Indeed, do household wage incomes behave differently than do wage rates by occupation and skill, perhaps due to household labor participation rate or employment rate differences between rural and urban locations? Do household total income gaps behave differently than wage income gaps, perhaps due to unobserved positive individual attributes (migrants self-select), and/or due to differences in investment income opportunities by location, and/or due to remittances from family abroad? Most importantly, are those large rural-urban income gaps explained mostly by wage gaps or mostly by endowment gaps?

2. Philippine urban-rural gaps, total inequality, and regional inequality: a comparative assessment

Based on total household income, inequality in the Philippines improved very slowly between 2000 and 2012 (Figure 1).4 The trends are broadly the same whether we use the Gini coefficient or Theil’s T Statistic. However, we will use Theil’s T Statistic in what follows since it is easily decomposable while the Gini is not.

Table 1 reports country urban-rural gap measures for three statistics: the average wage; the average household wage income; and the average household total income. The gap is measured as the difference between urban and rural, divided by urban: thus, the percentage by which urban exceeds rural. The sample is small in any of those categories, and only the Philippines reports all three.

3 The idea here is that skilled workers are more likely than poorer, unskilled workers to have the financial resources to invest in a move.4 Philippine household income data are available every three years. Thus, to update the assessment by Balisacan and Fuwa, our analysis uses data from 2002 to 2012.


FIGURE 1. Measures of inequality

TABLE 1. Cross-country comparison of wage and income gaps

Country/Area Year r-u gap Data used Source

Philippines 2011/12

34 Individual wage Labor Force Survey (2011)

44 HH total income Family Income and Expenditure Survey (2012)

54 HH wage income Family Income and Expenditure Survey (2012)

Nepal 2003 4 Individual wage Hertz et al. (2008)

Bangladesh 2000 9 Individual wage Hertz et al. (2008)

EU25 2003 12 HH total income Cameron et al. (2008)

Indonesia 2000 16 Individual wage Hertz et al. (2008)

India 2005 23 Individual wage Karan an Selvaraj (2008)

US 2011 24 HH total income Bureau of Labor Statistics, US Department of Labor

Vietnam 1998 26 Individual wage Hertz et al. (2008)

India 2005 31 Individual wage Karan an Selvaraj (2008)

Malaysia 2004 53 HH total income Ragayah (2008)

Cambodia 1999 56 HH wage income Morris (2007)

China 2002 58 Per capita income Sicular et al. (2007)

China 2011 81 HH wage income National Bureau of Statistics of China

0.50

0.60

0.30

0.40

2000 2003 2006 2009 2012

Gini coeff. Theil T

Source: NSO, WB staff estimates


But the Philippine results are revealing. The wage gap, 34 percent, is the smallest of the three. It reflects three influences: the fact that urban jobs are more skilled; that the urban cost of living is higher; and that there may be some labor market failure by specific job category. The household total wage income gap, 54 percent, is much bigger. It reflects the fact that urban immigrants self-select, that low-wage (high-wage) earners cluster in low-income (high-income) households, and that urban workers have higher hours worked per month. Surprisingly, the household total income gap, 44 percent, is lower than the household wage income gap. This cannot be explained by remittance rates since the share of remittance income to total household income was roughly the same for rural and urban households. In 2012, they were 17 percent of total rural household income and 15 percent of urban household income. Most importantly, the wage gap is significantly less than both income gaps.

Compared with its Asian neighbors, wage gaps in the Philippines are high: that 34 percent is the highest in Table 1; the reported wage gap is much higher than in Nepal (4 percent), Bangladesh (9 percent), and Indonesia (16 percent); and it is even higher than Vietnam (26 percent) and India (23-31 percent). The Philippine household wage income gap is only exceeded by China (81 versus 54 percent). Finally, the Philippine household total income gap of 44 percent is closest to Malaysia (53 percent), and both are much bigger than that of the United States (24 percent) and the European Union (12 percent).


FIGURE 2. Income gaps (in percent)

Given the country’s heterogeneous island geography with multiple languages, ethnic groups, and religions, one might expect that spatial income gaps

60

2000 2003 2006 2009 2012

urban-rural

55

50

45

40

35

30

Luzon-VisMin PHL-ARMM


(Figure 2) would account for a large share of the high Philippine inequality. But do they? Table 2 reports some country measures where total inequality is decomposed into between regions and within regions. The canonical case of Italy—with its persistent “North-South” or Mezzogiorno problem—records 13 percent of total inequality due to regional gaps. But the figure for the Philippines is higher, 20-24 percent in 2000, although it dropped to 11-14 percent in 2012. The figures are smaller for better integrated and richer countries, but they are bigger for China (38 percent) and Vietnam (36 percent).5 Although it is not shown here, the urban versus rural between share in total Philippine inequality is a bit more than 10 percent in 2012, which is almost equal to the regional between share of 11 percent.

TABLE 2. Within and between income inequality: the Philippines compared with other countries

Country/Area Year Within

region/

province

Between

regions/provinces

Source

Philippines 2011 80/76 20/24Family Income and

Expenditure Survey (2012)

Philippines 2012 89/86 11/14Family Income and

Expenditure Survey (2012)

Indonesia 1997 50 50 Akita (2003)China 1998 62 38 Akita (2003)Vietnam 1999 64 36 Minot et al. (2003)Indonesia 1999 78 22 Tadjoeddin (2003)Mozambique 1997 80 20 Simler and Nhate (2005)Indonesia 1990 83 17 Tadjoeddin (2003)Italy 995 87 13 Forster et al. (2005)Russia 1995 90 10 Forster et al. (2005)India 1983 94 6 Mishra and Parikh (1992)Czech

Republic1996 97 3 Forster et al. (2005)

Canada 1997 98 2 Gray et al. (2003)Poland 1999 98 2 Forster et al. (2005)

5 We exclude Indonesia from these comparisons since it has very different estimates for within and between province inequalities from two different sources (Akita [2003]; Tadjoeddin [2003]) and even for the same source [Akita 2003].


Thus, while urban and regional income gaps certainly contribute to inequality in the Philippines, it is the variance of household incomes within urban and within rural areas and regions that matters most (80-89 percent). A study by Alwyn Young [2013] used consumption data from country Demographic and Health Surveys to show that the urban-rural gap explains a much higher share of total consumption inequality, on average 40 percent in his low-income sample, and the Philippines was about in the middle of his 65-country pack [Young 2013:1748].6

3. Gap measurement errors

3.1. Getting the prices right

Since migrants respond to real wages and since living costs are lower in rural areas within all provinces and lower still in rural areas in the poorest provinces, some effort must be made to get the prices right. Typically, either the urban-rural gap estimates are nominal, or they are deflated only by the prices of tradable commodities (e.g. they exclude non-traded services which are priced low in rural areas since nominal wages are also low), and when deflated, they use provincial capital prices rather than rural prices. Since rural cost of living indices or consumer price indices are not available from official Philippine sources, we compute them using a modified version of the provincial consumer price indices reported by the National Statistics Office. We consider the official provincial consumer price indices as urban since the underlying price data are mainly collected from urban areas.

Our rural consumer price indices are computed using five consumption categories: agricultural food; other food; tradable goods and services7; semi-tradable goods and services; and non-tradable goods and services.8 Agricultural food prices are taken from the Bureau of Agricultural Statistics, which provides farm gate and retail prices of key commodities. Only rice has complete price data at the provincial level so it is used as a proxy for agricultural food, and rural prices are calculated as the average of the farm gate and retail price. The agricultural food price index in the initial year (2001) is constructed as the average of the farm gate and retail price divided by retail price. Succeeding years follow trends in the farm gate and retail price. The prices of other food, tradables, and semi-tradable goods and services follow the provincial consumer price indices. Non-tradable goods and services prices are assumed to follow the movement of nominal rural wages. The initial year price of non-tradables is computed as the ratio of rural to urban nominal wages multiplied by the urban price of these non-tradables. Rural non-tradable price movements in succeeding years follow the growth in rural

6 It should be noted that Young’s consumption measure is based on ownership of durables and housing quality [Young 2013:1734], a measure that would exaggerate gaps.7 For the purposes of this analysis, tradables include alcoholic beverages, tobacco, clothing, footwear, furnishings, household equipment, and communication.8 It is assumed that 30 percent of semi-tradables are not traded.


nominal wages. This method ensures that the prices of food and non-tradable goods and services are cheaper in rural areas.

FIGURE 3. Rural-urban income gaps (nominal and real in percent)

As expected, the urban-rural income gap falls when these improved rural site prices are used instead of provincial-capital prices (Figure 3). Over the 2000-2012 period, the urban-rural real income gap is 17 percent lower than the nominal gap, and higher urban prices account for around 6.1 percent of the urban-rural nominal income gap. That is, the average urban-rural cost of living difference/average urban-rural nominal income gap = 17 percent/52 percent = 33 percent. The Oaxaca decomposition offers another way to assess how much cost of living differences account for the urban-rural gap. Using individual household income data for 2000-2012, the decomposition tells us that about P38,000, or 34 percent of the P110,000 difference, is explained by cost of living differences (Table 3), a figure very close to the 33 percent using descriptive statistics.

TABLE 3. How much of the urban-rural income gap is explained by the cost of living differences?

Explanatory variable: prices

Average HH incomeUrban 229,999.3Rural 120,342.8Difference 109,656.5DecompositionExplained 37,734.9Unexplained 71,921.6Percent Explained 34.4

60

2000 2003 2006 2009 2012

nominal rural-urban


55

50

45

40

35

30

65

25

20

real rural-urban


3.2. Getting the wages right

What happens to the 34 percent nominal urban-rural wage gap when average wages (mean wage in urban barangays versus mean wage in rural barangays) are broken down by skilled or unskilled wage categories? The urban-rural nominal wage gap for unskilled labor is 31 percent; the real wage gap is 6 percent. The urban-rural nominal wage gap for skilled labor is 17 percent; the real wage gap is nil. Both figures are below the 34 percent gap for the average wage (Figures 4 and 5).9 Presumably, the measured real wage gap would decline even further if we could disaggregate jobs in greater detail. In short, the reported nominal urban-rural wage gap falls significantly when prices and wages are measured more accurately. Also, note that the skilled labor urban-rural wage gap declined over the past few years, while the unskilled wage gap remained constant (Figure 6). One reason for the recent decline in the skilled wage gap could be the dispersion of business process outsourcing firms from Manila to secondary cities like Cebu, Bacolod, Iloilo, and Davao.10

FIGURE 4. Average nominal rural and urban wages in 2011 (skilled, in pesos per day)

9 Note that we predicted the lower real wage gap for skilled workers, appealing to the notion that they were better able to finance the move.10 In 2007, 82 percent of the full-time business process outsourcing workers were located in Manila. However, the share fell to 75 percent in 2013.

190

urban

Source: Labor Force Survey

700

rural

600

500

400

300

200

100

0

86 or 31%

P p

er d

ay

}277


FIGURE 5. Average nominal rural and urban wages in 2011 (skilled)

FIGURE 6. Urban-rural skilled and unskilled ages gaps. (Average nominal rural and urban wages)

Finally, we can find no evidence of a Todaro effect in the Philippines.11 That is, underemployment plays no role in accounting for these wage gaps. Wages adjusted for underemployment—the daily wage divided by 8, multiplied by

11 Strictly speaking, the Todaro effect [1971] means that rural wages would be closer to urban wages if we took (allegedly higher) urban unemployment into account. Here, we use underemployment to adjust wages and loosely refer to it as the Todaro effect.

urban


700

rural

600

500

400

300

200

100

0

75 or 12%P

per

day

}

554629

45

2001

unskilled gap


40

35

30

25

20

15

50

10

5

skilled gap

0

2002 2003 2004 2001 2006 2007 2008 2009 2010


normal hours worked per day (only for underemployed workers working less than 8 hours a day)—differ from unadjusted wages only by about 0.5 percent. Of course, it’s possible that the explanation for the small difference is that workers already factor in underemployment when reporting their daily wages during the survey interview. Since the Todaro-adjusted wage is not statistically different from the unadjusted wage, the latter is used in all the regressions that follow.

4. What explains the Philippine wage gap?

As we have seen, urban-rural wage gaps diminish greatly when computed by skill and properly deflated by location-specific cost-of-living. What accounts for gaps when properly measured?

This section offers a more explicit statistical explanation of the wage gap using the Oaxaca and Fields decomposition methods. The period covered is 2001-2009. The unit of observation goes down to the individual and household level, and it is grouped by urban or rural location, which is in turn based on barangay characteristics. Individual attributes like education, age, and wages by occupation or skill are available for all years. We also use provincial variables, although they are available only triennially (i.e., 2003, 2006, and 2009). Our main data sources are the quarterly Labor Force Survey and the triennially Family Income and Expenditure Survey. Education, health amenities, and outcome variables are sourced from government administrative data. Poverty and Gini statistics are computed from the Family Income and Expenditure Survey. Infrastructure, political, and geographic data are sourced from the Asia Pacific Policy Center provincial data set. The Oaxaca decomposition uses the estimated coefficients from two separate regressions—one for rural barangays and one for urban (and for other group categories, respectively)—to account for the difference between the average urban-rural wage gaps (unadjusted by our new rural prices). The difference is broken down into the explained and unexplained components. There is reason to believe that the unexplained residuals can in large part be attributed to undocumented (positive) attributes of city immigrants [Young 2013]. The regressions are pooled Ordinary Least Squares. Multi-collinearity often results when all the provincial-level variables are used at the same time, but the regressions exploited below are limited to those that are free of multi-collinearity.

To better understand the contribution of these variables to the wage gap, we implement the Oaxaca and Fields decompositions. As noted above, the regressions underlying the decompositions are pooled Ordinary Least Squares, where year and region are added as controls, and the results are presented in Table 4.12 As expected, wages are higher on average for urban, skilled, and large

12 Pooled Ordinary Least Squares with year controls, and year and occupation controls, were also run. The results are broadly similar to the regressions used below. The regressions using occupation controls yield higher R2.


industry workers compared to their counterparts. Higher education, age, large firm employment, road density, and the regional minimum wage are also associated with higher wages. The last column in Table 4 uses the Fields decomposition index (the share of the total variance attributable to each explanatory variable in percent) to assess the importance of each variable. The most important variables are skilled work (23.3 percent), large industry employment (3 percent), college education (50.9 percent), age as a proxy for experience (11.5 percent), and provincial road density (10.4 percent).

TABLE 4. Explaining nominal wages: individual pooled data – Ordinary Least Squares regressions

Dependent variable: ln(wage)(1) (2) (3) (4) (4) Fields

DecompositionUrban 0.0544*** 0.0514*** 0.0600*** 0.0500*** 1.2

Skilled 0.4186*** 0.4189*** 0.4185*** 0.4188*** 23.3

Large industry 0.1577*** 0.1577*** 0.1581*** 0.1576*** 9.8

Sector of employment

Industry 0.2501*** 0.2512*** 0.2529*** 0.2497*** 3.0

Services 0.0186*** 0.0195*** 0.0204*** 0.0185*** 0.5

Individual Characteristics

Educational attainmentElementary undergraduate 0.1424*** 0.1406*** 0.1404*** 0.1419*** -2.9Elementary graduate 0.1786*** 0.1765*** 0.1764*** 0.1780*** -3.1

High school undergraduate 0.2493*** 0.2466*** 0.2468*** 0.2483*** -3.6

High school graduate 0.3419*** 0.3392*** 0.3395*** 0.3409*** -3.0

College undergraduate 0.5033*** 0.5004*** 0.5007*** 0.5022*** 4.0

College graduate 0.8291*** 0.8262*** 0.8266*** 0.8280*** 46.9

Age 0.0417*** 0.0418*** 0.0418*** 0.0417*** 29.9

Age2 -0.0004*** -0.0004*** -0.0004*** -0.0004*** -18.4

Provincial characteristics

Student to room ratio -.29167*** -1.8014*** -2.5462*** -2.2050*** 0.9

Number of doctors 0 0.0002** 0.0001 0.0001 -9.7

Remittance to HH income ratio -0.0012* -0.0017** -0.4

Domestic remittance to HH inc ratio -0.0019* -0.0035***

Distance from Manila 0 -0.0000* -0.0000* 0 1.9

Gini coefficient -0.1418*** -0.1136*** 0.5

Poverty incidence -0.0021*** -0.0015*** 2.1

LGU income (from local sources) 0.0076*** -0.0006 0.0059** 0.0022 -0.4

Total road density 0.2057*** 0.1559*** 10.4

Access to electricity 0.0004* 0.0014*** 0 1.7

Access to safe water -0.0002** 0 -0.0001 -0.0002 0.0

Political dynasty index 0 0 0 0 0.0

Conflict incidence -0.0043*** -0.0044*** -0.0051*** -0.0040*** 0.0

Frequency of typhoons 0.0021 -0.0023 -0.0034 0.002 0.3

Prices -0.0001 0.0001 -0.0002 0.0001 1.0

Min wage 0.0006*** 0.0011*** 0.0011*** 0.0007*** 4.1

Constant 3.3875*** 3.5585*** 3.2935*** 3.5501***


Dependent variable: ln(wage)(1) (2) (3) (4) (4) Fields

DecompositionR2 0.54 0.54 0.54 0.54

Observations 272150 272150 272150 272150

Controls

Time Y Y Y Y

Region Y Y Y Y

In addition, bivariate regressions of the log nominal wage with provincial level variables reveal that only the following are significant and have the expected signs: student to classroom ratio; grade 6 test scores; percentage of children who are fully immunized; distance of provincial capital from Manila; Gini coefficient; poverty incidence; local government revenue; road density; proportion of households with electricity; access to safe water; political dynasty; conflict; and number of typhoons per year. When these provincial-level environmental attributes are added to the regression equation reported in Table 4, students per classroom (a school quality proxy), remittances, distance of provincial capital from Manila, the Gini coefficient, poverty incidence, local government income, road density, and incidence of conflict are all significant and have the expected signs. However, except for road density, their explanatory power is very small since their presence raises only slightly and each Fields statistic is small. These results suggest that initial location conditions (e.g., poverty, inequality, typhoon incidence, size of the provincial economy, distance from Manila), location amenities (e.g., classroom size, road infrastructure, health facilities)13, and location remittance experience all influence barangay wage levels, but not by much.

Most importantly, after controlling for all these provincial and individual attributes, urban location has little else to offer (1.2 percent Fields effect). What contribute most to the urban-rural average wage gap are the individual characteristics relevant to high-wage firms, like schooling, skills, and experience.14 However, if provincial characteristics influence individual characteristics, then these decompositions understate the importance of location-specific human-capital-building infrastructure.

The Oaxaca decomposition in Table 5 confirms these results. Five groups are defined for the wage decompositions: urban-rural for all workers; urban-rural for unskilled workers only; skilled-unskilled15; formal-informal16; and workers

13 See Albouy [2008] for a comprehensive assessment of the impact of amenities and quality of life on wages across metropolitan areas in developed countries.14 How skills determine city competitiveness is explored at length in Glaeser and Mare [2001].15 Occupation codes from the Philippine Standard Occupational Classification Code and their respective average wages were used to determine which occupations are skilled or unskilled.16 Here we follow the definition in the 2013 Philippine Development Report, “Creating More and Better Jobs”, where those working in agriculture, transport, communications, wholesale trade, and retail trade are all treated as informal workers.


in large and small industries.17 Table 5 reports those underlying regressions: the variables included in the decomposition explain 93 percent of the urban-rural wage gap. These variables also explain 46 percent of the skilled-unskilled wage gap, 77 percent of the formal-informal wage gap, and 68 percent of the large-small industry wage gap.

TABLE 5. Individual pooled data: Oaxaca decomposition of wages for 5 groups

Urban ruralEduc only Age only Educ and

age onlyProvincial

variables onlyAverage wage

Urban 279.1

Rural 157.9

Difference 91.2

Deocmposition

Explained 84.9 45.5 5.3 40.4 58.4

Unexplained 6.3 45.7 85.9 41.8 32.8

Percent Explained 93.1 49.9 5.8 54.1 84.0

Skilled-UnskilledEduc only Age only Educ and

age onlyProvincial


Urban 469.1

Rural 165.4

Difference 263.7

Deocmposition

Explained 129.1 116.1 -4.8 129.9 9.6

Unexplained 151.5 167.5 269.4 151.5 271.1


Formal-InformalEduc only Age only Educ and

age onlyProvincial


Urban 278.7

Rural 178.4

Difference 102.3

Deocmposition

Explained 78.7 58.4 12.2 68.0 18.2

Unexplained 23.6 43.9 90.1 34.3 81.1


17 The distinction between small and large industry is used as a proxy for labor policy, in which large firms are more likely to be affected by minimum wage and contractual terms. The distinction is derived using the 2008 Informal Sector Survey. Industries in which informal workers comprise at least 90 percent are defined as small.


Urban-Rural Unskilled

Educ only Age only Educ and age only

Provincial variables only

Average wage

Urban 210.5

Rural 150.9

Difference 58.7

Deocmposition

Explained 58.2 20.1 3.4 23.3 40.4

Unexplained 1.5 39.7 56.3 36.4 10.3


Large-Small Industry

Educ only Age only Educ and age only


Average wage

Urban 331.6

Rural 164.0

Difference 167.6

Deocmposition

Explained 114.5 96.9 10.5 104.7 13.7

Unexplained 53.1 70.7 157.1 62.9 153.9


5. What explains the Philippine income gap?

Finally, we apply the Oaxaca decomposition to total household income from 2000 to 2009 using the same regression model and variables that we did for wages. To repeat our earlier finding, about P38,000 of the P110,000 urban-rural nominal income gap, or 34 percent, is explained by cost of living differences. Table 6 deals with the real income gap. Since the individual and provincial variables exhibit much the same statistical impact on income gaps as they do on wage gaps, we do not report their detail here in the text except to repeat that education and experience are the individual variables that dominate: by themselves they explain 71 percent of the household total income gap.

However, one income gap finding does emerge which contrasts dramatically with the wage gap findings: provincial environmental variables have a much bigger impact—50 percent of household income gaps are explained by provincial variables alone. This result strongly suggests that provincial variables have a powerful effect on labor participation rates, skill mix, occupation mix, property incomes, and remittances; they have little impact on wage rates themselves.


TABLE 6. Oaxaca Decomposition of urban-rural real total household income gaps

All Educ only

Age only Educ and age only


Average wageUrban 178,896.0Rural 115,311.8Difference 63,584.2DeocmpositionExplained 66,985.1 41,782.9 0.0 45,083.2 31,577.9Unexplained 0.0 21,801.4 63,584.2 18,501.0 32,006.3Percent Explained 100.0 65.7 0.0 70.9 49.7

6. Understanding spatial wage and income gaps: thinking in general equilibrium

Income inequality is higher in the Philippines than in most of its Asian neighbors, and spatial inequality accounts for a fairly large share of it. So what accounts for the large urban-rural wage and income gaps in the Philippines? Individual attributes of workers and households explain the majority of the gaps, and schooling, skill, and experience are the three individual characteristics that matter most. Provincial variables like typhoon incidence, government corruption, school crowding, and access to health facilities have only a modest impact on urban-rural wage gaps, but they have a powerful impact on urban-rural income gaps. Workers born in the cities and immigrants to the cities invest more in human capital than do rural workers, but this paper cannot tell us why. That is, it cannot say how much of this is due to better human-capital-building infrastructure supply and how much is due to individual demand for that infrastructure, and, if the latter, how much is due to market incentives.

Many questions remain unanswered. First, we have no direct evidence that explains why the urban-rural skilled wage gap fell substantially in the last few years of our sample, whereas the equivalent gap for unskilled wages remained essentially constant. To the extent that this has been driven primarily by more abundant skills in Metro Manila and contiguous regions, it implies that firms have found the location attractive. But it also suggests that opportunities have emerged for businesses to expand in rural areas where their added demand for skilled workers has helped close the wage gap. An investigation into the specific changes in policies, markets, or institutions that led to this result might suggest strategies that would further improve labor market integration between urban and rural areas, thus reducing spatial inequality.

Second, we do not yet know why secondary cities have relatively low skilled wages (Figure 7). This could be due to a shortage of investment opportunities for


skill-intensive firms in secondary cities. If so, we need to know more about the causes of the shortage.

FIGURE 7. Average nominal wages in 2011(skilled)

Third, we have shown that unemployment and underemployment account for very little of the urban-rural income and wage gaps, either for skilled labor, unskilled labor, or the average of the two. This is puzzling given that unemployment rates rise with education in the Philippines (Figure 8). We have also shown that those large urban-rural income gaps are driven mostly by education and experience. So why doesn’t an excess supply of educated workers in the cities lead skill-intensive firms to expand employment, thus driving down both the income and the wage gaps? Are skilled jobs rationed with their rates of pay sticky downwards?

Fourth, the relative stability of the unskilled urban-rural gap over the last decade must be explained by one or both of two forces: unskilled workers may not be exploiting the gap with much vigor; and firms are not exploiting it with much vigor either. Do constraints on worker mobility matter, such as poverty making investment in the move difficult, high minimum wages, labor regulations, or union restrictions? Are there barriers in capital markets that do not make it profitable for firms to invest outside well-established urban areas? And what about land markets?

NCR


700

elsewhere

600

500

400

300

200

100

0

PH

P p

er d

ay

}

495.2

648.1

549.4

11 other metros

}159 or 23%

112 or 16%


FIGURE 8. Unemployment by educational attainment

Joseph Capuno [2010] reported strong evidence of systematic spatial clustering of land values in the Philippines. That is, land prices tend to converge only within certain regions, especially in those areas of major population density such as the Metro Manila region. Capuno [2012:20] could only find evidence of statistically significant positive spatial autocorrelation for Metro Manila and three neighboring provinces; otherwise, he could observe no significant correlations in land prices. Moreover, he [2012:21] could find no evidence of spatial autocorrelation when considering the role of distance in land price integration forces.

It seems to us that the lack of wage and land price convergence across regions in the Philippines is not caused primarily by distance or geography but rather by the fact that different jurisdictions continue to have significant administrative autonomy, which raises barriers to market integration. These institutions are consistent with continued spatial segmentation of land and capital markets. Lack of labor market integration might, at least partially, be explained by mobility restrictions on firms, but also by constraints on land use.

Given the judicial and administrative constraints on the spatial mobility of firms and their land use, Philippine workers do an excellent job of sorting themselves to take advantage of the spatial opportunities. The fact that education, age, experience, and other individual characteristics remove much of the observed urban-rural wage gaps suggests fairly efficient labor market adjustment. The fact that firms do not locate or expand in areas with cheaper land and labor, or that trade does not promote national integration, seems to suggest that capital and



land market failure offer a better way to understand persistent urban-rural income gaps than does labor market failure. The persistence of spatial income gaps in the Philippines has much more to do with unequal spatial endowments than labor market failure. And the fact that individual endowments explain so much of spatial inequality in the Philippines suggests that local political economy forces are critical since they determine the quality of the education-health infrastructure.Given the large role that spatial infrastructure inequalities have in promoting the rural-urban gap, government programs to ameliorate these differences should come in the following areas:• Promotion of the development of secondary urban centers outside of the

national capital region;• Improvement of the investment climate, which may constrain companies from

expanding investment in existing urban areas that would be able to attract more rural unskilled workers and also retain many of the skilled workers who would otherwise work abroad as overseas Filipino workers;

• Improvement of infrastructure and transportation between areas identified as weakly integrated despite geographical proximity;

• Reconsideration of the current rules on the limited and highly politicized rezoning of agricultural lands to commercial so as to permit more investment;

• Relaxation of labor market rules that might hamper expansion of employment in existing businesses; and most importantly

• Promotion by the central government of programs that might serve to level the education-health infrastructure playing field between rural and urban areas.

*International Bank for Reconstruction and Development

**George Mason University and the Higher School of Economics in Moscow

***Harvard University, the University of the Philippines, and the University of Wisconsin

The authors gratefully acknowledge comments by and help with the data from Arsi Balisacan, Roger van den Brink, and Timothy Hatton. Most importantly, we thank Hal Hill for his years of support of research on the Philippines, including the research for this paper.


References

Akita, T. [2003] “Decomposing regional income inequality in China and Indonesia using two-stage nested Theil decomposition method”, The Annals of Regional Science 37(1); 55-77.

Albouy, D. [2008] “Are big cities bad places to live? Estimating quality of life across metropolitan areas”, nber Working Paper 14472, National Bureau of Economic Research.

Balisacan, A.M. [1999] “Reconstructing poverty profiles in the Philippines”, University of the Philippines School of Economics Discussion Paper, 9915.

Balisacan, A.M. and N. Fuwa [2004] “Changes in spatial income inequality in the Philippines: an exploratory analysis”, Research Paper 2004/34, unu-wider, United Nations University.

Berry, A. [1978] “Income and consumption distribution trends in the Philippines, 1950-70”, Review of Income and Wealth 24(3): 313-331.

Cameron, S., T. Merridew, and F. Pichler [2006] First European quality of life survey: urban-rural differences. Luxembourg: Office for Official Publications of the European Communities.

Capuno, J.J. [2012] “Agglomeration and sub-regional disparities under decentralization: evidence of spatial clustering of land values in the Philippines”, Review of Urban and Regional Development 24(3): 106-120.

de Dios, E.S. and J.G. Williamson [2014] “Has the Philippine forever lost its chance at industrialization?”, Philippine Review of Economics 51(2): 47-66.

Förster, M. and D.E.M. Mira [2005] “Income distribution and poverty in oecd countries in the second half of the 1990s”, oecd Social, Employment and Migration Working Papers 22:1-79.

Glaeser, E. and D. Mare [2001] “Cities and skills”, Journal of Labor Economics 19:316.

Gould, E.D. [2007] “Cities, workers, and wages: a structural analysis of the urban wage premium”, The Review of Economic Studies 74(2): 477-506.

Gray, D., J.A. Mills, and S. Zandvakili [2003] “Statistical analysis of inequality with decompositions: the Canadian experience”, Empirical Economics 28(2): 291-302.

Hatton, T.J. and J.G. Williamson [1992] “What explains wage gaps between farm and city? Exploring the Todaro model with American evidence, 1890-1941”, Economic Development and Cultural Change 40(2): 267-294.

Hertz, T., P. Winters, A. P. de la O Campos, E. J. Quiñones, B. Davis, and A. Zezza [2008] “Wage inequality in international perspective: effects of location, sector, and gender”, esa Working Paper 8/08. Rome, Italy: Food and Agriculture Organization of the United Nations, Agricultural and Development Economics Division.

Karan, A. K. and S. Selvaraj [2008] “Trends in wages and earnings in India: increasing wage differentials in a segmented labour market”, Asia-Pacific


Working Paper Series, International Labour Organisation Sub-regional Office for South Asia, New Delhi.

Minot, N., B. Baulch, and M. Epprecht. [2003] Poverty and inequality in Vietnam: spatial patterns and geographic determinants. Washington, dc: International Food Policy Research Institute.

Mishra, P. and A. Parikh [1992] “Household consumer expenditure inequalities in India: a decomposition analysis”, Review of Income and Wealth 38(2): 225-236.

Morris, E. [2007] Promoting employment in Cambodia: Analysis and options. International Labour Organization, ilo Subregional Office for East Asia.

National Statistics Office (nso) [2012] Labor Force Survey (2001-2011) data file.National Statistics Office (nso) [2012] Family Income and Expenditure Survey

(2003, 2006, and 2009) data file.Novotný, J. [2007] “On the measurement of regional inequality: does spatial

dimension of income inequality matter?”, The Annals of Regional Science 41(3): 563-580.

Ragayah, H.M.Z. [2008] “Income Inequality in Malaysia”, Asian Economic Policy Review 3(1): 114-132.

Sicular, T., Y. Ximing, B. Gustafsson, and L. Shi [2007] “The urban-rural income gap and inequality in China”, Review of Income and Wealth 53(1): 93-126.

Simler, K.R. and V. Nhate [2005] “Poverty, inequality, and geographic targeting: evidence from small-area estimates in Mozambique”, International Food Policy Research Institute, fcnd Discussion Paper, 192.

Tadjoeddin, M.Z., W.I. Suharyo, and S. Mishra [2003] “Aspiration to inequality: regional disparity and centre-regional conflicts in Indonesia”, In uni/wider Project Conference on Spatial Inequality in Asia, Tokyo, Japan (pages 28-29).

Todaro, M.P. [1971] “Income expectations, rural-urban migration, and employment in Africa”, International Labour Review 104(5).

Weber, A.F. [1899] The growth of cities in the nineteenth century: A study in statistics (Vol. 11). Published for Columbia University by Macmillan.

Williamson, J. G. [2013] “Demographic dividends revisited”, Asian Development Review 30(2): 1-25.

World Bank (wb) [2013] Philippine development report: Creating more and better jobs. Philippines: Asia-Pacific region.

Yankow, J.J. [2006] “Why do cities pay more? An empirical examination of some competing theories of the urban wage premium”, Journal of Urban Economics 60(2): 139-161.

Young, A. [2013] “Inequality, the urban-rural gap, and migration”, Quarterly Journal of Economics 128(4): 1727-1785.


PRE

American private direct investment in the Philippines after independence

Gerardo P. Sicat

The Philippines has lagged behind East and Southeast Asian economic growth successes in the course of contemporary times. A historical perspective is presented to explain this lag. When the country entered the definite road to independence from American rule in 1934, it adopted a political Constitution that limited foreign capital to a minority role in three industries – land, public utilities, and the exploitation of natural resources. During the colonial period and after political independence in 1946, American foreign direct investment in the country was dominant and large. This continued after independence as US citizens were granted equal rights of Filipinos and American businesses received war damage payments to rehabilitate their investments. This was further helped by bilateral preferential trade with the United States during twenty-eight years of economic adjustment. In actual practice, the restrictive provisions in the Constitution invited many virulent forms of nationalistic economic policies of exclusions of foreigners in other lines of economic activity and led to inward-looking economic development policies. These policies discouraged American foreign direct investments causing disinvestment in key areas. The restrictive economic provisions in the Constitution continue to this day even as the economy is opening more to world markets.

JEL classification: P45, D9Keywords: Philippine economic history and development, foreign direct investment,

American foreign direct investment, Philippine-American economic relations

In Hal Hill’s work on the Philippine economy, he has called attention to the following conundrum: Why does the Philippines, despite its many good attributes, lag behind in economic development when compared to East Asian high economic


achievers?1 In this essay, I hark back to what I have called the “original sin”, the biggest long-term mistake in our nation-building: the restrictive economic provisions in the Philippine Constitution.2

It is not as much a mystery. When we take a historical perspective, we link developments to the question of why American and other foreign direct investments have not come in sufficient volume to help push the growth of industry, incomes, employment and productivity. Those restrictive provisions led to protective and inward-looking policies that characterized subsequent policies promoting industrialization and economic development. In that respect, they served as a lever for bad policies that moved the country in many wrong directions.

In turn, because those provisions further strengthened the vested interest in the continuation of these policies, their reform could not take hold. Today, the problem of amending these provisions in the Constitution remains as a principal problem hindering still future progress.

Before the Second World War, United States private industrial capital in the Philippines was dominant. The public social and economic infrastructure built by the colonial administration was extensive and growing. The country was the envy of early economic development in East Asia among colonized economies. When the grant of full political independence in 1946 happened immediately in the aftermath of a ruinous war of occupation by Japan, the country was economically devastated. However, the rehabilitation from that devastation was relatively quick.

Total us direct investments3 in the world at yearend in 2012 is estimated at us$4.45 trillion. Three-quarters (74 percent) of this total is located in highly advanced countries. Only 14 percent of American direct investments is located in Asia, principally also in the developed countries of the region. In this context, direct investments to the Philippines were hardly a speck in the numbers. A recent work summarizing data on flows of foreign direct investments (including us direct investments) going to Southeast Asian countries from 1970 to 2011 reports that Singapore, Thailand, Indonesia, and Malaysia have been the largest receivers of foreign direct investments. It also observes that the Philippines has consistently received a relatively small amount of these investment flows over the same period.4

To those drawn into the discussion on investment and economic policy in the Philippines, the relative insignificance of foreign direct investment numbers is all too familiar.

1 Hal Hill expounded on this point in seminars he delivered at the University of the Philippines School of Economics. His work on the Philippines is sometimes indistinguishable with that of Arsenio Balisacan because they have worked together in major pieces. 2 The author tried to elaborate on this point in several short essays found in Sicat [2013:186-187]. In a presentation on the topic of restrictive economic provisions on foreign capital before an annual meeting of the Philippine Economic Society circa 2004, he called these provisions the “ original sin” of development policy.3 See Jackson [2013]. 4 See Sjöholm [2013].

24 Sicat: American private direct investment in the Philippines after independence

1. The triad of us-Philippine relations

The Philippines was one of the most devastated economies at the end of World War II. War confiscations and conscriptions, market disruptions, inflationary spiral especially towards the closing months, destruction due to the battle for liberation—all these led to output and physical capital losses that were substantial. The total output of the country at the end of the war in 1945, was reduced to 30 percent of the level of the pre-war gdp.5

Before the start of the World War II in the Pacific, the preparations for Philippine independence were in full swing. In 1934, the us Congress passed the law that spelled the road toward full political independence. A year later, the political Constitution (which had built into it the restrictive economic provisions on foreign investments) that was crafted by an elected national convention composed of Filipino leaders was approved by the American president as required. Hence, in 1936, the ten-year transitional Philippine Commonwealth took office. The Commonwealth stage was a semi-independent political status under the American dominion before the grant of full independence on July 4, 1946 as the Republic of the Philippines. This path toward independence was interrupted by a destructive war of occupation by Japan covering four years.

Economic rehabilitation from the war proceeded rapidly during the next decade after independence, starting in 1945 when Japan’s occupation ended. Aside from a sustained volume of American military expenditures early in this period, three major actions propped up the economy. Such measures would involve deep economic entanglements of the Philippines with its former colonial ruler. These actions—the triad of relations—were to influence future events strongly.

First was a trade and economic adjustment to wean the country from its dependence on the us economy, an outcome of the long American occupation of the country. The us Congress enacted the Philippine Trade Act of 1946 (us Public Law 371, 79th Congress) to amend its trade and tariff policies to accommodate the new parameters that would be contained in the commercial trade agreement with its former colony. The law granted the Philippines a period of declining trade preferences and fixed quotas on certain products like sugar, coconut oil, and other major products that would end in 1973. In the early discussions of this adjustment period before the war, the proposed adjustment period was shorter, twenty years being the longest. The war destruction in part led to this extension of the period of adjustment to twenty-eight years.

Second was the payment of war damage transfers6 to assist in speeding up the rehabilitation of the domestic economy. The war damage payments were embedded

5 See Sicat [2015:191-204]. 6 The author uses the word “transfers” in contrast to war damage payments as reparations, which connotes monetary retribution as payment made by a defeated party after the war, as in the Japanese reparations. The us considered the payment of war damage claims (which had to be formally declared first) as assistance to help the recovery effort, more in the context of the Marshall Plan for Europe after the war.


in the law called the Philippine Rehabilitation Act (Public Law 370, 79th Congress), which contained the war damage payments and the mechanisms for delivering the program. This companion law on Philippine independence appropriated a large sum of money—us$520 million—in two parts: us$120 million to help finance and restore public buildings, roads, and social infrastructure; and us$400 million to pay war damage compensation to assist in the rehabilitation of the economy that heavily suffered during the war.

Finally, the American government required that before these new provisions were to take effect, the new Philippine republic had to give Americans the same rights as Filipino citizens. This meant in polite treaty language, “national treatment”, a common feature of commercial agreements between states. Essentially, this meant protection from confiscatory national policies and guarantee from discriminatory policies. The restrictive provisions in fact were directed at foreign control of the economy, which was dominated by American investments at the time. President Franklin Delano Roosevelt approved the Philippine Constitution as prelude to the process leading to Philippine independence early in his first term in office at the bottom of the Great Depression in 1934. Faced with more pressing economic issues at home, he approved the restrictive provisions in the Philippine Constitution. It therefore would fall on Harry S. Truman (the successor to the presidency upon Roosevelt’s death in 1944) to exact parity rights as a condition to protect American investor interests after Philippine independence. For the Philippines to yield to this demand, a constitutional amendment had to be secured by plebiscite. The plebiscite was made and the government of Manuel A. Roxas, the first president of the Republic, triumphed in his sponsorship of the yes vote.7

The 1935 Constitution limited foreign capital participation in three economic markets—utilities, land, and exploitation of natural resources—to a maximum of 40 percent, meaning that such industries must have 60 percent Filipino control of the capital equity. This was a key intent of the “nationalistic” provisions of the Philippine Constitution when it was enacted in 1935 in reference to corporate activities. Had the basic law of the land simply remained silent on this issue, it would still have been possible to enact nationalistic provisions, but they would not need to be cast in solid stone to accommodate changes in national moods and perception.

Thus, a package of three important economic actions was needed before the road toward full economic rehabilitation and adjustment of the newly independent republic could proceed. To nationalistic Filipino leaders, the grant of equal rights to Americans was economic blackmail. To the Americans, these actions were pragmatic elements to help a former colony recover as well as to protect the rights of its citizens who continued to have vital interest in the Philippines. To the incumbent, newly elected leaders of the Philippine government (led by

7 The parity rights amendment was held on March 11, 1947 with 79 percent of voters casting approval.


President Manuel A. Roxas, who was then the newly inaugurated first president of the republic), this was a realistic compromise to ease the way for the newly independent country.

Another element of relations with the us at the inception of the Philippine republic was the military bases agreement. This could be considered the fourth element of the post-independence conditions with the former colonial master. The us was given a military foothold for 99 years to operate military bases in the country. Two major military bases were principally included in this agreement, the Clark Air Base and the Subic Naval Base, in addition to a number of military installations that were installed during the long American colonial period.

From an economic viewpoint, this agreement helped to provide for the country an important American military shield. A steady flow of earnings of foreign exchange resulted from us military operational spending in the bases throughout the post-war period when the military bases were stationed in the country. From a defense standpoint, the newly independent country was kept from undertaking a heavy domestic expenditure on defense matters. Defense expenditure was a major expense among newly independent countries.

2. The war damage payments by citizenship of recipients

The war damage payments made to American citizens give a clue to the amount of American private capital in the Philippines. The losses were therefore compensated at the original value of acquisition. When the payments were made, the replacement cost of capital was already three times the cost of capital before 1940. They were disbursed in pesos at the fixed exchange rate of 2 pesos to a dollar.

About two-thirds of the war damage compensation was to pay war damage to the private sector on the basis of claims of war damage filed before the Joint Philippine-American War Damage Commission8 (henceforth referred to as the Commission). The first job of the Commission was to pay as much of the small claims first. Approximately one million claimants, out of a total of 1.25 million claimants, filed claims for us$500 or less. According to the final report9 of the Commission, these claimants were paid a total of us$131,766,435 as of end of December 1950, or, converting the amount at the 2 pesos to one dollar fixed exchange rate, P263,532,870. The total amount of these small payments amounted to 32.9 percent of the us$400 million that was appropriated for private payments. It was safe to assume that all the small payments were received by Filipino citizens.

8 The Commission had three members: two Americans and a Filipino, with one of the American commissioners serving as chairman. The activities of the Commission and their actions with respect to implementing its mandate are explained in the main study. 9 See Philippine War Damage Commission [1951].


The question of how much war damage compensation was received by Americans meant that that the citizenship of the recipients had to be known. The law on the war damage payment was only to qualified American and Filipino citizens and citizens of other nations that were not enemies, to include also corporations owned by such citizens. The report of awards did not track the citizenship of the awardees. In order to make an estimate of awards by citizenship, an effort is made to classify the citizens that filed claims and who were issued approvals for those claims.

The task was to classify by citizenship of awardees of the “large” payments, that is, those exceeding us$500 and above. To get a distribution by citizenship of all these awards, I had to tag each of the recipients—individuals or enterprises—with the appropriate citizenship: Filipino, American, Chinese, and the like. The greatest area of doubt was in determining whether some recipients were Filipino or American citizens.

The individual recipients of war damage awards could be fairly identified by their names, although this was not fully accurate. The names of business establishments were another matter. Most of the establishments were corporations. From the names of these recipients, I tried to form some basis for the citizenships by identifying each recipient with a citizenship label. The classification by citizenship was essentially educated guesswork to determine overall magnitudes that yielded approximate orders of magnitudes.

I labeled familiar Filipino names “Filipino” by citizenship. American names, and those I associate with American enterprises, I labeled “American”. Chinese names were easy to identify when they used proper names. Uncertainty was greatest in identifying the citizenship of enterprises, for which I could only add the label “American or Filipino”. These recipients were separated from those tagged more definitely as either American or Filipino. In all these labeling efforts, acceptable errors were likely to be made.

The corporate names of the firms were more challenging. There was little to go by that could give the citizenship of the ownership of corporate enterprises. Sometimes the corporate names gave a clue. This was clear enough for some businesses, as when the corporate names were emblazoned with Chinese names or titles. Most of the time, an educated guess sufficed, for instance, from studies listing industries and establishments10, from personal and imperfect memory growing older, and from introspection on the business of the corporation and the likely citizenship of the owners.

10 The study of Kunio Yoshihara [1985] on Philippine industrialization used some names of Chinese, American, and Filipino enterprises during the 1960s. Many of these firms existed before the war. A more assiduous study would have been to use an establishment set of names from the Securities and Exchange Commission in an earlier period or some listing from the Bureau of Commerce and Industry. I had no resources for hiring research assistants, and neither did I have the luxury of time to study this aspect of the problem. In any case, relative accuracy, not definite accuracy, was what I was after.


American investments were heavily in sugar centrals, mining, logging and timber exploitation; in public utilities such as power generation and distribution, transportation, both land, sea and air; in domestic and international trading; in manufacturing operations that processed agricultural products; in construction and metallic fabrication; and in banking and finance. They were typically the larger industrial enterprises operating in the country, although there were also some Filipino-owned industrial enterprises in similar fields.

There were many instances when it was difficult to tag the enterprise definitely, in which case they were identified as “Filipino or American”. Exact judgment depended on subjective judgment, a dubious but necessary method in order to establish a reasonable measure of allocation in the absence of total information. For a norm, that probability could be 50-50. An optimistic judgment that the group of “American or Filipino” recipients thus tagged contained more Americans represents “60 percent likelihood of being American”. A less optimistic assessment on the opposite side could be one in which 60 percent likelihood of being Filipino”. I use these subjective judgments in providing a “high” and “low” estimate of American citizen recipients.

Finally, there were awardees whose citizenship did not belong to the three above-mentioned. These awards were made principally to religious institutions that received war damage compensation. I labeled most of these “Other” citizens. Catholic Church institutions could have been grouped under “Filipino”, but they fall under the hierarchic control of the Vatican. Therefore, all the war damage payments to archdioceses, bishoprics, and foreign Catholic religious institutions that received awards were classified as “Other citizens”.

3. Summary of war damage payments by citizenship

Table 1 presents a comprehensive table. Part A is the raw table showing the citizenships of the large recipients of war damage as approved by the Joint Philippine-American Commission. The citizenships of recipients separated those grouped under American, Filipino, “American or Filipino”, Chinese, and Other.

Part B of the table uses the “high” scenario in which American citizens received 47 percent of the large payments made by the US government. In this estimate, Filipinos received 42 percent. These numbers are seen in the last column under Part B. The ratio of awards to the amount of claims (seen in the second to last column) is slightly higher for American awardees, probably not out of bias but by virtue of the greater reliability of claims made by larger claimants—corporations with much better record keeping.

Part C of the table shows the “low” scenario result. From the viewpoint of the immediate impact on the economic rehabilitation efforts, the issue of who received more of the war damage payments between Filipino and American citizens did not matter as much. However, this exercise is helpful in identifying the amount of private American capital in place in the country before the war.


There were no reliable estimates of American private investments that I could find before the war. An estimate due to the American Chamber of Commerce in the Philippines was P537 million, of which P200 million was in sugar, coconut, and abaca plantations, mining, and lumber operations. (In us currency, this was us$268 million at the fixed rate of two pesos to a dollar.11 ) Another estimate of us$160 million in direct investments and us$36 million held in Philippine bonds and securities (or nearly us$200 million) was due to a presentation made by the us National Foreign Trade Association president before a hearing on Philippine independence issues.12 These were very sizable magnitudes of investment in an economy that was undergoing steady economic development before the war.

TABLE 1. Summary of US war damage payments (“large awards”), by citizenship of awardees

Citizenship of awardees Value of claims submitted in thousand 1940 pesos

Amount of war damage approved in thousand 1940 pesos

Ratio of award to claims

Percentage distribution of awards to total

A. Raw estimatesFilipino 203.02 87.41 43.1% 30.2%American 166.46 85.00 51.1% 29.4%American or Filipino ("AF") 161.77 85.90 53.1% 29.7%Chinese 26.87 7.67 28.5% 2.6%Others 50.50 203.02 46.7% 8.1%TOTAL 608.63 289.56 47.6% 100.0%

B. High estimates (Americans 60% of "AF")Filipino 267.73 121.77 45.5% 42.1%American 263.53 136.54 51.8% 47.2%Chinese 26.87 7.67 28.5% 2.6%Others 50.50 23.59 46.7% 8.1%TOTAL 608.63 289.56 47.6% 100.0%

C. Low estimates (Americas 40% of "AF")Filipino 300.08 138.95 46.3% 48.0%American 231.17 119.36 51.6% 41.2%Chinese 26.87 7.67 28.5% 2.6%Others 50.50 23.59 46.7% 8.1%TOTAL 608.63 289.56 47.6% 100.0%

D. Estimated high estimates calibrated to 2015 US Dollar (based on Part B estimates)Filipino 4,015.92 1,826.52 2,007.96 913.26American 3,952.95 2,048.05 1,976.47 1,024.02Chinese 403.08 115.00 201.54 57.50Others 757.53 353.84 378.77 176.92TOTAL 9,129.48 4,343.40 4,564.74 2,171.70

Source: From basic table of war damage payments. G.P. Sicat, “US War Damage Payments Income Flows After Independence and Philippine Economic Development,” UP School of Economics, forthcoming, 2016.

11 See Hartendorp [1953]. 12 See United States [1938].


My objective in presenting these estimates is to show that for the economic rehabilitation during the early years of the Philippine republic, a sizable amount of American capital was returned to their proper owners. Since the war damage payments were linked with the rehabilitation of the damaged economic establishments, this return of American capital also signified the continuing commitment to the future of the Philippine economy. The war damage payments also helped to identify orders of magnitude that had a direct connection to the amount of capital in place before the war.

The Commission in its final report indicated that replacement cost of damaged capital at the time of payment was three times more than before the war, even though the approved payments represented only about 52.5 percent of the submitted claims. When replacement cost was therefore taken into account, the recipients effectively received only around 20 percent of the value of replacement costs. The award of war damage payments could therefore act as supplying merely a significant stimulant for the rehabilitation process. On this point, the report of the commission said: “They permitted many industries and other activities to attract new capital or obtain credit to start rebuilding the industrial commercial, and other agricultural enterprises which provide employment and otherwise benefit the Philippine economy.”13 Some slippage from this objective or procedure might have occurred, but most of the payments of war damage were made after proof of rehabilitation work was made available.

In other words, although the war damage payment might have constituted a fifth of the total investment requirement to rehabilitate an enterprise, in effect it performed a multiplier impact on the rehabilitation effort. Such an amount of new payments would have implied an expenditure multiplier close to 5, but it surely was much less since the payments would have triggered a high bill of imports of materials and equipment not available in the domestic economy. The country’s productive capacity had been rendered very low due to the ravaged productive capacity after the war.

Finally, a point to be made here is that the amount of money transferred to private pockets at that critical time was very significant in relative value even when viewed today. Using the price multiplier for the value of the US dollar in 1940 compared to the much inflated 2015 us dollar, here are some results.14 The us$520 million valued at the 1940 us dollar (which was allotted for public and private war damage) is roughly the equivalent of us$7.8 billion in 2015 dollars.

13 See Philippine War Damage Commission [1951]. The report further stressed that “before second and final payments were made on claims for which combined payments would amount to us$25,000 or more, the Commission … communicated with the claimants to make certain that the first payments had been expended in accordance with their sworn statements and that, second payments would also be used for rehabilitation.” 14 I use an inflator-multiplier of 15 to compare the pre-war us$ to the current us$ for 2015, taking into account the changes in prices in 1940 to the present in the Philippines. In the paper calculating the war damage from the Japanese occupation (see footnote 6), I used the same inflator. In this context, these calculations are relatively conservative.


Since the amount was converted at the fixed exchange rate of 2 pesos to one dollar, the equivalent of this sum in 1940 Philippine pesos was P15.6 billion. The amount of us$400 million which was appropriated for private war damage compensation, using the same approximations, meant us$6 billion in 2015 us dollars. In the peso values of 1940, this amounted to P12 billion. (A proper adjustment would of course also be reckoned in the much depreciated pesos of 2015, which I do not show.) By whatever measures, these numbers are very significant, especially as they were made for a newly independent country with a population size of 19 million. (In 2015, the Philippine population is nearing 100 million.)

Part D of Table 1 shows the same calculations for the amounts of payments made by citizenship of the awardees.15 Following the discussion for the “high” case estimates of American awardees, we specifically track the amounts paid (shown in last column of Part D). American citizens received approximately us$1,024 million, or around us$1 billion in US dollar values of 2015. The total amount of payments for all awardees was US$2.171 billion measured in 2015 us dollars. (The peso values of the awards that are calibrated to 2015 us dollars but are converted at the 2 to 1 dollar exchange rate of the 1940 pesos are shown in column 3 of Part D.)

4. Three reasons

From the viewpoint of the newly independent nation, the trade adjustment of twenty-eight years brought about a stable relationship in trade and economic relations with the strongest economy in the world at an opportune time of global growth after World War II. Further, parity rights, which gave Americans the same opportunities as Filipinos, promised to continue attracting American investments. At the same time, the extraordinary war damage payments provided new income flows essential for rehabilitation. In particular, as it pertained to American investments, the income flows awarded to Americans were linked to the rehabilitation of their damaged businesses in the country.

This question then arises: Why has American capital not come to the country in droves after the war and during the long interim period to the end of special relations in 1973? Three reasons come to mind why this has transpired: • First, the restrictive provisions in the Constitution abetted other policies that

further relegated the role of foreign capital in other parts of the economy, thereby rendering the development process more difficult and perilous;

• Second, the evolution of American firms in the country reacted to the time-bound adjustment process more as a signal to divest rather than to be more fully engaged in the economy; and

15 Since I am interested only in understanding the orders of magnitudes for the war damage payments then, I simply return to the 1940 fixed exchange rate of the us$ to the peso. The average exchange rate for the pesos to the dollar in 2015 is P46 per us$1. The contemporary pesos would be very much depreciated.


• Third, inward-looking development strategies were implemented by industrialization efforts based on import substitution. To implement it the government undertook nationalist policies that equated promotion of industry with the exclusion of foreign economic participation.

4.1. Restrictive policies

The restrictive economic provisions were designed to spell out the boundaries of foreign investment participation in certain economic sectors, implying that those not so mentioned were to welcome foreign capital. In its place, the legal language of restrictive provisions provoked measures of economic nationalism that encouraged exclusion or limited participation rather than suggest the attraction of foreign capital to speed up development in the rest of the economy. In practice therefore, the provisions incited other restrictive economic policies.

A wider network of economic policies limiting foreign investments in other sectors ensued during the early years of political independence. Commercial banking was closed to foreign participation, except for those already allowed before. The licenses for new banks were issued only to Filipino capital even under low capitalization requirements. For many years until the 1970s, no new foreign banks were allowed to enter the economy, except the four foreign banks that were in operation at the time of independence. It would take the 1990s, when more openness to the participation of foreign banks in the country would be allowed.

When import and foreign exchange controls were instituted in 1949, the allocation decision became one of administrative selection rather than market solution.16 In an environment of economic controls, selectivity implied that resource allocation happened through regulation and administration rather than through the market. In this climate of policy, industrial and agricultural projects were the result of selection criteria in which nationality was a critical element in the choice.

A further consequence of the restrictive economic provisions was that they encouraged policies that were discriminatory over foreign economic interests. In 1955, retail trade nationalization was passed by Congress. The new law placed restrictions that excluded foreigners from the retail distribution sector of the economy. The measure was designed to encourage Filipino retailers to take over the small sari-sari corner store from the Chinese merchants. But the retail distribution network extended to big stores too, and this led to the exclusion of foreigners from the retail distribution industry.

16 See Golay [1961] for a longer discussion of nationalistic measures during the first two decades of independence.


The rhetoric of industrial and economic policy in the 1950s was “Filipino First”, echoing the demands of those who wanted more protection for Filipino enterprises and the exclusion of foreigners from the nation’s economic development. Though such rhetoric was removed from the official language in the investment promotion programs, the industrial and investment priority schemes for new industrial projects emphasized that priority areas of investments were reserved for Filipino enterprises under the Board of Investment. Only industries classified as “pioneer” were listed as investment areas where foreign direct investments would be allowed up to 100 percent equity from foreign enterprises. In many East Asian countries, especially at the beginning of their industrial promotion efforts, businesses that were 100 percent owned by foreigners were typically allowed to receive government incentives in order to attract new foreign capital.

4.2. The divestment of American firms

Before the war, American investments were strong across the spectrum of economic sectors. Specifically, they were dominant in agriculture (sugar, coconut processing, abaca, and other plantation agriculture), natural resources extraction, especially mining and forestry, and in public utilities. Targeted constitutionally by the restrictions on foreign ownership, the grant of parity rights meant that they could remain in American hands even after independence. However, the future prospects for these investments in spite of parity rights became bleak in the face of nationalistic regulations and policies.

The strength of the economic rehabilitation efforts after the war emanated partly from the participation of American companies that were awarded war damage payments. Agricultural enterprises, especially in sugar and coconut industries, were able to recover easily and enabled export earnings to begin to recover. These were big contributors to the export sector. The same story applied to the mining companies that resumed their operations after the war. The re-entry of mining firms into the production picture helped the mining industry recover.

The public utilities were among the biggest recipients of war damage payments and their restoration of service was quick. Meralco—the electricity retail distribution monopoly in the Manila area—was one of the biggest recipients of war damage. The company was able to undertake resumption of service and to re-expand destroyed facilities. The telephone and telecommunications sector re-established services. These utilities were most exposed to dealings with the local regulators.

By the late 1960s, American companies in the country began to feel the pinch. Nationalistic economic policies and regulations foreshadowed future difficulties for them, and so it was better to move ahead of 1974. Even if parity rights gave American investors some comfort as they faced the future, it was a kind of restless comfort. As the period of special economic relations was coming to a close at the historical time line of 1974, the end of parity rights was written on


the wall. Eventually, the public utilities saw the future as bleak for full control of management, and they sold to partners toward the years before the end of parity rights.

During this period then, the incentive to reinvest and to expand existing capacities began to weaken as more restrictive economic measures were added to those already in place. Thus, the stream of divestments began. Some American investors sold the land they owned to other Filipinos. Companies that had a desire to remain in business accepted the implications on land policies. They created a land company that was owned by their workers’ provident funds to which they sold the land. In turn, the land company that was owned by at least 60 percent by Filipinos leased back the land for the company’s use. This roundabout method of dealing with the land question made investors essentially temporarily satisfied, but it added complications to the company’s structure and, hence, its decision making. Even this solution did not please those that wanted complete divestiture. In the latter case, the local partners often had the inside track and benefited handily. The lack of a strong market to buy up these assets meant that the local partners had the upper hand and bought the assets at favorable prices.

Some American investments found business opportunities in the domestic economy, and they ventured further into purely local markets. New investments took place behind the logic of import substitution, but they reflected small-scale commitments of new capital since the domestic market was limited. Formerly imported goods had better economic logic being assembled for the domestic market so as not to lose the market. This was the case for many industrial products that thrived in food, cosmetics, pharmaceuticals, and energy refining, where the buyers were home consumers. The end of parity rights was not a problem for these industries for they had a comfortable position participating in the domestic market. The same domestic protection given to local industries was accorded these investments. American companies that had an early foothold before with imports of these goods established their manufacturing and processing plants in order to keep their domestic market share. This industrial foothold anticipated a future in the country’s growing market. But these industries were relatively small in themselves because they simply undertook production for the local market like most of the industries promoted by the protectionist industrial policies.

Plantation agriculture in old lines of agriculture continued in part because of the early economic foundation laid by export industries destined for the American market. This was the case for the quota-based industries, such as sugar and the traditional coconut-based processing industries, and in the case of the pineapple industries, which had natural markets in other countries as well. The banana industry was such a case, with markets intended for third countries like Japan. However, the banana industry—which started originally as a proposal to set up an investment by the United Fruit company—got shot down. It was resuscitated when domestic interests took over production while they allowed marketing functions


to the American investors. Dole Plantation got an earlier foothold, with large land areas made available through government lease at the beginning. Later expansions led to innovations in contract farming with local landowners. What has thrived well was the profitable contract farming arrangements with local planters, which involves other technical and other production tie-ups and marketing agreements. In general, because of land reform issues which limited future land use to smaller plots of land, even plantation agriculture became almost a closed field for investment expansion.

It was only after the late 1970s that more enthusiastic American capital began locating in new investment areas. These new investments arose from the globalization of the international division of labor in the electronics industry. Philippine entry into this lucrative opening for foreign direct investments in general was the establishment of export processing zones. These export processing zones enabled industrial locators to speed up export or import processing, the acquisition of land for factories, the isolation from much bureaucratic red tape and control measures that are often associated with tight, inward-looking policies. By the mid-1990s, the investment attraction framework within export processing zones accelerated further and, finally, the Philippines is able to again attract new flows of foreign direct investments, including American capital. In short, it opened up industry for the participation of a wider range of foreign direct investments without being hampered by the restrictive economic policies of an inward-looking industrial framework. In this, American capital had found a way to come back more strongly.

Despite the improved capacity to attract foreign capital-engaged electronic parts assembly, the country lost its competitiveness in expanding the garment-making industries. In this case, the economic policies that continued to hamper the growth of this industry were tied up with labor market policies, especially those linked with minimum wage setting. American investment in garment making was a strong presence very early during independence and should have expanded more. But the import substituting framework for industrial growth favored the highly protected textile industries and disfavored the importation of raw materials from the world for the garment makers. And then, rising domestic wages for formal employment in organized manufacturing further edged out industries engaged in exports that were heavily labor-intensive.

4.3. Inward-looking policies

The inward-looking approach promoted measures to replace imported goods rather than an industrialization that was based on comparative advantage. This program was a hollow strategy. Most of the promoted industries using domestic capital were mainly import-dependent operations that assembled imported raw material inputs. With little domestic feedback on production and lacking scale in production size, these new industries simply produced profits for their investors


while raising the cost of the products to Filipino consumers. The opportunities that were built on tariff protection and other forms of subsidies to domestic industrialists also worked for import-substituting American investors.

The protectionist industrial policies emanated from the early import and exchange controls and by tariff protection. These policies created a large group of domestic industries that depended mainly on protection. As a result, future efforts to reform the framework to allow for amendments that opened the industries to more competition encountered strong opposition.

The special trade preferences, for instance, heightened the economic isolation of the country in international trade. The special preferences made it difficult for the country to multilateralize its approach to international economic issues. The special concessions to trade under the us-Philippine bilateral tariff preferences and quotas for specific industries (sugar and coconut industries) created the excuse to resist multilateral trade policies. Thus, the country did not join the General Agreement of Tariffs and Trade.17 Industries that benefited from tariff preferences were more concerned about enjoying those preferences. Of course, the General Agreement of Tariffs and Trade was for multilateral trade and not bilateral preferential trade. And it would be many decades later, in 1994, at the birth of the World Trade Organization, the successor to the General Agreement of Tariffs and Trade, before it could join fully the world’s trade rules.

In one sense, the Philippines was uniquely blessed with special preferential treatment with the United States for that market represented an infinite potential for economic growth, if the trade opportunities were fully exploited for domestic reasons. Under the revised Laurel-Langley Agreement, the terms of tariff preferences were even better than those originally under the original agreement. But the import-substitution climate of domestic industrial protection and extreme nationalistic measures influenced by the original restrictive economic provisions of the Constitution swamped all reasons for exploiting an open trading system.

Without the blessings of a similar preferential trade agreement with the United States, the economies of South Korea, Taiwan, Hong Kong, and Singapore (and later East Asian countries)—by simply espousing a more open foreign direct investment policy for themselves—were able to reap major benefits by attracting heavily American direct investments which manufactured labor-intensive consumer products for the us markets. Because of the unlimited growth prospects that were opened for their exports to the rich American market, these countries realized their industrial revolutions much sooner and more quickly than the Philippines.

University of the Philippines School of Economics

17 The multilateralization of trade commitments would only come about with the entry of the Philippines into the World Trade Organization. The Philippines had participated in the negotiation of the later rounds of multilateral tariff negotiations, beginning with the Tokyo Round and the Uruguay Round that ultimately led to the formation of the World Trade Organization in 1994.


The author is Professor Emeritus at the University of the Philippines School of Economics. This paper is part of his longer study of the US war damage payments to the newly independent Republic of the Philippines after World War II. Collection of data arose out of an unplanned research effort in 2003 when the author found the individual records involving the recipients of war damage payments in the official minutes of meetings of the Joint Philippine-US War Damage Commission that were lodged in the US National Archives in Washington D.C. The festschrift for Hal Hill was the immediate incentive for him to restart the study. The author uses segments from this study to restate an old theme concerning the missed opportunities arising out of Philippine-US economic relations.

References

Balisacan, A. and H. Hill [2003] “Introduction”, in A. Balisacan and H. Hill, eds., The Philippine economy: development, policies and challenges. Quezon City: Ateneo de Manila University Press.

Golay, F.H. [1961] The Philippines: public policy and national development. New York: Cornell University Press.

Hartendorp, A.V.H. [1953] History of industry and trade of the Philippines. Manila: American Chamber of Commerce.

Jackson, J.K. [2013] D.C. direct investment abroad: trends and current issues. Washington D.C.: Congressional Research Service,

Philippine War Damage Commission [1951] “Rehabilitation of the Philippines: final and ninth semiannual report of the United States Philippine War Damage Commission, Manila, Philippines, March 31, 1951.”

Sicat, G.P. [2013] Weighing in on the Philippine economy and social pderogress. Mandaluyong: Anvil Publishing.

Sicat, G.P. [2015] “The Philippine economy during the Japanese occupation, 1941-1945”, in M. Boldorf and O. Tetsuji, eds., Economies under occupation: the hegemony of Nazi Germany and imperial Japan in World War II. London & New York: Routledge. 191-204.

Sjöholm, F. [2013] “Foreign direct investments in Southeast Asia”, Working Paper Series 987, Research Institute of Industrial Economics.

united states (US) [1938] Joint preparatory committee on Philippine affairs, report of May 20, 1938. Washington D.C.: Government Printing Office.

Yoshihara, K. [1985] Philippine industrialization: foreign and domestic capital. Singapore: Oxford University Press.


PRE

Singapore’s five decades of development: lessons and future directions

Chia Siow Yue

This paper looks at Singapore’s transition from a Third World economy at the time of political independence in August 1965 to a First World economy five decades later. It is a unique city-state with a dearth of land and natural resources and pursuing a government-led, mnc (multinational corporation)-led, and export-led development strategy. Government and governance have been outstanding, and the city-state overcame its constraints by adopting open trade and investment policies, making the region and the world its economic hinterland. Five decades later, while facing the same demographic and land constraints, it has to overcome the challenges of transitioning into a private-enterprise-driven and innovation-driven economy.

JEL classification: F1, F2, H1, J1, O1, O2Keywords: Singapore development model, economic policy, social policy,

city-state economy

1. Introduction

Singapore gained its political independence on August 9, 1965 and celebrates its 50 years of nationhood in 2015 with much introspection as well as aspirations in going forward. Notwithstanding its small nation status, its rapid transition from a Third World to a First World economy has attracted worldwide attention from the scholarly, policymaking, and business communities.

In 1965, the economic and political challenges facing Singapore were more challenging than those faced by many newly independent developing economies. Its city-state economy was small, with land area of only 580 square kilometers and a population of less than 2 million; there was a dearth of natural resources, including crucial water and energy. Political independence had cut off its Malay Peninsula hinterland and its two economic pillars—entrepôt trade and British military base—faced dismal prospects. There was high unemployment


and budget deficits, militant trade unions, and ethnic unrests. Many thought Singapore was a “basket case”.

However, there were also some positive initial conditions. Singapore inherited the British legal system and use of the English language in government and business and has a wel1-developed seaport, commercial and financial expertise, a highly monetized economy, and no backward agricultural sector. Its per capita income was only second only to Japan in Asia. Furthermore, as with other Asian newly industrializing economies at the time, Singapore also benefited from the growth of global production networks and supply chains that spearheaded its industrialization. Above all, Singapore was lucky to have a visionary and non-corrupt political leadership and a pragmatic and effective economic management team.

Going forward, the overall economic challenge for Singapore is how to successfully transition to an innovation-driven economy and meet the rising aspirations for better quality of life and social inclusiveness. While there are many constraints and challenges, Singapore also has a foundation of high-quality institutions and is educationally and financially in much better shape than five decades ago. Following the September 2015 General Elections, the government is establishing a Futures Committee to explore the pathways for the next phase of Singapore’s economic development.

Section 2 of this paper analyzes the policy and performance record of the past five decades. Section 3 examines the domestic and external constraints, challenges, and opportunities in going forward. Section 4 concludes.

2. Performance record and policies pursued

2.1. Transition from Third World to First World economy

This transition from 1965 took five decades. In the first two decades, the political leadership undertook the daunting task of nation building, economic restructuring, and forging social cohesion. It adopted a development model characterized by being government-led, foreign mnc-led and export-led.

The government built a world-class, non-corrupt public service with recruitment and promotion based on meritocracy; reformed the labor movement resulting in industrial peace and a flexible labor market; and built low-cost public housing to remove urban slums and provided home ownership for the masses. Singapore faced threats to its traditional economic pillars of entrepôt trade and servicing the British military base, and the government played a proactive role in jump-starting the economy through industrialization. The very small domestic market and policy lessons from failed import substitution experiments elsewhere led to the choice of an export-led industrial strategy. The lack of manufacturing expertise and the daunting task of transforming trading entrepreneurs into industrial entrepreneurs capable of international marketing led to reliance on foreign mncs.

This would enable Singapore to access foreign investment capital, technology and managerial expertise, and integrate into global production networks and supply

40 Chia: Singapore’s five decades of development: lessons and future directions

chains. Where mncs seemed hesitant to enter, government-linked-corporations were formed to fill the void.

As a labor shortage emerged by the late 1970s, Singapore’s industrial strategy evolved towards higher-skill and higher-value added manufacturing and services. In more recent decades, the government has steered the economy through the 1991 Strategic Economic Plan, the 2003 Economic Review Committee Report, and the 2010 Economic Strategies Committee Report. In particular, the 2010 report developed strategies for Singapore to achieve sustained and inclusive growth to 2020. • Economic growth targets: Annual gdp growth of 3-5 percent and annual

productivity growth of 2-3 percent. Economic growth to be driven by productivity in the face of growing labor shortage and rising real wages.

• Human skills: Singaporeans with the skills, expertise, and flair to take on higher quality and higher- level jobs across the whole range of occupations.

• Singapore companies: Deeper base of globally competitive Singapore companies with leadership in market niches across Asia; a key base for global companies to grow and manage their pan-Asian operations and for Asian enterprises that are expanding internationally.

• Innovation: Vibrant climate of innovation, seeking commercial success through design, new products, and services, and tapping knowledge from a broader base of public and private sector r&d.

• Leading global city with its unique characteristics.

In the May 2011 General Elections, the ruling People’s Action Party lost a significant percentage of the popular vote, due to public dissatisfaction over the rising cost of living and housing prices; stagnating wages for the low-income workers and widening income gap; and crowding out by foreign labor of public housing, public transportation, and public healthcare facilities. In the interim years to the August 2015 General Elections, the government attempted to address many of these grievances through pumping up public housing, public transportation, and public healthcare facilities, increasing subsidies and transfers to the low income and the elderly, improving the wages of low-income workers, and slowing down the inflow of foreign workers. In the recent August 2015 General Elections, the People’s Action Party government was returned to office with a much higher percentage of the popular vote, reflecting satisfaction with the ability of the government to deliver on its promises.

2.1.1 Growth and global competitiveness

High gdp growth was achieved for a sustained period, with full employment and low inflation delivering rising incomes and living standards. Economic growth averaged 7.5 percent annually in the period 1965-2014, to reach a per capita gdp of S$71,318 (us$56,284) by 2014. Annual gdp growth moderated from 7.6


percent in the decade 1990-1999 to 4.9 percent in the decade 2000-2009 as the Singapore economy matured and moved towards its technological frontier and as the economy was buffeted by many cyclical shocks. For 2011-2014, annual economic growth further slowed to 3.6 percent. Average labor productivity growth slowed from 3.4 percent in the 1990s to less than 1 percent in recent years.

The 2015 Global Competitiveness Report ranked Singapore second behind Switzerland in overall competitiveness among 144 countries. Singapore ranks high on basic requirements, institutions, infrastructure, macroeconomic environment, health and primary education, efficiency enhancers, higher education and training, goods market efficiency, labor market efficiency, and financial market development. Singapore scored less well on the technological front in terms of technological readiness and innovation and sophistication.

2.1.2. Structural change

Singapore’s current economic structure is shown in Table 1.

TABLE 1. Singapore’s changing economic structure

2000 2014 2000 2014

S$ million Percentage distribution

Total GDP, current market prices 162,584 390,089 100.0 100.0

Goods producing sectors 52,434 92,055 32.3 23.6

Manufacturing 40,699 67,817 25.0 17.4

Construction 8,863 18,961 5.5 4.9

Utilities 2,719 5,148 1.7 1.3

Services producing sectors 93,154 259,448 57.3 66.5

Wholesale and retail trade 20,405 64,440 12.6 16.5

Transport and storage 15,464 25,359 9.5 6.5

Accommodation and food services 3,521 8,161 2.2 2.1

Information and communications 5,697 14,915 3.5 3.8

Financial services 15,748 46,026 9.7 11.8

Business services 16,725 58,168 10.3 14.9

Ownership of dwellings 5,759 17,018 3.5 4.4

Taxes on products 11,238 21,568 6.9 5.5

Source: Yearbook of Statistics Singapore, 2011 and 2015

In 1965, the leading industries were domestic market oriented (printing and publishing, and food and beverage manufacturing). Industrial development was export-led and foreign mnc-led. By the 1980s, the leading industries comprised petroleum refining, marine transport equipment, and electronic products and components; Singapore was well integrated into regional production networks and global supply chains. By the 2000s, the leading industries were electronic and optical products, pharmaceutical and biological products, marine transport equipment, and machinery and equipment.


Singapore also grew as a regional services hub in finance and sea and air transport, linking Southeast Asia to the rest of the world. Its competitive advantages are a strategic geographical location; well-developed physical and telecommunications infrastructure; expertise in commerce, finance and infrastructure management; well-educated and English-speaking workforce; conducive legal environment; minimal restrictions on right of establishment; favorable tax regime; absence of controls on capital flows and foreign exchange transactions; and political, social and economic stability. It serves as the regional headquarters of many American, European, and Japanese mncs. Emerging service activities include healthcare, education, and creative industries.

2.1.3. Social well-being

Singapore ranked 9th in the undp’s 2014 Human Development Index with improvements in recent decades in life expectancy at birth, expected years of schooling, mean years of schooling, and Gross National Income per capita. The Boston Consulting Group’s 2015 study ranks Singapore 10th in overall well-being, the only non-European nation to make the top 10.

Singapore was found to have made particularly strong progress in education and governance, but it lagged in economic stability and income equality.

Singapore’s Gini coefficient rose to 0.478 by 2012, before it fell to 0.464 in 2014. After adjusting for government transfers and taxes, it moderated from 0.432 in 2012 to 0.412 in 2014, still high by Organisation for Economic Co-operation and Development standards. The ratio of household income from work per household member at the top and bottom percentiles rose to 9.64 in 2008 before declining to 8.92 in 2014 before government transfers and taxes and to 6.02 after. Also in 2014, the median monthly resident household income from work per household member was S$2,380, with 8.2 percent of resident households earning under S$2,000.

Various government measures introduced in recent budgets include increasing social transfers, especially to the working poor through the Workfare Income Supplement scheme and the Workfare Training Support scheme. The Workfare Income Supplement scheme provides support for low-wage workers and encourages them to work by supplementing their work income and Central Provident Fund (cpf) savings and encourages them to enhance their skills and employability. The Workfare Training Support scheme encourages employers to send their older low-wage workers for training.

However, with rising costs of living, low-income households are pressured to make ends meet. And with expectations of continuing volatile growth and retrenchments in times of economic recession and corporate restructuring, the absence of unemployment assistance means unmitigated hardships for the unemployed. Measures to help the elderly, especially those without adequate social safety nets, include the Pioneer General healthcare subsidies and the Silver


Support scheme. Measures have also been introduced to equalize education and training opportunities to increase inter-generational social mobility.

2.2. Effective role of the state in development

Government intervention following political independence in 1965 was extensive and went beyond the provision of public and merit goods as the public sector also played a leading role as investor and catalyst for Singapore’s development.

International reports such as the World Economic Forum’s Global Competitiveness Report, Transparency International’s Corruption Perceptions Index, and the World Bank’s Ease of Doing Business Index have ranked Singapore highly in quality of economic management and lack of government corruption.

The single-tier city-state government, a dominant-party legislature, and close integration of government ministries and agencies make for cohesive top-down decision-making and swift implementation of policies and measures. Public sector manpower recruitment is based on meritocracy, and public sector corruption is well controlled. Statutory boards were created to provide added administrative and financial flexibility, and many were given significant autonomy to carry out their tasks. Government-linked-corporations were established to lead the development of strategic sectors such as banking, telecoms, air and maritime transport development, industrial estates, and industry clusters. Pragmatism is manifested in government willingness to continually calibrate the balance between state intervention and market discipline. The leadership and policymakers continuously interact with international organizations, foreign governments and leading business leaders and thinkers around the world to establish friendly ties and identify emerging technological, economic and business trends globally and regionally so as to plan appropriate and timely policy responses.

Prudent fiscal and monetary policies were pursued. A low tax regime encouraged investments, savings, and the work ethics, while tax evasion is severely punished and budgetary expenditures were kept in check by eschewing the welfare state. There is zero external public debt and a large accumulation of forex reserves.

Innovative institutions were encouraged to tackle unique problems that Singapore faced early on. For example, the institutions to promote industry, investment, and export include the following:• The Economic Development Board was established in 1961 to spearhead

Singapore’s industrialization and foreign direct investment drive. Its International Advisory Council includes global heads of mncs to advise on international and regional strategies. Its network of overseas offices in North America, Western Europe and Asia target sectors, activities and firms for investment promotion. It functions as a one-stop investment center, assisting


investors to reach commercial production in Singapore in a matter of weeks. It focuses not only on pre-investment promotion but also on post-investment services, keeping existing investors satisfied so that they are encouraged to stay, reinvest and expand. Since 1993 the Economic Development Board also led in promoting outward investment drive by Singaporean firms.

• spring Singapore promotes Singaporean enterprises through assistance in financing, capability and management development, technology and innovation, and accessing new markets. As the national standards and accreditation body, it develops and promotes internationally recognized standards and quality assurance infrastructure that build trust in Singapore enterprises, products, and services.

• International Enterprise Singapore promotes the overseas growth of Singapore-based enterprises and international trade. With a global network, it offers services to help enterprises export, develop business capabilities, find overseas partners, and enter new markets. At the same time, it works to position Singapore as a base for foreign businesses to expand into the region in partnership with Singapore-based companies.

The National Wages Council fosters tripartism, industrial peace, and wage flexibility. It was formed in 1972 as a tripartite body (representing employers, trade unions and government) when a tight labor market led to concerns over possible rise in industrial disputes and excessive wage demands. It forges national consensus on wages and wage-related matters and its annual recommendations served as guidelines for negotiation between employers and unions. Following the 1985-1986 recession, the National Wages Council shifted from issuing quantitative to qualitative guidelines to enable more flexibility in wage negotiations and introduced a flexible wage system with variable components linked to company performance. Following on the Asian financial crisis of 1997-1998, further wage flexibility was introduced in the form of a monthly variable component to enable companies to respond more readily to a volatile business environment.

cpf provides home ownership and a social safety net. It is a compulsory self-funded scheme inherited from Singapore’s colonial past, initially intended to meet Singaporeans’ retirement needs but subsequently expanded to include housing and healthcare needs. The employer and employee contributions to the cpf have been tinkered with from time to time to reflect cyclical upturns and downturns in business costs and profitability. With expanded use of cpf funds, the amount left for old age retirement has been seriously eroded, raising concerns over its adequacy to meet the retirement needs of a rapidly ageing population.

The Housing and Development Board provides public housing for the masses. It was established in 1960. By 2014, 82 percent of Singapore’s population was


living in Housing and Development Board-built flats, with 80 percent home ownership. Public housing has transformed the physical and social landscape of Singapore. It has enabled slum removal and efficient town planning in land-scarce Singapore as well as social and ethnic integration. From time to time, the affordability of public housing became a hot political issue. While rising property values benefit the existing homeowners, there was growing concern among young couples whose home ownership aspirations were checked by rising prices and a long waiting period. In recent years, a slate of new policies has been introduced to improve affordability and availability, particularly for young couples and low-income families.

2.3. Outward- and forward-looking sectoral policies

2.3.1. Industrial development

The industrialization strategy is export-led and fdi-led. Export orientation was necessitated by Singapore’s small domestic market. A strategy of trade openness enables export production according to comparative advantage with exploitation of economies of scale and agglomeration, while zero import tariffs on capital and intermediate goods help keep production costs competitive and those on consumer goods benefit consumers through lower retail prices. Singapore is a key node in the region’s production and distribution networks. High trade-dependence, however, also makes Singapore vulnerable to changes in the external environment and necessitates nimble economic management to moderate the effects of transmitted business cycles.

Singapore’s export manufacturing necessitated dependence on fdi. The city-state lacked industrial or crafts traditions, and transforming trading entrepreneurs into industrial entrepreneurs with technological and international marketing capabilities was time consuming. fdi enabled Singapore to have access to foreign technology as well as industrial capital, management, marketing and integration into global production networks and supply chains. Attracting fdi became a policy priority. As most developing countries were then strongly influenced by the dependency school and eschewed the role of mnc, the first-mover advantage enabled Singapore to successfully leapfrog into export manufacturing in the late 1960s.

With the emergence of labor shortages by the late 1970s, Singapore’s industrial strategy shifted towards higher skill and value added manufacturing. The Manufacturing 2000 program targeted manufacturing at not less than 25 percent of gdp and called for the development of industry clusters, aimed at upgrading capabilities across the entire value chain. The Industry 21 initiative identified electronics, chemicals, engineering, life sciences, education and healthcare, headquarters, communications and media, and logistics as industry clusters to be nurtured. This transformation of industry-mix by 2010 is seen in Table 2.


TABLE 2. Singapore’s industry clusters, 2010

Percentage distribution

Out

put

, S$

mill

ion

Valu

e ad

ded

, S

$ m

illio

n

Em

plo

ymen

t nu

mb

er

Valu

e ad

ded

p

er w

ork

er,

S$0

00

Out

put

Valu

e ad

ded

Em

plo

ymen

t

Total manufacturing 270,494.7 56,863.8 419,963 135 100.0 100.0 100.0Electronics 94,193.6 17,865.0 80,466 222 34.8 31.4 19.2 Semiconductors 57,029.4 11,904.3 40,662 293 21.1 20.9 9.7 Computer peripherals 8,966.9 2,235.6 10,659 210 3.3 3.9 2.5 Data storage 14,487.2 1,977.5 13,360 148 5.4 3.5 3.2 Information communications and consumer electronics

11,375.6 1,111.7 8,149 136 4.2 2.0 1.9

Other electronic modules and components

2,334.6 635.9 7,636 83 0.9 1.1 1.8

Chemicals 80,506.5 6,081.4 23,826 255 29.8 10.7 5.7 Petroleum 41,627.9 1,234.1 3,535 349 15.4 2.2 0.8 Petrochemicals 29,991.0 1,972.0 4,735 416 11.1 3.5 1.1 Specialty chemicals 6,794.5 2,248.4 9,846 228 2.5 4.0 2.3 Others 2,093.0 627.0 5,710 110 0.8 1.1 1.4Biomedical manufacturing 23,253.5 11,126.7 13,738 810 8.6 19.6 3.3 Pharmaceuticals 19,668.0 9,700.4 5,369 1807 7.3 17.1 1.3 Medical technology 3,585.5 1,426.3 8,369 170 1.3 2.5 2.0Precision engineering 26,546.2 7,620.3 92,323 83 9.8 13.4 22.0 Machinery and systems 14,073.9 4,186.1 38,814 108 5.2 7.4 9.2 Precision modules and components 12,472.3 3,434.1 53,509 64 4.6 6.0 12.7Transport engineering 25,058.7 8,339.4 115,478 72 9.3 14.7 27.5 Marine and offshore engineering 16,030.6 4,875.7 91,140 54 5.9 8.6 21.7 Aerospace 7,207.7 2,826.5 18,213 155 2.7 5.0 4.3 Land 1,820.4 637.3 6,125 104.1 0.7 1.1 1.5General manufacturing industries 20,936.2 5,831.0 94,132 62 7.7 10.3 22.4 Printing 2,656.7 1,340.3 17,406 77 1.0 2.4 4.1 Food, beverages, tobacco 7,283.3 1,868.1 26,500 71 2.7 3.3 6.3 Miscellaneous industries 10,996.2 2,622.7 50,226 52 4.1 4.6 12.0

Source: Yearbook of Statistics Singapore, 2011

Industrial estates and business parks ensure the efficient use of scarce land resources, enable businesses to enjoy the economies of scale and agglomeration from locational clustering, and allow speedy start-up of commercial operations with the ready availability of land and industrial amenities. The Jurong Industrial Estate was established in the late 1960s and is the forerunner of many smaller and specialized industrial estates, business parks, and science parks. The latest developments are the Jurong Island petrochemical cluster and the Biopolis biotech cluster.


2.3.2. Services hub development

The role of services in the Singapore economy rose to reach 66.5 percent of gdp by 2014, with leading services in trade, business services and financial services.

Blueprints to develop the Singapore services sector were first outlined in the 1991 Strategic Economic Plan, which emphasized trade, transport, logistics, telecoms, and financial services. This was followed by the 2003 Economic Review Committee Report, which has the following recommendations: • Position Singapore as world-class education and healthcare hub, with

emphasis on attracting a strong cluster of education centers with world-class universities, executive learning centers, corporate training centers, and distance learning providers.

• Build the information communications and media cluster into a global hub in Asia for the digital economy offering a wide range of initiatives and developments in ict, media, e-commerce, and the Internet. Parallel with efforts to strengthen the telecommunications infrastructure is the active promotion of software development, Internet builders, application service providers, portals, and intermediaries.

• Build on Singapore’s reputation as the premier location for mncs to attract them to base regional and business headquarters.

• Nurture a logistics and supply-chain management cluster to develop Singapore into a leading Asia- Pacific integrated logistics hub and to build up supply chain capabilities by getting logistics players, cargo airlines, value-added distributors, and manufacturers to locate their supply chain centers for Asia in Singapore. These world-class logistics services enhance manufacturers’ global supply chains. A free trade zone logistics park has been established at Singapore Changi Airport, and a chemical logistics hub set up on Jurong Island.

The 2010 Economic Strategies Committee Report further recommended that Singapore carry out the following:• Grow manufacturing-related services such as headquarter-related activities,

r&d, Intellectual Property (ip) management, and product lifecycle management by capitalizing on the convergence of manufacturing with services.

• Strengthen Singapore’s position as a leading global-Asia financial and business hub that connects the global and Asian business community.

• Establish Singapore as the leading business hub in Asia with strengths in risk management and trading, asset management and private banking, loan syndication, project and infrastructural financing, and capital raising as well as information communications technology, accounting, legal, and consulting services.


• Seek to seamlessly integrate physical trade with related services such as trade finance, risk management, supply chain management, certification, and distribution.

• In maritime and aviation areas, enhance activities in insurance, financing, legal, and arbitration services.

• Develop Singapore as a leading consumer business center with clustered companies in marketing, branding, consumer research, and market intelligence, capitalizing on Singapore’s Asian cultural affinity, and develop Singapore into the pan-Asian location of choice.

• Establish Singapore as the location for future-ready urban solutions, by leveraging on Singapore’s own future urban needs and track record in urban planning, focusing on areas such as urban mobility/smart transportation, energy efficiency and management, renewable energy, and water and waste management. Such initiatives can also foster collaboration among Singapore-based companies to provide solutions that could be scaled up and exported.

2.4. Global and regional trade and investment integration

Singapore has been able to leverage on its strategic geographic location with infrastructure development, financial, and commercial expertise, and a free trade policy to become trading, transportation, and financial hubs. Its growth performance to date is due in no small part to the pursuit of economic openness. Table 3 shows Singapore’s economic openness through trade in goods and services and inward and outward direct investment.

TABLE 3. Singapore’s economic openness

2000 2010 2014Total merchandise trade (S$ million) 470,001 902,063 982,701.9Total merchandise exports (S$ million) 237,826 478,841 518,922.7 Domestic exports (S$ million) 135,938 248,610 273,492.1 Re-exports (S$ million) 101,888 230,231 245,430.6 Domestic/total exports (percentage share) 57 52 53Total merchandise imports (S$ million) 188,142 423,222 463,779.1Merchandise trade/GDP ratio (%) 292 297 252World Trade Organization Most Favored Nation average tariff (%) 0 0Total services trade (S$ million) 101,125 284,252 357,298.1Exports of services (S$ million) 49,213 152,929 177,935Imports of services (S$ million) 51,912 131,323 179,362.5Services trade/GDP ratio (%)\ 63 97 92 1999 2010 2014Inward FDI stock (S$ million) 276,819 625,780.4 853,339.5 Inward FDI/GDP ratio (%) 194 219Outward FDI stock (S$ million) 92,720 425,207.7 531,691.1 Outward FDI/GDP ratio (%) 132 136 Outward/Inward FDI ratio (%) 33 68 62



2.4.1. The pursuit of free trade agreements (ftas)

Singapore’s trade in goods and services amounted to over 300 percent of gdp, and the import contents of consumption, production, and exports are extremely high. It is a key node in the global and regional production networks in the electronics and machinery industries. Top export destinations are asean (particularly Malaysia and Indonesia), eu, us, as well as Northeast Asia (China, Japan, Hong Kong, Taiwan, South Korea). The major import sources are asean, eu, us, China, and Japan.

Table 4 shows that domestically produced exports comprise largely of refined petroleum and petroleum products and technology-intensive products, such as chemical and biotech products and high-end electronics. Exports of services include financial, maritime, aviation, logistics, and tourism services, while education and health services are also emerging exports. Singapore also serves as an entrepôt of Southeast Asia, providing efficient and timely shipment services for a sizeable portion of the region’s trade in commodities and manufactures. Its advantages are its container port (second in Asia after Shanghai), frequency of shipping and airline routes, logistics, financial services, and other distribution services. As a result, shipment costs to destinations in the Americas and Europe are often cheaper via Singapore than via national ports in the region.

TABLE 4. Singapore’s domestic exports, 2014

Value (S$ million) Percentage share

Food, beverages, and tobacco 6,522 2.38

Crude materials 2,074 0.76

Petroleum and products 106,986 39.12

Animal and vegetable oils 217 0.08

Chemicals and chemical products 49,383 18.06

Machinery and equipment 73,567 26.90

Electronics 44,059 16.11

Integrated circuits and parts 24,784 9.06

Personal computers and parts 9,313 3.41

Diodes and transistors 3,412 1.25

Disk drives 1,910 0.70

Telecoms equipment 1,454 0.53

Non-electronic machinery, equipment 29,508 10.79

Professional, scientific, optical equipment 9,689 3.54

Others 25,054 9.16

Total domestic exports 273,492 100.00



Singapore has been prolific in signing and implementing ftas with countries and regions across the world. These ftas benefit both ftas and foreign companies in Singapore who are goods exporters, service providers, and investors, providing them with improved market access and investment protection.

There is little domestic resistance to ftas, as businesses are accustomed to import competition. Services liberalization leads to inflow of foreign service providers, which improves the efficiency, range, and quality of services available to Singapore’s goods producers and consumers. There are also indirect benefits as trade and investment expansion leads to higher economic growth, job creation, and spin-offs for domestic enterprises. National and preferential treatment and investment protection measures in ftas also facilitate more Singapore-based companies, especially small and medium-sized enterprises (smes), to venture abroad, particularly to asean countries.

2.4.2. Inward and outward direct investment integration

From the 1960s to the 1970s, inward fdi helped to close the domestic saving-investment gap and finance net imports of goods and services. By the mid-1980s, Singapore had become a net capital exporter, and inward fdi contributed to technological and managerial know-how and integration into regional production networks and global supply chains. Inward fdi contributed importantly to Singapore’s increasing export sophistication, particularly in electronics, petrochemical, chemical, and pharmaceutical and financial sectors. Wholly foreign and joint venture firms are much more export-oriented than domestic firms.1

Singapore’s technological development can be divided into four phases.2 First, from the early 1960s to mid-1970s, there were few innovation links between foreign mncs and the rest of the Singapore economy. Second, from the mid-1970s to late-1980s, there was rapid growth of local technological development within mncs and development of local supporting industries. Third, from the late-1980s to late-1990s, there was rapid expansion of applied r&d by foreign mncs, local firms, and public r&d institutes. Lastly, since the late-1990s, there has been emerging emphasis on high-tech startups and basic r&d development. However, Singapore’s ability to innovate and pioneer new technologies still lags behind the world frontier.

Singapore began the outward investment drive in 1993. On the push side, a maturing Singapore economy facing severe land and labor constraints and rising costs needed to develop “an external wing” to sustain its growth performance. On the pull side, East Asia has become the world’s most dynamic economic region and

1 This is in part due to the fact that domestic firms acted as local suppliers to Singapore-based mncs rather than exporting directly.2 See Wong [2003].


offers Singapore investors abundant natural resources, low-cost labor, and rapidly expanding markets. Additionally, Singaporean domestic enterprises with limited outward investment experience and managerial resources would find it easier to regionalize due to geographic proximity and cultural and linguistic familiarity, which help reduce information and transaction costs and need for management oversight. Furthermore, Singapore government agencies have developed close relations and extensive networks in the region, and its regional projects, information networks, and contacts could facilitate investments by the Singapore private sector. Singapore’s ftas also help facilitate overseas investments.

Table 5 shows inward and outward fdi stock. Inward fdi stock rose rapidly to reach S$853.3 billion by 2013 or 226 percent of gdp. Singapore played host to thousands of foreign mncs. Reinvested earnings and expansion investments accounted for a growing share of fdi inflows. Manufacturing’s sectoral share has been falling, and fdi is much larger in services. Major sources of fdi are the advanced industrial economies, with eu, Japan, and the us accounting for a major share. Over the past decade, the eu’s share has risen rapidly, particularly from the Netherlands and the uk, while the Japan and us shares have shrunk. Outward fdi stock grew to S$531.7 billion by 2013, or 141 percent of gdp, and investments were mostly in services, particularly financial and insurance services. The outward investment in manufacturing reflects the deindustrialization in Singapore due to land and labor constraints. The outward push has also responded to the strengthening Singapore dollar and improved investment climate in the asean region, China, and India. Investments in overseas financial activities reflect Singapore’s growing role as an international and regional financial center. In the post-1997 Asian financial crisis, cash-rich Singapore firms were busy acquiring regional business assets as the crisis countries relaxed their mergers and acquisitions restrictions. Singapore’s outward direct investment (odi) traditionally has a strong Asian bias, but the emphasis has declined from a 75.9 percent share in 1981 to 54.4 percent in 2013. China’s share grew rapidly from less than 3 percent in 1985 to 19.4 percent in 2013, mostly taking place after 1992, almost equal to asean’s share of 19.8 percent.

2.4.3. Spatial connectivity

To serve Singapore’s highly trade-dependent manufacturing and services activities, the city-state invested heavily in a wide range of infrastructure including transportation, logistics, and telecoms systems. To maintain its competitive edge, there is constant upgrading and expansion of infrastructure. Singapore’s world-class seaports and airports provide crucial links with the region and the world. Singapore also has a well-developed information and communications technology infrastructure that contributes to its international competitiveness as a knowledge-based economy.


TABLE 5. Singapore’s stock of inward and outward foreign direct investment, by industry

1999 2010 2013 1999 2010 2013

Inward FDI S$ million Percentage distribution

Total 170,820.8 625,780.4 853,339.5 100.00 100.00 100.00

Manufacturing 58,139.6 133,590.7 150,791.1 34.04 21.35 17.67

Construction 1,505.1 1,468.3 3,227.6 0.88 0.23 0.38

Wholesale and retail trade 26,008.7 108,721.5 146,273.6 15.23 17.37 17.14

Accommodation and food services

1,993.7 3,811.8 3,976.5 1.17 0.61 0.47

Transport and storage 5,958.6 36,793.8 36,858.2 3.49 5.88 4.32

Information and communications

873.7 6,418.4 8,595.1 0.51 1.03 1.01

Financial and insurance services

64,647.4 270,176.8 293,621.8 37.85 43.17 34.41

Real estate activities 5,620.7 20,083.1 29,560.5 3.29 3.21 3.46

Professional, technical, and administrative services

5,674.9 35,173.7 40,578.7 3.32 5.62 4.76

Others 398.4 9,542.4 14,166.5 0.23 1.52 1.66

Outward FDI

Total 92,719.9 425,207.7 531,691.1 100.00 100.00 100.00

Manufacturing 22,869.5 88,635.7 109,943.8 24.67 20.85 20.68

Construction 797.4 1,342.3 2,238.9 0.86 0.32 0.42

Wholesale and retail trade 5,921.3 26,571.0 45,725.0 6.39 6.25 8.60

Accommodation and food services

1,692.4 3,798.6 4,442.2 1.83 0.89 0.84

Transport and storage 3,408.7 10,363.6 13,730.8 3.68 2.44 2.58

Information and communications

2,257.8 17,958.4 21,935.5 2.44 4.22 4.13

Financial and insurance services

44,717.5 206,204.1 210,207.9 48.23 48.49 39.54

Real estate activities 6,869.4 35,390.9 44,215.7 7.41 8.32 8.32

Professional, technical, and administrative services

2,737.0 7,617.2 10,205.3 2.95 1.79 1.92

Others 1,448.8 27,326.0 34,480.4 1.56 6.43 6.49


2.5. Education and skills levels of the workforce

Although Singapore has four official languages (English, Mandarin, Malay, and Tamil), English is the language of administration and business and the medium of instruction in tertiary institutions.

In the early decades of industrialization, vocational and skills training in vocational institutes and polytechnics equipped the workforce with industrial skills. The government also leveraged on the training capabilities of mncs


and their home governments by setting up joint training institutes to provide industrial skills in demand. Singapore was able to minimize the mismatch between skills supplied and skills needed. As the economy advanced towards a knowledge-based-economy, the educational level and skills-set required changed correspondingly.

The educational attainment of the resident labor force has improved rapidly in the past decade, as the less educated retired and the new labor force entrants are much better educated. High-skilled professionals, managers, executives, and technicians have grown more rapidly (with 53.1 percent share in 2014). The proportion of university degree holders in the workforce correspondingly increased from less than 2.4 percent of the labor force in the 1970s to 31 percent by 2013.

Tertiary education has expanded rapidly since the mid-1980s. By 2015, of the age cohort that started formal education, 45 percent gets into polytechnics and 30 percent into universities, making a total of 75 percent with tertiary education. Table 6 shows percentage distribution of polytechnic and university enrolment by course. The lion’s share found in engineering sciences and business administration. In more recent decades, the government realigned its focus and revised the school curriculum to concentrate on developing students’ creativity and critical thinking.

Singapore has also been improving on its educational quality. The World Economic Forum competitiveness rankings for Singapore shows that its educational institutions are among the best in the world. The Organisation for Economic Co-operation and Development Program on International Student Assessment shows Singapore students in top rankings in mathematics, science, and reading. Quacquarelli Symonds ranks Singapore’s public universities among the top in East Asia.

2.6. Prudent policies and macro-economic stability

Table 7 shows Singapore’s strong financial resources as evidenced from the government’s strong budgetary position (reflecting prudent monetary and fiscal policies over the years), the household sector’s assets including the cpf savings of individuals, and the strong net international position with no government external debt and high official foreign reserves. In addition, Singapore is one of the few countries with an aaa sovereign credit rating by Standard and Poor.


TABLE 6. Enrolment in polytechnic and university courses

1998 2010 2013 1998 2010 2013University first degree courses Number Percentage distribution Education 630 1,841 1,296 1.90 3.22 2.15 Applied arts 1,076 1,262 - 1.88 2.09 Humanities and social sciences 5,354 9,694 10,642 16.18 16.97 17.66 Mass communications 474 675 702 1.43 1.18 1.16 Accountancy 2,351 3,203 4,042 7.10 5.61 6.71 Business and administration 3,834 6,426 6,476 11.59 11.25 10.75 Law 629 1,423 1,491 1.90 2.49 2.47 Natural, physical, and mathematical sciences 2,952 7,640 7,761 8.92 13.38 12.88 Medicine, dentistry, health sciences 1,203 2,553 3,081 3.64 4.47 5.11 Information technology 1,515 3,207 3,888 4.58 5.61 6.45 Architecture and building 1,177 1,786 1,884 3.56 3.13 3.13 Engineering sciences 12,975 17,294 17,426 39.21 30.28 28.92 Services 299 312 - 0.52 0.52 Total 33,092 57,117 60,263 100.00 100.00 100.00 Polytechnic diploma courses Education 174 274 - 0.78 1.09 Applied arts 1,378 1,192 1,805 2.59 5.37 7.19 Humanities and social sciences 275 538 - 1.24 2.14 Mass communications 595 481 606 1.12 2.17 2.41 Business and administration 10,731 4,696 5,695 20.18 21.14 22.69 Legal studies 306 125 159 0.58 0.56 0.63 Science and related technologies 1,324 1,070 1,408 2.49 4.82 5.61 Health sciences 2,093 1,999 2,504 3.94 9.00 9.98 Information technology 5,263 3,568 3,420 9.90 16.06 13.63 Architecture and building 1,859 539 752 3.50 2.43 3.00 Engineering sciences 29,139 7,743 7,453 54.79 34.86 29.70 Services 489 352 483 0.92 1.58 1.92 Total 53,180 22,214 25,097 100.00 100.00 100.00


TABLE 7. Singapore’s financial indicators

2010 2014Government sector Government operating revenue, S$ million 46,060 57,020 Government total expenditure, S$ million 45,338 51,728 Government overall budget surplus, S$ million 980 4,998 Government domestic public debt, S$ million 321,182 387,251 Household sector Household net worth 1,191,581 1,467,398 Central Provident Fund due to members 185,888 275,364 International position Net international investment position, S$ million 651,406 710,039 Government external public debt, S$ million - - Official foreign reserves, S$ million 288,954 340,438 Official foreign reserves, US$ million 225,754 256,860



3. Going forward: challenges and opportunities

While Singapore has generally been a successful economic development story over the past five decades, it also inherits a legacy of constraints and challenges going forward. There are two more immediate economic challenges: first, the economy is on a declining path of sustainable growth, and this is in the face of aspirations for higher living standards and social security and an uncertain and sluggish global and regional economic environment; and second, there is a need to push for growth driven by innovation and local enterprise. In the more distant future, Singapore has to worry about the impact of climate change with an accompanying rising sea level and how this will impact on a small island nation.

3.1. Slower growth rate and productivity growth

In January 2015, the Singapore prime minister announced that the targeted growth rate going forward over the next several years has to be reduced to 2-3 percent instead of the 3-5 percent target set by the 2010 Economic Strategies Committee.

The Singapore economy has matured and its growth trajectory will mirror that of other small advanced economies. Domestic constraints of land and labor and a rapidly ageing population have resulted in high costs and deteriorating competitiveness in the absence of rapid improvements in productivity and a vibrant private entrepreneurial sector. Furthermore, the external environment has become less benign and accommodating: regional economies are becoming more competitive and challenging Singapore’s hub status, and there is slowing demand for Singapore’s goods and services by the advanced economies and by China as they go through periods of slower growth themselves. However, Singapore has the advantage of a solid foundation of high-quality institutions, well-educated workforce, and ample financial resources to pull through.

Productivity growth performance has not met expectations and requires a reinforced push to transform business operations and to intensify skills training. Labor productivity growth fell from an annual 3.4 percent in the 1990s to 1.1 percent in 2000-2009 and 0.3 percent in 2009-2014. One culprit for the poor productivity performance was the large influx of low-skilled labor in the past decade, which disincentivize business upgrading, particularly in the construction and retail services sectors. In addition to putting a cap on foreign labor inflows in order to force businesses to upgrade, the government has also several initiatives including tax benefits, grants, and training subsidies to help companies and workers invest in productivity, innovate, and deepen skills and expertise. In particular, SkillsFuture is launched as a national movement to enable Singaporeans to learn new skills for new jobs as the economy continues to restructure and meet the challenges of globalization and disruptive technological change.


3.2. Widening income gap and inadequate social safety net

The Singapore government has long eschewed Western style social welfare and preferred to focus on providing education, housing, and healthcare, to equalize opportunities. But the income gap has widened in the past decade, necessitating remedial measures in recent years as it could undermine social cohesion.

While obviously more needs to be done to raise the living standards and well-being of low-income groups and the elderly poor, the government also needs to ensure that its rising social expenditures remain fiscally sustainable. It has traditionally exercised fiscal prudence so that there is overall no net public debt or external debt.

Some years ago, the goods and services tax was introduced to ensure adequate tax revenues without raising personal and corporate income taxes, which could disincentivize work and investment. However, the 2015 budget increased the top marginal income tax rate from 20 percent to 22 percent explicitly to provide funds for increased government social expenditures. More taxes on the rich and user-charge fees, as well as further measures to equalize opportunities to education, training, and remunerative employment, can be expected.

3.3. Demographic and labor constraints

Singapore’s natural population growth has slowed dramatically with the total fertility rate (tfr) declining below replacement rate for decades, which together with rising life expectancy, has resulted in a rapidly ageing population and elderly healthcare costs. Population of working age citizens (20-64 years) peaked in 2005, but the elderly aged 65 and above is growing rapidly. The ratio of working age to elderly has declined from 13.5 in 1970 to 4.8 currently and will plummet to 2.1 by 2030. Government policy efforts to reverse tfr had produced no sustained result and stood at 1.3 percent in 2014. The ageing and shrinking population means a shrinking labor force and a less vibrant and innovative society.

Population and labor force growths have been heavily dependent on immigration, with foreigners accounting for one-third of the labor force in recent years. Going forward, labor force growth can come from increasing the female and the elderly labor force participation rates and hiring foreign workers.

Female labor force participation has been rising (58.4 percent in 2014) and could be raised further, as it has not reached the levels of many western societies. However, higher female labor force participation could impact negatively on the tfr. It could also be argued that better working conditions for working mothers—such as flexible hours of employment, more family-friendly policies by employers, more reliable and affordable childcare facilities, ease of re-entry married women into the workforce—could increase both female labor force participation and tfr.


With rising life expectancy (82.8 years in 2014), Singaporeans are being encouraged to postpone retirement. In any event, with cpf balances inadequate to fund most retirement needs, working longer is a necessity for many. The government has gradually lifted the official retirement age from age 55 to 60; it is currently age 65. To incentivize the private sector to employ older workers, the employer’s cpf contribution rates for older workers have been reduced. The government is also encouraging movement away from a seniority-based wage system to a productivity-based wage system to encourage the employment and re-employment of older workers.

To meet the growing shortage of skills, particularly in the transition to an innovation-based economy, tertiary education and skills training have been ramped up with expansion of universities, polytechnics, and various skills training institutes and continuing changes in educational curriculum and pedagogy. To reduce the demand for labor, there should be faster and wider adoption of labor saving production technologies and practices.

An open policy towards foreign labor has enabled Singapore to overcome the labor constraint since the late 1970s. The large influx in recent years is unsustainable because of physical and social limits, with vocal concerns on foreigners crowding out locals for jobs, housing, education, health and transportation services, and recreational space. Too heavy a dependence on foreign workers has also contributed to the poor productivity performance of the Singapore economy in the past decade, particularly in the construction and some service sectors.

For an innovation-driven economy, Singapore needs a large and expanding pool of foreign human talent, even though local talent is growing with the rapid expansion and revamp of tertiary education and training institutes. Foreign talent is being recruited through liberalized immigration policies, easing requirements for permanent residence and citizenship, offer of scholarships and research fellowships at Singapore tertiary and research institutions, recruitment missions by universities and government agencies to the main centers of learning abroad, and improving the living and cultural attractions and tax regime of Singapore for foreign expatriates.

While the inflow of foreign talent surged, there were even bigger inflows of low-skilled workers into labor-intensive manufacturing, services, and construction sectors. Additionally, the Singaporean workforce is becoming increasingly better educated and eschewing unattractive and low-paying jobs. Government policy aims at moderating employer demand for foreign workers through the use of work permits, foreign worker levies, and dependency ceilings, and there have been increases in these levies from 2010 to make foreign workers more expensive for employers. The government also announced that foreigners will continue to make up the current one-third of Singapore’s workforce in the next 10 years, with some flexibility to take account of booms and recessions. The large presence of foreign labor, both professionals and low-skilled employees, was a hot political issue in the May 2011 General Elections. Since then, the government has been further


challenged to fine-tune the balance between the needs of businesses for foreign professionals and workers, particularly in an economic upturn, and the spatial and social limits of allowing a continuing influx. Cutting off ready access to foreign workers will exert pressure on businesses to adopt labor-saving technologies, operations, and practices and will speed up economic restructuring.

3.4. Land and natural resource constraints

Land reclamation is costly and reaching a limit so Singapore would need to make more efficient and innovative use of its land and space. Additionally, there is the longer-term threat of climate change and rising sea level, impacting on Singapore’s coastlands and resulting in alternative sea routes bypassing the Straits of Malacca.

Going forward, Singapore needs to do the following:• Enhance land and space productivity. To gain the greatest benefits from its

limited land, Singapore has to create more space upwards and underground and be more creative in urban planning and design to ensure that Singapore remains a livable and business-friendly city. Sectoral policy towards production of goods and services will have to increasingly focus on value added to economize on land and space demands, such as high-value added and clean manufacturing and high-end tourism, healthcare, and education markets. This will also ensure Singapore’s continuing competitiveness vis-à-vis other less advanced and lower-cost asean countries.

• Enhance water planning and development. Many parts of Asia will face shortage of water as populations, cities, and agricultural and industrial needs grow. Singapore has historically imported most of its water supplies from neighboring Malaysia. However, over the years, Singapore has developed alternative water resources, including collection of rain water, recycling of waste water, and desalination of sea water. The city-state will become self-sufficient in water by the time the second water agreement with Malaysia expires in 2061. In fact, water technology and management has become a new Singapore growth industry and export of services.

• Enhance the region and the world as Singapore’s hinterland: Fortunately, Singapore is neither “landlocked” nor “sea-locked”. Instead, it is favorably located in the dynamic Asian region. Continuing improvement of physical (land, sea, and air) and information technology connectivity will minimize the cost for cross-border movement of goods and people and facilitate access to ideas and technologies. Support for global and regional free trade will help overcome the dearth of land and natural resources and achieve economies of scale and scope to overcome its small market size. Singapore must continue to support trade liberalization in the World Trade Organization and regional and bilateral ftas.


• Meet the rising costs of energy as well as reduce its carbon emissions. As an energy-deficient economy, Singapore has resorted to using the price signal in energy pricing, adopting clean technologies and energy-efficient usages in factories, offices, and homes, and designing of energy-efficient buildings. At the business level, Singapore should promote industries and services and processes that are less energy-intensive and/or use less of fossil energy. At the household level, consumers should be incentivized to save on energy use, such as installing energy-efficient household equipment. At the transportation level, Singapore pioneered efforts to restrict private car ownership and use and heavily taxed fossil-fuel consumption. Exploring energy alternatives include the nuclear option, which will be especially challenging for a densely populated city-state.

3.5. Weak local enterprises

High-income Singapore faces the challenge of weak local enterprises and entrepreneurship. Unlike South Korea and Taiwan, Singapore in the earlier phases of development and industrialization failed to nurture local private enterprise and chose to depend on foreign mncs and government-linked-corporations. By the mid-1980s, local enterprises were confronted with higher cost structures exemplified by rising land and space costs, labor shortages, and rising wage costs. Also, major public sector infrastructure projects were tendered out to large European, Japanese, Korean, and prc firms, creating a vicious cycle of local firms lacking the experience and track record to bid for them domestically or regionally.

The 1988 sme Master Plan marked the first coordinated national effort to upgrade local smes and to promote domestic entrepreneurship. Since then, a plethora of sme assistance schemes have been hatched and implemented.3 Furthermore, the sme 21 Report outlined the need to nurture innovative high growth world-class smes able to compete in the global marketplace; enhance sme productivity; and create a knowledge-based pro-enterprise environment. A multi-agency sme 21 Implementation Committee was formed. The Technopreneurship 21 initiative was announced in April 1999 specifically to boost development of technopreneurs. Another Master Plan, named sme21, was created to take Singapore smes into the 21st century. As a result of the two sme

3 These include the following: the Local Industry Upgrading Programme which helps local enterprises achieve greater efficiency through the transfer of management skills and technological know-how from mncs and large local companies; the Economic Development Board-Joint Venture Matching Service to help smes seek strategic alliance with foreign organizations for growth; the Business Development Scheme to encourage smes to seek business opportunities overseas for marketing arrangements, technological tie-ups, and other business partnerships; the Small Industry Technical Assistance Scheme to help smes improve their productivity and technological standards through grants to defray part of the costs of approved upgrading projects; and the Promising Local Enterprise Programme that aims to build 100 local companies with at least S$100 million sales turnover in ten years.


Master Plans, Singapore has a spectrum of some 63 assistance schemes to assist local enterprises.

Government agencies involved in smes include the following: International Enterprise Singapore, a statutory board responsible for taking smes overseas; spring Singapore, a first stop for all smes, which are then directed to the relevant agencies; A*Star that fosters scientific research and the exploitation of technology through incubator units; Jurong Town Corporation that provides industrial space and has incubator space as well; Information Development Authority which develops, promotes, and regulates the information technology and telecoms market as well as assists the adoption of online and e-commerce technology by smes. In addition to these government agencies, there are private sector institutions, tertiary institutions, and non-government organizations, which smes can access.

Notwithstanding these multi-pronged efforts, the World Economic Forum 2011-2012 Competitiveness Report continues to show poor rankings for Singapore in the following: local supplier quantity; local supplier quality; and control of international distribution. Due to the dominance of foreign mncs and dominance of the services sector in the economy, there are few internationally well-known Singaporean manufacturing brand names. unctad’s list of top firms from developing countries has few Singaporean names and they are mainly services firms or government-linked-corporations such as Singapore Airlines, Singapore Telecommunications, Keppel Corporation, Sembawang Corporation, Singapore Changi Airport, and the Maritime and Port Authority of Singapore.

The actual and effective assistance provided to smes paled in comparison to that offered to mncs and government-linked-corporations. Furthermore, local smes lack a sizeable domestic market to test their ideas, processes, and products and to grow before expanding abroad.

Many schemes to promote and facilitate local entrepreneurship and sme development and availability of venture capital have been introduced in recent years. There is also a noticeable trend of young university and polytechnic graduates entering directly into business ventures instead of taking up jobs with the public sector and foreign mncs. So there is hope for Singapore with more homegrown entrepreneurs.

3.6. Shifting to an innovation-driven economy

According to the World Economic Forum, Switzerland, Singapore, and Sweden are the top three in the Global Competitiveness Index rankings in 2011-2012.• Switzerland tops the Global Competitiveness Index rankings; its most

notable strengths are in innovation, technological readiness, and labor market efficiency. Switzerland’s scientific research institutions are among the world’s best, and the strong academia-business collaboration together with high corporate r&d expenditures and strong ip protection ensure the


commercialization of research into marketable products and processes. Its rate of patenting ranks 7th worldwide. Productivity is enhanced by a business sector and a population that are proactive in adapting the latest technologies. Competitiveness is also buttressed by excellent infrastructure, well-functioning goods markets, and highly developed financial markets.

• Sweden, like Switzerland, has been placing significant emphasis on innovation-led growth. The quality of its public and private institutions are tops in efficiency, transparency, and ethics. Goods and financial markets are very efficient. Combined with a strong focus on education and training and high level of technological absorption, Sweden has developed a sophisticated business culture and is a global leading innovator.

• Singapore is 2nd in Global Competitiveness Index rankings (after Switzerland) and top among Asian economies. While Singapore is tops in rankings on institutions, efficiency of its goods and labor markets, financial market development, world-class infrastructure, and education, it lags in innovation and business sophistication and have considerable catching up to do. Singapore needs to encourage stronger adoption of the latest technologies and measures that support the sophistication of its companies.

Singapore has to transition rapidly from being an investment-driven to an innovation-driven economy. At the investment-driven stage, the key drivers are a strong business environment, openness to trade and investment, well-developed physical infrastructure, legal and regulatory framework, and a solid skill base. But at the innovation-driven stage, the key drivers are a strong innovation system, ip protection, innovative and techno-savvy entrepreneurs, and sophisticated domestic producers and consumers.

Before the 1990s, r&d activities in Singapore remained low for a number of reasons: industrial growth in the 1960s and 1970s depended more on cost efficiency than on innovative capability; the mnc subsidiaries in Singapore had ready access to the processes and technologies from their overseas parents, and they generally preferred to conduct r&d in their home base; and there were no performance requirements to conduct local r&d or partner with local institutions in r&d. Singapore also lacked a critical mass of scientists and researchers to provide a stimulating research environment for both mnc and local enterprise r&d activities.

Recognizing the uphill task, the Ministerial Committee on Research and Development in 2005 reviewed the r&d strategies and directions for Singapore. The committee recommended that national r&d efforts should be driven by 5 key strategic thrusts.4

4 This is excerpted from a speech by Dr. Tony Tan, outgoing Singapore deputy prime minister and incoming chairman of the National Research Foundation, on August 11, 2005.


• Provide more resources for r&d. Singapore’s gross expenditure on research and development was 2.15 percent in 2003, which lags significantly behind other leading innovation countries. Singapore must intensify r&d efforts and achieve a gross expenditure on research and development of at least 3 percent of gdp within the next 5 years.

• Need to focus on a small number of strategic areas to develop a critical mass of research capabilities in industries where it can be economically competitive. Existing key clusters are electronics, chemicals, marine engineering, and biomedical sciences. A vibrant research environment will help identify emerging growth areas, such as in environment and water technologies and interactive and digital media.

• Need for balance between investigator-led and mission-oriented research in selected strategic areas. Basic investigator-led research is broadly aligned with the long-term strategic interests of Singapore, and the Ministry of Education’s Academic Research Fund was raised from S$550 million for fy2001-2005 to S$1.05 billion for fy2006-2010. Mission-oriented research would be closely integrated with industry development and investment promotion strategies, and the budget for A*Star was raised from S$4 billion for fy2001-2005 to S$5.4 billion for fy2006-2010.

• Encourage more private sector r&d. Incentive packages would be reviewed to better attract more global r&d centers and activities to Singapore, supported by a high quality support framework, including a strong base of scientific and research manpower and sophisticated IP protection regulations. The aim is to have two-thirds of r&d, mainly development, performed by the private sector, and one-third of r&d, mainly research, performed by the public sector agencies.

• Strengthen the nexus between r&d and business. Universities and research institutes must improve on their ability to commercialize their research results and have closer collaboration with industry. There is also a need to promote technology innovation in local enterprises through stronger co-funding frameworks between public and private sectors. In particular, polytechnics with strong applied research and downstream capabilities and industry networks could be encouraged to link with industry associations to collaborate on r&d initiatives.

The government elevated r&d into a national priority by setting up the Research, Innovation, and Enterprise Council and the National Research Foundation. The council announced that it plans to spend S$16.1 billion over 2011-2015 on research, innovation, and enterprise, as compared to an allocation of S$13.55 billion in 2006-2010. Research and innovation will have to underpin the competitiveness of Singapore’s industries, catalyze new growth areas, and transform the economy. Increasingly, intellectual capital will be critical to


Singapore’s next phase of economic development. More will be done to facilitate collaboration between industry and public research institutions to foster greater commercialization of r&d. A larger portion of r&d funding will be awarded on a competitive basis to projects that will strengthen capability and contribute to economic and social outcomes. The National Research Foundation has been established with key responsibilities to implement the strategic thrusts and to fund longer-term research in strategic areas. The foundation is provided with funding of S$5 billion for fy2006-2010. This, together with the increased budget for A*Star and the Academic Research Fund, will more than double the total public sector r&d budget from just under S$5 billion for fy2001-2005 to almost S$12 billion for fy2006-2010. Singapore aims to increase gross expenditure on research and development to 3.5 percent of gdp by 2015 through greater private sector r&d activity. The foundation’s research priorities are in environmental and water technologies; biomedical sciences; and interactive and digital media.

Despite the quantum leap, Singapore’s r&d efforts remain modest in absolute terms, reflecting the small size of the economy, its strong services orientation, and weak core of large manufacturing corporations. A challenge for Singapore’s r&d ambitions is that it lacks the breadth and depth of talent to compete against the bigger and more advanced economies of the us, eu, Japan, and Russia as well as the emerging economic powers of China, India, and South Korea.

4. Conclusion

In the past five decades, Singapore’s non-corrupt and visionary political leadership has brought about political stability and social cohesion, and, together with pragmatic and efficient economic management, has enabled Singapore to achieve well beyond the economic potential of a small city-state that lacks natural resources. However, in the process, Singapore’s development depended heavily on foreign mncs and neglected the nurturing of domestic enterprises. Its heavy dependence on foreign labor delayed the expansion of tertiary education and disincentivized industrial upgrading as businesses had ready access to a large pool of low-wage labor from neighboring countries. Critics have also criticized Singapore as being a “nanny state” with its extensive and intrusive role in the economy and society.

Singapore is a now a high-income country and is on the innovation-driven stage of economic development. Going forward, Singapore has to overcome domestic constraints and challenges as well as operate in a more challenging regional and global environment. The government will have to hold back on its dominant role and let the private sector and civil society grow and become more creative and risk taking. Given the best educational opportunities in the world, and a liberalizing trade and investment environment, and with government playing a less intrusive but more supportive role, the next phase of Singapore’s


economic development should be more private-led and local-enterprise led. Whether the Singapore city-state will continue to excel over the next few decades remains to be seen.

Singapore Institute of International Affairs

References

Chia, N.C. [2015] “Adding a basic pillar to the Central Provident Fund system: an actuarial analysis”, Singapore Economic Review Special Issue: a fifty year retrospective on the Singapore economy 60(3): 1530037-1 - 1530037-26.

Chia, S.Y. [2015] “Globalisation and regionalisation: Singapore’s trade and FDI”, Singapore Economic Review Special Issue: A Fifty-Year Retrospective on the Singapore Economy 60(3): 1530034-1 - 1530034-23.

Chia, S.Y. [2012] “Singapore”, unpublished background monograph prepared for the ASEAN 2030 project coordinated and funded by the Asian Development Bank Institute.

Chia, S.Y. [2011] “Inward and outward fdi and the restructuring of the Singapore economy”, in S. Chalongphob, Y.C. Park and S.J. Kang, eds., Foreign Direct Investments in Asia. uk: Routledge.

Economic Strategies Committee (esc) [2010] Report of the Economic Strategies Committee: highly skilled people, innovation economy, distinctive global city.

Phang, S.Y. [2015] “Singapore’s housing policies: responding to the challenges of economic transition”, Singapore Economic Review Special Issue: a fifty year retrospective on the Singapore economy 60(3): 1530036-1-1530034-25.

Singapore Department of Statistics [various years] Yearbook of statistics Singapore.

Singapore Department of Statistics [2014] Key household income trends 2014.Singapore Ministry of Trade and Industry [2003] Economic review committee

report.Spring Singapore. Assistance programmes for smes. Downloaded from

website: http://www.spring.gov.sg/Transparency International (ti) Global corruption report. Downloaded from

website: https://www.transparency.org/research/gcr/World Bank (wb) [2015] Doing business 2015 report. Downloaded from

website: http://www.worldbank.org/World Economic Forum (wef) [2010] The global competitiveness report 2010-

2011. Downloaded from website: http://www.weforum.org/reports/global-competitiveness-report-2010-2011-0


PRE

The Philippines: on the road to being an emerging economy

Dante B. Canlas

In the 1980s, the Philippines was viewed as a failure in terms of its goal to industrialize. But in the past few years of the current century, the country’s economic prospects improved and the country is predicted to be joining the next group of “breakout nations”. The paper looks at the short-run macroeconomic policy reforms and long-term growth-oriented policies since 1986 that have contributed to improved growth performance and bright economic prospects. A responsible budget deficit-reduction program and investments in factors that support long-run growth, such as human capital, have been helpful. Moving forward, stabilization policy has gotten more challenging in an environment of mobile international capital, flexible exchange rates, and “quantitative easing”. In the long run, policies conducive to technological progress are essential.

JEL classification: E32, O11 Keywords: growth, convergence, stabilization, technological progress

1. Introduction

Robert E. Lucas, Jr. [1993] in his article entitled “Making a Miracle”, cites the Philippines as an example of a country that failed to transform itself into an economic miracle. Lucas uses the latter phrase to describe a process of productivity growth and industrial transformation associated with profound improvements in living standards of the country’s population. Lucas has in mind the transformation that at the time had been occurring in South Korea, Taiwan, Hong Kong, and Singapore in East and Southeast Asia—these economies had witnessed a growth rate in real per capita income of 5 to 7 percent each year, in about three decades from 1960. South Korea and Taiwan were transformed from predominantly agricultural economies. Today, they are often held up as models of successful industrialization. Hong Kong and Taiwan had no agricultural sector

66 Canlas: The Philippines: on the road to being an emerging economy

to begin with and relied primarily on export-led growth. I refer to a country showing signs of undergoing such a transformation as an emerging economy.

About two decades later, Ruchir Sharma [2012], in his book, Breakout nations: in search of the next economic miracles, includes the Philippines in his list of developing countries likely to become an economic miracle in the second decade of the 21st century. The main basis for inclusion in the list hinges on exhibiting “growth that surpasses expectations”. This brings to mind the so-called bric countries—namely, Brazil, Russia, India, and China—which some observers before Sharma had tagged as breakout nations or emerging economies.

What did the Philippines do to merit significantly improved economic prospects? Is the country’s real per capita income showing a tendency towards convergence, i.e., catching up with the established economic miracles? There are many possible answers. This paper examines the macroeconomic and growth-oriented policies that underpin the significant improvements in the growth prospects of the Philippines.

A quick look at the growth performance of the Philippine economy over the past three decades, juxtaposed with two other middle-income Southeast Asian economies—namely, Thailand and Indonesia—provides some insights into the positive assessment that the Philippine economy is currently enjoying. Table 1 shows data spanning three decades that were accessed from the World Bank’s World development indicators. This period witnessed three economic crises in the Philippines: the debt default of the country in 1983; the Asian financial crisis of 1997; and the global financial crisis of 2008. In the 1980s, the average real per capita income each year of the Philippines declined 0.8 percent, compared to growth rates of 6 percent in Thailand and 4.46 percent in Indonesia. The Philippines recovered in the 1990s with an average annual growth rate of 0.58 percent. The recovery gained strength in 2001-2010, growing 2.52 percent each year.

TABLE 1. Real per capita income and average annual growth rate

Country 1980 real per capita income, y, (in 2000 US$)

Growth rate of y, 1981-1990 (in %)



2010 real per capita y, (in 2000 US$)

Philippines 1,088 (0.80) 0.58 2.53 1,375

Thailand 772 6.0 3.64 3.26 2,774

Indonesia 356 4.46 3.15 3.52 1,034

Source: World Bank, World development indicators, accessed August 2011


In 1983, the Philippines experienced a balance-of-payments crisis that forced the government to declare a moratorium on foreign-debt servicing. To overcome its liquidity problems, the national government arranged a standby loan agreement with the International Monetary Fund (imf). The latter’s preferred financial programming techniques and conditionality lending practices anchored on tight fiscal and monetary policies resulted in a deep recession in 1984-1985. Real gdp growth declined 11 percent during the two-year period. The economy recovered in 1986, which gathered strength up to 1989. However, in view of the deep recession in 1984-1985, the 1980s proved to be a lost decade for the country.

Political shocks in 1990, including attempted coups d’état against the President of the Republic and natural disasters, caused economic activities to decline in 1991. A new Philippine President was elected in 1992, but an electric power shortage in 1992-1993 kept the economy weak, and it did not recover until 1994. The growth was sustained up to 1997, despite the emergence that year of the Asian financial crisis that was triggered by the devaluation of Thailand’s baht. The growth rate of real gdp slowed down, ending flat in 1998. A new president was elected in 1998, but lack of fiscal discipline under the new administration caused the budget deficits of the national government to balloon, which caused interest rates to rise and growth to slow down. Still, given the strength of the recovery in 1994-1997, a moderate growth rate in per capita income was posted in the 1990s.

Entering the first decade of the 21st century, several shocks—both economic and non-economic—again battered the economy. In 2000, allegations of the President’s involvement in an illegal numbers game led to the Philippine Senate impeaching the President, resulting in his eventual unseating in 2001. Nonetheless, the economy posted a positive growth rate in real gdp that year. Then the September 11 terrorist attack took place, an event that led the us federal government to declare war against terrorism. That war had serious repercussions on the Philippine economy, in view of the conflicts with Muslim rebels in the southern part of the country. In any event, the economy proved sufficiently resilient by posting an annual average growth rate in real per capita gdp of 2.53 percent over the period 2001-2010.

The economic growth in 2001 ushered in the longest recovery in the past few decades. Policymakers responsible for the economy managed to temper the output volatility, keeping business fluctuations moderate. The economy dodged a recession in 2009, with real gdp managing to grow 0.1 percent, in spite of the 2008 global financial crisis that was triggered by the collapse of the sub-prime housing loan market in the us. Still, the economy failed to realize its capacity output, as real gdp grew only by an average of 4 to 4.5 percent each year. The failure to reach full potential is often attributed to poor governance, mainly an inability to fight corruption in high places during the period 2001-2010.


In 2010, real gdp grew a strong 7.6 percent. This moderated to 3.6 percent in 2011. In 2012, real gdp recovered and grew 6.6 percent, the highest among Southeast Asian economies. Real gdp grew 7.1 percent in 2013 and 6.1 percent in 2014. Clearly, the country’s growth performance improved considerably over the period 2012-2014, with real gdp growth averaging 6.6 percent each year. The remarkable growth performance of the economy in the past three years is one reason for the improved assessments of the country’s macroeconomic performance in the future. Many observers of the Philippine economy attribute the high-growth rate to a combination of market reliance and good governance under the incumbent administration.

For a developing economy to catch up with more developed ones, its real per capita income must exhibit growth on a sustained basis. To provide a quantitative tone, in a country where real per capita gdp grows 6.9 percent each year, a doubling of real per capita gdp every 10 years can be expected. Several factors, both economic and non-economic, matter for growth of real per capita income in the long run [Barro 1994]. Among the economic factors, it is important in the short run to avoid unwanted business fluctuations through appropriate macroeconomic policies [Fischer 1993]. In the long run, it is vital for the economy to overcome the usual tendency towards diminishing marginal productivity in both factors of production: labor and capital. As the new endogenous growth theory emphasizes, such growth is rooted in capital accumulation, whereby capital is defined broadly to include not only physical, but also human, technological, and social overhead capital (Romer [1986, 1990]; Lucas [1988]).

For the Philippines, having experienced recurrent boom-and-bust cycles that forced the government to tap standby lending facilities of the imf on a recurrent basis, it is useful to revisit the macroeconomic policy reforms that have contributed to an economic recovery and laid the foundation for that recovery to gather strength. Meanwhile, among the long-run factors, it is widely recognized that human capital investment, especially education, is vital to economic growth. Acting in combination with other forms of capital, human capital can lead to increasing returns ([Romer 1986]; [Barro and Sala-i-Martin 1990]). Moreover, since education is a core value of the Filipino people that is enshrined in the Philippine Constitution of 1987, an investigation of the role of education in the growth process is warranted.

To begin with, this paper examines the various short-run macroeconomic policy reforms that the Philippines has pursued to minimize unwanted business fluctuations. Then, in relation to growth over the long run, the paper assesses the role that human capital accumulation, focused on education, has played. There is no question about government support for basic education. But in view of the important contribution of higher education to technological progress, the major source of long-run growth, public policy needs to pay more attention to it [Canlas 2015].


Moreover, since economic policy reforms dating back to 1986 have sought to integrate the Philippines with the rest of the world’s economy, the growth effects of foreign-trade policy reforms are useful to investigate. For instance, since 1986, political administrations have in succession pursued trade and investment liberalization, a policy measure that can be relied on to trigger growth over the long run.

The second section discusses the macroeconomic policy reforms designed to minimize unwanted business fluctuations. The third section looks at key aspects of capital accumulation that have long-run growth effects, focusing on human capital. The fourth section concludes the paper.

2. Macroeconomic policy reforms

Macroeconomic polices are intended to provide a stable and predictable environment for the private sector. Minimizing risk and uncertainty is vital in enabling households and enterprises to realize their consumption and investment plans, elements of aggregate demand that support growth.

Both fiscal and monetary policies have long been recognized as constraints to sustained growth in the Philippines, a point that was stressed in Canlas, Khan, and Zhuang [2008] and reiterated by the National Economic and Development Authority [2010]. Fiscal policy is often singled out as the most binding constraint to the country’s sustained growth, particularly, the relatively low tax effort, defined as the proportion of tax revenues to gdp. Fiscal and monetary policies, however, are closely related, and are commonly viewed in tandem.

Fiscal policy encompasses measures that affect the size of the public debt, including taxation and government spending. Monetary policy spans policies affecting the composition of that debt between money and non-interest-bearing liabilities of the government. When the monetary authority, the Bangko Sentral ng Pilipinas (bsp), accommodates by lending to the government to finance the latter’s budget deficit, money supply rises, which constitutes the non interest-bearing liabilities of the government. However, when the monetary authority does not accommodate, then the government through the National Treasury issues interest-bearing debt claims. In the short run, the choice of financing a deficit may entail differential impacts on aggregate output and the general price level.

2.1. Fiscal policy

Any deficit in the national government budget has to be financed. When the government borrows to finance the deficit, the public debt rises, resulting in heavier debt servicing. The fiscal history of the Philippines indicates persistent deficits in the national government budget. In the 1970s, for instance, the Philippine government pursued an expansionary fiscal policy in an attempt to ward off the recessionary


effects of the two oil price shocks in 1974 and in 1979. The Philippines entered the 1980s with a large public debt, largely foreign. When interest rates rose on a global basis in the early 1980s as a result of disinflation moves by the US Federal Reserve Board, the government found debt servicing quite burdensome, eventually forcing the government to default on foreign-debt servicing.

Following the debt moratorium and tapping of a standby loan from the imf, the government resorted to fiscal and monetary tightening. That ushered in the recession of 1984-1985, the longest in the postwar economic history of the Philippines, in which real gdp declined 11 percent during the two-year period.

The fiscal position of the government has long been at the core of the liquidity crises repeatedly faced by the Philippines. In a small open economy like the Philippines, the current-account deficit is identically the same as the budget deficit of the public sector and the savings-investment gap of the private sector. Table 2 shows such a decomposition of the current-account deficit three years before the government declared a debt moratorium. It is shown that the budget deficit of the government as a percentage of gdp, after a near-balance in 1980, rose to 4.8 percent in 1981 and ballooned to 5.3 percent in 1982. It moderated to 3.3 percent following the declaration of a debt moratorium in 1983.

TABLE 2. Current account, consolidated public sector account, and savings-investment gap as percentage of GDP

Year Current account

Consolidated public sector account

Savings-investment gap

1980 (5.9) 0.46 (6.4)

1981 (6.2) (4.8) (1.4)

1982 (9.0) (5.3) (3.7)

1983 (8.3) (3.3) (5.0)

Sources of basic data: Bangko Sentral ng Pilipinas for current account; Department of Finance and Department of Budget and Management for consolidated public sector account. Savings-investment gap is calculated as a residual.

The problem posed by the budget deficit of the national government returned in 1990, when the budget deficit reached 4.8 percent of gdp. Fiscal policy was subsequently tightened, forcing the consolidated public sector account to moderate to 2.1 percent of gdp in 1991. Real gdp growth slowed down to 0.5 percent that year.

A large government budget deficit as a percentage of gdp may also be an offshoot of weak economic growth. The tax system has automatic stabilizers that lose their vitality in an economy that’s slowing down. The stabilizers include revenues from personal income and corporate income taxes. This was evident in 2009, when the budget deficit of the national government swelled again to 3.9 per cent of gdp as a result mainly of the economic slowdown that emerged in the


aftermath of the 2008 global financial crisis. And so in 2010, the fiscal position of the government was still deemed as the most binding constraint to growth in the Philippines, triggering earnest calls even in government for fiscal reforms.

It is well recognized that government spending at all levels has to be reined in, while protecting core values in education, health, and infrastructure. But more importantly, it is widely agreed that insofar as raising the tax effort is concerned, there is still plenty of room for improvement. Table 3 shows figures on the tax effort (total tax revenues divided by gdp for the given year) for selected years. In the years prior to the 1983 foreign debt crisis, the tax effort was declining and averaging only about 10 percent each year. In 1986, a series of measures aimed at improving tax collection and tax administration started. From about 10.6 percent in 1985, the tax effort improved, peaking at 17 percent in 1997 before it weakened again.

TABLE 3. Tax effort (in %), selected years

Year 1980 1981 1982 1985 1990 1995 1997

Tax effort 11.4 10.3 9.9 10.6 14.1 16.3 17.0

Note: “Tax effort” is defined as total tax collection divided by GDP.Source of basic data: National Statistical Coordination Board, Philippine statistical yearbook.

In 1986, following the restoration of democratic political institutions, the administration under the leadership of President Corazon Aquino instituted a tax reform package. One major motivation for these reforms stemmed from the recurrent liquidity crises of the past that were rooted in the recurrent budget deficits of the government. In addition, the introduction of foreign-trade policy reforms gave rise likewise to the need to strengthen the internal tax system [Canlas 2000]. The Aquino administration started an import liberalization and tariff reduction program. As border or customs tariffs and trade taxes declined, strengthening the internal tax system became an imperative, particularly, personal and corporate income taxes and indirect taxes. Since reliance on foreign-trade taxes was being reduced, internal tax collection had to increase to compensate for the foregone border taxes. Table 4 shows the declining long-term shares of customs tariffs to total taxes and the increasing shares of internal taxes.

TABLE 4. Share of foreign-trade and internal taxes (in %)

Tax type 1980 1985 1990 1995 1997

Internal 56.9 69.4 68.5 67.7 76.4

Customs 38.0 27.8 30.2 31.4 23.0

Others 5.1 2.8 1.3 0.9 0.6

Total 100 100 100 100 100

Source: National Statistical Coordination Board, Philippine statistical yearbook


In 1987, President Aquino issued an Executive Order that introduced a 10-percent value-added tax on all sales transactions not explicitly excluded from the first Executive Order on value-added tax. In addition, a Republic Act was enacted that amended the National Internal Revenue Code. The latter introduced a “simplified net income taxation” for the self-employed and professionals in an attempt to widen the tax base, accompanied by a reduction in marginal tax rates.

The Ramos administration, which succeeded Aquino’s in 1992, resumed the tax-reform program. It expanded the coverage of the value-added tax law. In 1998, the end of the Ramos term, a Republic Act called the Comprehensive Tax Reform Program was enacted. The program further sought to expand the tax base while reducing marginal tax rates. One outcome of the reform measures covering tax policy and administration during the Ramos years was a significant increase in the tax effort. In 1997, the tax effort peaked at 17 percent, enabling the national government to post a budget surplus for the first time since the late 1970s.

In 1998, the Estrada administration took over the reins of government. The budget deficit of the national government expanded as spending rose without any accompanying tax measure. Mr. Estrada was unseated in 2001, but the imperative to raise tax revenues re-emerged.

The Macapagal-Arroyo administration assumed power in 2001. The tax effort weakened during this administration, down from 16.9 percent prior to the 1997 Asian financial crisis to 12.4 percent in 2004. And so pressure mounted to enhance tax collections. In response, the administration caused the enactment of a Republic Act called the Reformed Value-added Tax, which expanded coverage and raised the rate to 12 percent from 10 percent in 2007.

In 2010, the tax effort remained a major challenge to the administration of President Benigno S. Aquino. The latter had shepherded the enactment of a law in 2012 that restructured taxes on cigarettes and alcoholic beverages and indexed them to inflation. Accompanied by disciplined government spending, the budget deficit of the national government declined to about 2 percent of gdp in 2012.

Moving forward, the tax reform program will continue to rely a great deal on internal, rather than, border taxes. For direct taxation, this strategy calls for strengthening collection and administration of personal income and of corporate income taxes. A noticeable trend is the growing share of indirect taxes to total tax revenue mainly on account of improvements in value-added tax collection.

With the share of indirect taxes increasing, questions about tax incidence generally crop up. Indirect taxes are considered regressive since the burden on low-income taxpayers is disproportionately large. As a proportion of earned income, poor people pay a larger amount in taxes. There is little that can be done to reduce such burden, except to make government spending progressive, that is, low-income taxpayers should benefit more from any additional unit of government spending. In line with making government spending progressive, the PNoy administration has prioritized some spending programs in the national


government budget for low-income households, such as, the “conditional cash transfer” program for the poor.

2.2. Monetary policy

At this point, there is wide agreement among economists that monetary policy should be based on rules rather than discretion [Kydland and Prescott 1977]. The main reason for rules is to minimize the uncertainty and volatility that emanate from discretionary monetary policy. In doing so, monetary policy minimizes the risk from various sources that the private sector confronts.

Monetary policy rules in the Philippines have been evolving and undergoing refinements over time [Canlas 2012]. In the early 1970s, following the decision to allow the world’s major currencies to go on a generalized float against one another, the monetary authority followed suit by adopting a flexible exchange rate system. However, the central bank as monetary authority reserved the right to intervene in foreign-exchange markets to keep the exchange value of the Philippine peso against the us dollar within a targeted band. The central bank observed exchange-rate targeting as a monetary policy rule.

Under exchange-rate targeting, in which the central bank tried to keep the nominal exchange rate within a narrow band, the system in place resembled a fixed exchange-rate system [Canlas 2002]. When the central bank embarked on an expansionary monetary policy, the country’s inflation rate exceeded that of the us, thereby making imports from the latter attractive. As the demand for imports increased, the demand for us dollars likewise rose. When the central bank accommodated, the official foreign reserve assets that it held declined. The erosion continued amid non-stop inflationary policy. The situation eventually triggered a speculative attack against the Philippine peso, thereby accelerating the erosion of official foreign reserves. When the latter fell to a critically low level, the economy experienced serious liquidity problems, resulting in a balance-of-payments crisis and an abandonment of narrow exchange-rate targeting.

The central bank then allowed the exchange rate to move within a wider band. At the same time, it changed its policy stance and embraced monetary-aggregate targeting as monetary policy rule. One factor that influenced the adoption of such a monetary policy rule stemmed from the fact that in the 1990s, the Philippines was still tied to an imf program. In this context, the conduct of fiscal and monetary policies was based on financial programming techniques that the imf imposed on client countries experiencing balance-of-payments crises.

From the standpoint of the imf, a balance-of-payments crisis stems from an excess of aggregate demand over aggregate supply. Demand-management techniques are thus deemed essential. Inflationary monetary policy is to be curtailed. Money demand and supply need to balance. Disinflationary monetary policy means money supply growth has to decline. In the short run, however, disinflation results in adverse output effects.


Disinflation and recession occurred in the early 1980s, after the Philippine government declared a moratorium on foreign debt servicing and it tapped a standby loan from the imf for the needed liquidity. Sudden tightening of monetary (and fiscal) policy led to the recession of 1984-1985.

Monetary-aggregate targeting had its downside. Demand for money could be unstable at times, making such targeting ineffective. In 2002, the new bsp as monetary authority adopted inflation targeting.1 Under this monetary policy rule, the bsp announces a target inflation rate and uses its array of monetary policy tools to achieve the inflation target. When, for instance, the actual inflation rate exceeds the target, the monetary authority may cause interest rates to rise to tighten money supply and curb inflation. It is important for the bsp to maintain constant communication with the public so that its intentions are properly communicated and its messages shorn of uncertainty to the extent possible.

Inflation targeting requires that the bsp has both policy and instrument independence. Policy independence means that the fiscal authority does not influence bsp policy making. By instrument independence, the bsp must be equipped with enough tools of its own to achieve its inflation targets. In line with this independence, an institutional reform in 1993 created an independent bsp. One of the important provisions of the law calls for a monetary board; majority of the board’s members are appointed by the Philippine president from the private sector. Prior to this reform, members of the president’s cabinet who sat in an ex-officio capacity dominated the monetary board.

A prior issue is this: Is inflation a monetary phenomenon? Friedman said that “inflation is always and everywhere a monetary phenomenon”. With a quantity theory of money in mind, if the central bank increases the money supply from a position of balance, then the real money stock exceeds the demand for it. To restore balance, the general price level must rise, which means that inflation rate, defined as the percentage change in the general price level, must rise.

Canlas [1992] tested a quantity theoretic model of inflation in the Philippines. The theoretical model implies a one-for-one impact of money growth on the inflation rate. Some econometric techniques—including first-order differencing for stationarity, finding the optimal lag length, and Granger causality tests—are put to work. Time-series data are used. For the period 1973-1990, the regression results show that money growth has a contemporaneous positive effect on the inflation rate, but it is less than one-for-one. The optimal lag length of the effect of money growth on inflation using Hsiao’s test [1981] is shown to be one year; in subsequent regressions, however, the effect of money growth lagged one-year is insignificant. Overall, there is empirical support for an inflation model based on the quantity theory of money.

1 For an exposition on the intellectual development of inflation targeting, see Svensson (1999a, 1999b).


Since the adoption of inflation targeting as a monetary policy rule in 2002, the general price performance of the Philippine economy has improved tremendously. Over the period 2005-2012, the inflation rate averaged only 5 percent each year, except in 2009, when relative-price shocks emanating from food and energy products intervened and caused the inflation rate to increase to 8 percent that year. Since then, the inflation rate has again moderated to an average of about 4 percent each year.

Monetary policy can be trusted to promote growth insofar as it dampens inflationary expectations. As the latter is dissipated, interest rates decline with salutary effects on investment. An increase in investment is expected to support growth. However, an increase in physical capital is subject to diminishing marginal productivity, which in the long run dampens growth. The challenge, therefore, is to overcome this tendency towards diminishing marginal productivity in order to generate long-run growth.

3. Towards long-run growth

Proper conduct of macroeconomic policies can be trusted to minimize unwanted business fluctuation, a phenomenon to which all market-oriented economies are exposed. Having stabilized the economy, the associated major challenge is ushering in self-sustaining growth. This section looks at factors that contribute to long-run growth.

3.1. Investing in human capital

A major force that propels growth in the long run emanates from the accumulation of human capital, which is commonly associated with the knowledge and skills, or efficiency units, which people bring to the workplace. Given diminishing marginal productivity of physical capital, accumulation of the latter alone is not sufficient to yield positive growth in the long run. The model of Solow-Swan [1957] emphasized the importance of investment for growth.2 It is insightful in terms of accounting for growth in a given country, but it is not designed to account for growth convergence in a large cross-section of heterogeneous countries. Modern economic growth theory, or what is often referred to as endogenous growth, has thus extended the Solow growth model in trying to account for long-run growth in a number of developed countries and to account for convergence (or lack of it) in a large sample of countries.3

Human capital accumulation takes various forms, including investments in education, training, health, and nutrition. Health investments enable people

2 See Canlas [2003] for tests of the Solow model in the Philippine context. 3 For an exposition on the origins and key doctrines of endogenous growth, see Romer [2000].


to develop their physical skills, such as manual dexterity and visual acuity.4 However, it is education and training that are popularly linked to human capital investments. Various levels of education, for instance, are credited with helping raise literacy and cognitive skills and equipping people with knowledge and scientific skills. Moreover, human capital investment is associated with other aspects of household behavior, including desired family size and labor force participation (Becker, Murphy, and Tamura [1990]; Tamura [2006]). As a result, many countries, especially those that have succeeded in achieving growth in real gdp over a long-run period, have made it a point to invest in quality education at all levels.

In empirical studies that seek to determine the sources of growth, much of the growth stems from total factor productivity (tfp). The latter refers to the efficiency in use of all factors of production, e.g., labor and capital. The role of education in enhancing tfp is well understood and widely accepted in the economics profession. It seems clear that education and training equip people with efficiency units that are useful in the workplace [Becker 1961, 1964]. On the job, people accumulate more skills, which further raise their human capital, as Mincer [1962] emphasized.

At the same time, however, family members’ uses of time include non-market activities—such as attending to the education, health, and nutritional needs of children—which lead to the accumulation of human capital that eventually gets transmitted within the household across generations [Becker 1965]. In countries that are able to produce across time highly trained and educated workers who can master production techniques that continuously emerge in a modernizing economy, aggregate production efficiency is increased in the long run, with positive effects on tfp.

Following Solow [1957], Denison [1962], and Jorgenson and Griliches [1967], the growth rate of output is decomposed into the sum of the growth of labor and capital plus tfp, which is taken as a residual. Early empirical studies tended to show that much of the growth rate of output stems from growth of tfp. The latter is viewed as technological progress, which emanates from a variety of factors, including research and development (r&d) or knowledge production, and realization of scale economies. Technological progress follows as a matter of course from knowledge production, which is human-capital intensive and has the dimensions of a public good [Grossman and Helpman 1989]. That is, knowledge generated by one firm may spill over to all firms in the industry. And so as tfp improves, the tendency towards diminishing marginal productivity is overcome, resulting in positive growth rates of output over a long period of time.

4 For empirical tests supportive of the positive growth effects of education on output growth using an endogenous growth model, see Benhabib and Spiegel [1998].


Attempts to decompose the sources of output growth in the Philippines shows that in the 1980s and the 1990s, much of the growth of output was due to increases in labor and capital, while tfp contributed only a small proportion, as shown in Table 5. In the 1980s, tfp was a minus 1.62 percent. This turned positive 0.25 percent in the 1990s, accounting for 9.9 percent of the growth rate of real tfp. Over the period 2001-2006, tfp gained strength and increased to 2.4 percent, which is about 50 percent of real gdp growth rate.

TABLE 5. Total Factor Productivity in the Philippines

Period Growth of capital Growth of labor Total Factor Productivity

1981-1990 2.05 1.37 (1.62)

1991-2000 1.77 0.87 0.25

2001-2006 1.12 1.24 2.41

Source: Table 2.9 in Canlas, Cham, and Zhuang [2009]

To account for the improvements in tfp, it is insightful to note the trends in educational attainment in the Philippines and its impacts on the skill composition of the labor force. A rising share of skilled workers normally contributes to tfp improvements.

3.2. Investing in education

Recognizing the significance of education in the process of economic growth and development, the Philippine Constitution of 1987 provides for allocating the biggest proportion of the expenditure program of the national government budget to education. The national government has long been the major provider of education in the Philippines, particularly, basic education, which consists of elementary and secondary education levels. The latter are provided without out-of-pocket costs to parents who enroll their children of school age in public schools.

At the tertiary education level, the national government runs a network of state colleges and universities, wherein tuition fees are charged, but at subsidized rates. The rest of the demand for tertiary education is met through private colleges and universities.

Moreover, the government supports skills training through a system of private post-secondary schools offering technical and vocational education and training. These private institutions co-exist with government-run institutes.

Table 6 shows the distribution of workers by educational attainment. It is seen that the proportion of unskilled workers (those with only elementary schooling) has over the past two decades been declining. Meanwhile, the share of semi-skilled to skilled workers (those with at secondary and tertiary schooling) has been rising.


TABLE 6. Employment by educational attainment (in %)

Education 1991 2001 2006

Elementary school graduate 24.2 19.2 16.8

High school graduate 18.1 23.2 24.9

Tertiary and above 10.9 13.2 14.4

Source of basic data: National Statistics Office, Philippine labor force survey, various issues

Investments in education and training transmit growth through another mechanism. By narrowing the differentials between earnings, income inequality is reduced. Persistent income inequality tends to be deleterious to growth in so far as such inequality raises political pressure to implement redistributive programs that are financed by raising taxes. Raising taxes on personal and business incomes tends to erode the marginal productivity of labor and capital that can be privately captured, with dampening effects on capital accumulation, both human and physical, and on growth. However, by reducing the proportion of poor individuals and households, political pressure for redistribution may be eased.

3.3. Technological progress, trade policy, and growth

It is widely known that growth comes not just by producing increasing quantities of the same good, but also by producing new products of ever-increasing quality [Stokey 1995]. The growth of Apple Company, for example, comes not from producing the same first-generation Mac Book or iPad but from making newer and more powerful versions of these products. Moreover, further growth is obtained not by selling in the same market but by accessing new markets for an expanding array of new products. Trading countries differ in their endowments of knowledge capital. The developed countries with large endowments of knowledge capital do a good deal of r&d and are thus able to produce and trade a wide range of new products [Grossman and Helpman 1989]. Over time, developing countries that adopt open trade regimes may acquire production knowledge about some of these new products through imitation, with salutary effects on output growth. Clearly, technological progress and trade policies matter a lot for growth.

Trade in intermediate products is one of the fastest growing segments of international trade. These intermediate products are normally the product of investments in r&d; they lie at the core of an increasing number of new products of higher quality. In the crop sector of agriculture, for example, the discovery of high-yielding seed varieties has significantly raised the yields per hectare of rice and corn. In information technology, the microchip revolution has enabled the manufacture of supercomputers with huge memories and storage capacity and


allowed various industries in the service trades, such as banking and finance, and retail and wholesale trade, to exponentially increase their productivity.

The rise of intermediate products emerging from r&d is taking place in a system involving separable and multi-stage production processes. This has given rise to outsourcing. Firms in developed countries subcontract to firms in less developed countries the labor-intensive stages of production, for instance. But this can also be done under a vertical-integration arrangement if the firm from the developed country finds it more efficient to do so.

Considering that r&d yields knowledge that spills over to entities beyond the r&d proponents themselves, there is a public policy concern that it may be under-produced. Third parties are able to capture the returns from r&d but the original r&d proponents may not be fully compensated. People tend to invest only up to the marginal rate of returns that they can appropriate. If the private returns fall short of the social rate, then there is room for some form of subsidies to avoid under-investment in knowledge production.

3.4. Philippine trade policy reforms

In 1986, the Philippine government ushered in some trade policy reforms aimed at integrating the economy with the global economy and make local firms internationally competitive. To kick-start this reform process, the government liberalized imports and embarked on a tariff-reduction program. Later on, it also liberalized entry of foreign direct investments.

In the 1950s, trade policy was based on import substitution. The government encouraged domestic firms to manufacture goods that at the time were being imported in large amounts, ostensibly to save on scarce foreign exchange and to “learn by doing”. Local firms were protected from competing imports through high import tariffs and quantitative restrictions, using infant-industry arguments. The firms, however, relied on imported capital equipment and intermediate products. Since they were selling their products solely in the domestic market, the goal of saving on foreign exchange was not achieved. Moreover, since the Philippines was starting with a low human capital base, new products that could be exported did not emerge. The foreign exchange earnings of the country were derived largely from exports of agricultural and mineral products. Since these foreign exchange earnings were not adequate to finance the import needs of the import-substituting industries, the country experienced recurrent liquidity problems, accompanied by the collapse of a fixed peso-us dollar exchange rate. Eventually, reliance on import substitution declined. Import liberalization and tariff reduction combined to shrink the size of import-substituting industries.

In 1994, the Philippine acceded to the World Trade Organization, which was anchored on a “most-favored nation principle”. Under this principle, tariffs and other commercial policies extended to one member country of the World Trade Organization cannot be withheld from other member countries. In


addition, the Philippines has actively participated in regional preferential trading arrangements. One of these is the asean Free Trade Area-Comprehensive Effective Preferential Tariff, which envisions non-tariff barriers among member countries, starting with manufactured products, and later on, extended to trade in agricultural goods and in services. The arrangement has evolved: In 2015, asean is committed to instituting a single-market arrangement.

Today the Philippines is a small open economy that is integrated with the rest of the world economy through trade in commodities, securities, and national currencies. Greater international labor mobility is also envisioned. The country adopts a flexible exchange-rate system and allows international mobility of capital. It has significantly reduced its public debt as a proportion of gdp, and its sovereign debt papers are nearing investment-grade status.

4. Concluding remarks

Several development observers regard the Philippines at this point as an emerging economy that is poised to join the Southeast and East Asian economies that are now considered newly industrializing. This optimism derives largely from the structural reform program that has been pursued by political administrations in succession since 1986. The policy reforms may be broadly classified into the following: short-term macroeconomic policies designed for stable growth; and long-term structural policy reforms aimed at sustained and broad-based growth.

The importance of short-run stabilization may be appreciated by taking a historical perspective. For nearly four decades since the 1950s, growth in the Philippines was hampered by inappropriate macroeconomic policies that resulted in unwanted volatility and business fluctuations. Political administrations since 1986 have in succession tried to address these concerns. Beginning in 1986, following the dismantling of martial-law rule and the restoration of democratic political institutions, the administration exerted efforts to strengthen the fiscal position of the national government by instituting a responsible deficit-reduction program. Monetary policy was made independent with the enactment of a law that established the bsp as a monetary authority independent of the fiscal authority. These measures have ushered in economic recoveries from balance-of-payments crises and gdp declines, while putting a lid on inflation.

Moreover, attention has been paid to growth-enhancing reforms, such as increased investments in capital broadly defined to include human and knowledge capital, in an effort to achieve long-run growth. In addition, the government pursued trade liberalization policies that are anchored on import liberalization and tariff reduction. The government also liberalized foreign direct investments, which allowed a hundred percent ownership in several industry areas; restrictions or limits on foreign ownership were explicit in the so-called negative lists that over time are being shortened. The privatization of government


firms that previously monopolized some industry sectors—such as oil and petroleum, commercial banking, and telecommunications—was pursued.

The economy has responded positively to these policy reforms. In 2001, the economy recovered from a political shock and posted a gdp growth that has been sustained in spite of the intervention of shocks like the September 11 terrorist attack, the global financial crisis of 2008, and the Eurozone debt crisis of 2012. The economy has achieved sufficient resilience as it became integrated with the global economy through trade in commodities, securities, and national monies. Entering the second decade of the 21st century, the economy is expected to achieve a sufficiently high growth rate in real gdp with stable prices amid continuing challenges like the weak recovery in the US, the debt problems of some Eurozone countries, and the growth slowdown in China.

The current administration is restoring credible leadership at the top of the political ladder, and this is serving the economy well. Not only is the current political leadership committed to market-friendly policies; it is likewise aggressive in addressing concerns about good governance, including fighting corruption and strengthening the legal and judicial system. The latter is vital to making the infrastructure program of the government at both the national and local levels succeed. Since the program is anchored on public-private partnership, contractual performance and sound adjudication in case of contractual disputes are critical. Investors find these policy commitments quite appealing, which are necessary conditions to becoming an economic miracle.

Philippine National Oil Company Alternative Fuels Corporation

References

Barro, R. [1994] “Economic growth and convergence”, cs Press for International Center for Economic Growth, Occasional Paper no. 46.

Barro, R. and X. Sala-i-Martin [1990] Economic growth. New York: McGraw-Hill, Inc.

Becker, G. [1964] Human capital: a theoretical and empirical analysis. New York: Columbia University Press for the National Bureau of Economic Research.

Becker, G. [1965] “A theory of the allocation of time”, Economic Journal 75:493-517.

Becker, G., K. Murphy, and R. Tamura [1990] “Human capital, fertility, and economic growth”, Journal of Political Economy 98:S12-S37.

Benhabib, J. and M. Spiegel [1994] “The role of human capital in economic development: evidence from aggregate cross-country data”, Journal of


Monetary Economics 34:143-173.Canlas, D. [1992] “Inflation in a low-income country: tests based on the quantity

theory of money”, University of the Philippines School of Economics Discussion Paper no. 9209.

Canlas, D. [1998] “Tax reforms in the Philippines”, Asian Tax Reforms: Issues and Results, proceedings of an international symposium, Tokyo: Hitotsubashi University.

Canlas, D. [2002] “Monetary policy and exchange rate in the Philippines”, Augustine H.H. Tan, ed., Monetary and financial management in Asia in the 21st century. New Jersey: World Scientific.

Canlas, D. [2003] “Economic growth in the Philippines: theory and evidence”, Journal of Asian Economics 14:759-769.

Canlas, D. [2012] “Business fluctuations and monetary policy rules in the Philippines: with lessons from the 1984-1985 contraction”, presentation delivered as the Bangko Sentral ng Pilipinas Sterling Chair in Monetary and Banking Economics.

Canlas, D., R. Cham, and J. Zhuang [2008] “Development performance and policy”, in D. Canlas, M.E. Khan, and J. Zhuang eds. chapter 2.

Canlas, D., M.E. Khan, and J. Zhuang eds. [2008] Diagnosing the Philippine economy: toward inclusive growth. London: Anthem Press.

Canlas, D. [2015] “Investing in human capital for inclusive growth: focus on higher education”, paper presented on September 22, 2015 in observance of Development Policy Research Month, Philippine Institute for Development Studies, Metro Manila.

Denison, E. [1962] “Sources of growth in the United States and the alternatives before us”, Committee for Economic Development, Supplement Paper no. 13.

Fischer, S. [1993] “Does macroeconomic policy matter? Evidence from developing countries”, CS Press for International Center for Economic Growth, Occasional Paper no. 27.

Grossman, G. and E. Helpman [1989] “Product development and international trade”, Journal of Political Economy 97:1261-1283.

Hsiao, C. [1981] “Autoregressive modeling and money-income causality”, Journal of Monetary Economics, 85-106.

Jorgenson, D. and Z. Griliches [1967] “The explanation of productivity change”, Review of Economic Studies 34:249-280.

Kydland, F. and E. Prescott [1977] “Rules rather than discretion: the inconsistency of optimal plans”, Journal of Political Economy 473-491.

Lucas, R. Jr. [1988] “On the mechanics of economic development”, Journal of Monetary Economics 22:3-42.

Lucas, R. Jr. [1993] “Making a miracle”, Econometrica, 61:251-271.Mincer, J. [1962] “On-the-job training: costs, returns, and some implications”,

Journal of Political Economy 70:50-77.


National Economic and Development Authority (neda) [2010] A strategic framework and action plan for inclusive growth. Metro Manila.

Romer, P. [1986] “Increasing returns and long-run growth”, Journal of Political Economy 94:1002-1037.

Romer, P. [1990] “Human capital and growth: theory and evidence”, Carnegie-Rochester conference series on public policy 32:251-286.

Romer, P. [2000] “The origins of endogenous growth”, Journal of Economic Perspectives 8:3-22.

Sharma, R. [2012] Breakout nations: in pursuit of the next economic miracles. New York: W.W. Norton and Co.

Svensson, L. [1999a] “Inflation targeting as a monetary policy rule”, Journal of Monetary Economics 607-654.

Svensson, L. [1999b] “Inflation targeting: some extensions”, Scandinavian Journal of Economics 337-361.

Solow, R. [1957] “A contribution to the theory of economic growth”, Quarterly Journal of Economics 70:65-94.

Stokey, N. [1995] “r&d and economic growth”, Review of Economic Studies 62:469-489.

Swan, T. [1957] “Economic growth and capital accumulation”, Economic Record 32:334-361.

Tamura, R. [2006] “Human capital and economic development”, Journal of Development Economics 79:26-72.


PRE

Development progeria: the role of institutions and the exchange rate

Sarah Lynne S. Daway* and Raul V. Fabella*

Convergence is more the exception than the rule in the development landscape. As a possible explanation, we posit development progeria: the phenomenon where a low-income country exhibits the industrial share dynamics of high-income mature economies where the Non-traded Goods Sector outgrows the Traded Goods Sector and the share of the non-traded goods sector outstrips the share of the traded goods sector. We argue that this seems to be the case of the Philippines in the last 25 years.

We then inquire into the drivers of this phenomenon. One possibility is the Rodrik hypothesis: that market and institutional distortions hamstring the Tradable goods sector more than they do the Non-tradable goods sector. The other possibility is the exchange rate policy being favorable or unfavorable to the Tradable goods sector. Using cross-country data for countries with per capita income of us$10,000 or less, we show that these two factors cannot be rejected as drivers of development progeria.

JEL classification: 014, 043, F31Keywords: Development progeria, institutions, real exchange rate, low-income economies

1. Introduction

1.1. What is development progeria?

Development progeria is the phenomenon where a low-income economy exhibits the industry share dynamics considered normal in advanced high-income economies: that is, where the share of the modern Tradable goods sector falls while that of the Non-tradable goods rises in the course of development. In medicine, progeria is a genetic malfunction where afflicted six-year olds exhibit


the physical characteristics of sixty-year-olds. Like advanced mature economies that are characterized by slow growth, economies afflicted by development progeria also grow slowly and thus have poor prospects for catching up with mature economies. In contrast, economies on the non-progeriac or convergent trajectory experience an extended period of growing industrial share of Tradables coincident with the retreat of Non-Tradables share in total output before they finally graduate to mature economy status.

1.2. The Philippines in the past 25 years

For the most part of the quarter-century following the 1986 overthrow of the authoritarian regime of Ferdinand Marcos, the Philippines has chafed under the moniker “sick man of Asia.” It sank to the bottom of the asean 5 in per capita growth, rate of poverty reduction, and investment rate [Usui 2011]. Figure 1 shows the share trajectories of Agriculture, Manufacturing, Industry, and Services.

Source of basic idea: World Bank Development Index

FIGURE 1. Trajectory of industry sector shares in GDP: Philippines, 1986-2009

It is clear that over the 25-year period from 1986-2009 the share of the Services sector gained continually while the shares of Industry and Manufacturing stagnated or lost out. To compare the experience with those of select neighbors and mature economies in the same period, Figure 2 below, which is taken from Fabella and Fabella [2010], presents the change in the percentage share of the Manufacturing, Industry, and Services sectors in the period 1986-1996 by countries.

50.0

40.0

30.0

20.0

10.0

0.0

Sha

res

(%)

Agriculture

60.0

Manufacturing

Industry

Services

’86 ’87 ’88 ’89 ’86 ’87 ’88 ’89 ’90 ’91 ’92 ’93 ’94 ’95 ’96 ’97 ’98 ’99 ’00 ’01 ’02 ’03 ’04 ’05

86 Daway and Fabella: Development progeria: the role of institutions and the exchange rate

Source of basic idea: World Bank Development Index

FIGURE 2. Change in percentage shares, 1986-1996

Starting with the mature economies of Germany and the United States, the Services sector share rose while the shares of both Industry and Manufacturing for both periods fell. These are archetypes of late mature economy trajectory. South Korea, a new Organisation for Economic Co-operation and Development economy, saw its Services sector and Manufacturing shares rise but still with a slight increase in Industry share. Indonesia and Thailand saw their Service sector shares falling in the face of rapidly rising Industry and Manufacturing shares in both periods. This is the archetypal catch-up trajectory. Malaysia exhibited a rising Service sector share in this period but with rising Industry and Manufacturing shares. China exhibited the same pattern.

Figure 3 is taken from Fabella and Fabella [2010] and Fabella [2013]. It shows the trajectories for the second period (1996-2009).

Malaysia reverts somewhat to the archetypal catch-up trajectory with a falling Service sector share to go with a rising Industry but falling Manufactures share. China continues the pattern where both the Industry and Manufacturing sectors gain to go with a rising Service sector share. It is an economy firing on all cylinders.

Cha

nge

in %

sha

res

10.0

8.0

6.0

4.0

2.0

0.0

-2.0

4.0

-6.0

-8.0

Manufacturing

Industry

Services

Indon

esia Th

ailan

dMala

ysia

Philipp

ines Chin

a

South

Korea

German

y

USA


Sources: Fabella and Fabella [2010] and Fabella [2013] (Transactions of the NAST)

FIGURE 3. Change in percentage industry shares, 1996-2009

Finally, we turn our attention to the Philippines. The trajectory exhibited by the Philippine industrial structure in the last quarter-century mimics that of late mature economies: in both periods, Industry and Manufacturing shares declined while Services share rose. By the end of the second period, the Services sector share stood at the 55.4 percent, a feature common to Organisation for Economic Co-operation and Development and mature economies. Premature economic ageing accompanied its journey throughout the 25-year period. This also means that its prospect for convergence with mature economies is poor.

The question we confront is this: How does development progeria get engendered? We explore how the growth of the Tradable goods and the Non-tradable goods sectors are differentially affected by the exchange rate and the quality of governance. In the next section, we formally generate the relationship.

2. The model

2.1. A small open economy

We consider a model of the small open economy with two sectors, the tradable sector T and the non-tradable sector N, each using two factors, labor L and capital K, where K is a specific factor of production and thus is not mobile across sectors. We assume full employment of labor, L = L

T + L

N, where L

T is the labor used in T

and LN is used in N. Suppressing K, we can write the production technology of T

and N respectively as:

Cha

nge

in %

sha

res

8

6

4

2

0

-2

4

-6.0

Manufacturing

Industry

Services

Indon

esia

Thail

and

Malays

ia

Philipp

ines

China

South

Korea

German

y USA


T = Af(LT), f ' > 0 and f" < 0,

N = Bg(LN), g' > 0 and g" < 0,

where A > 0 and B > 0 are positive Hicks-neutral technical parameters. The equilibrium allocation of L (we assume an undistorted labor market) is given by the first order conditions:

Af '(LT) = w,

Bg'(L – LT) = w,

which combines to give Af '(LT) = Bg'(L – L

T). The latter can be solved for the

equilibrium LT* (and thus L

N*). Since progeria is at its core about differential

growth rates, we are interested initially in how T and N will grow as L grows. Totally differentiating and solving for (dL

T*/dL), we get:

(dLT*/dL) = [Bg"]/[Af" + Bg"] > 0.

The algebraic structure of (dLT*/dL) is identical to that of the “power of the

contract” in contract theory, where the efficient allocation of risk dictates that more risk is shouldered by the agent who suffers least from risk. If g" = 0, or the marginal productivity of labor in N is constant, but f" ≠ 0, all of the additional labor will be efficiently absorbed in N. Thus, the sector which suffers the least fall in marginal productivity from additional hiring should employ most of the additional labor at equilibrium.

2.2. The power of T

The “power of T” is the power of the tradable goods sector T to efficiently employ the increase in labor endowment. It is given by (dL

T*/dL).

Note that the power of T is the inverse of the capacity of N to efficiently absorb the additional labor. Thus, additional labor will be efficiently absorbed by that sector where the wage rate falls less with additional supply.

2.3. Market or institutional distortions and development progeria

It is a known fact that least-developed countries (ldcs) are generally beset by a myriad of institutional and market distortions. Rodrik [2008] argued and showed evidence that these market and institutional failures on the supply side adversely affect T and N differentially, T worse than N. We assume here that T and N are affected through the technical parameters A and B, respectively.


Let the universe of supply side market and institutional failures (but excluding the labor market) be summarized by a single parameter D > 0 and let A(D) and B(D) represent the relationship between the Hicks-neutral parameters and D. The higher is D the smaller are both A and B so that A' <0 and B' < 0. It is understood that D is lowest in mature developed economies.

While D adversely affects both T and N, T is more adversely affected by D than is N. We define this differential response: The Rodrik Differential Response Condition: Suppose D* and D** are two levels of D, D* < D**, then

A(D*)/B(D*) > A(D**)/B(D**)

or in continuous terms,

–[A'(D)/A(D)] > – [B'(D)/B(D].

In other words, the Hicks-neutral productivity of T is pulled down more than that of N by a rise in D. In terms of the production frontier, a higher D shifts the production frontier inwards throughout, but a drop in the intercept in the T axis is more than the drop in the N axis.

We are interested in how a rise in D ceteris paribus affects the composition of output at equilibrium. The following is shown in another paper [Fabella 2015].

Lemma 1: At equilibrium output basket, a rise in D ceteris paribus will reduce the labor used in T,

that is, (dLT*/dD) < 0, if the Rodrik Differential Response Condition holds.

Therefore, the share of T in total output will fall and that of N will rise as D rises. We now turn our attention to the response of relative growth rates of sectoral outputs when D rises ceteris paribus. In the same paper by Fabella [2015], the following is further shown.

Lemma 2: (dT/T) – (dN/N) < 0 as D rises ceteris paribus, if the Rodrik Differential Response Condition holds.

The growth rate of T decreases more than that of N when D rises. When local market and institutional distortions are large (D high), the share of Tradables will, over time, be lower and the share of Non-tradables higher. The more distorted are the market and institutional environments, the more likely is development progeria.


When local market and institutional distortions are large (D high), the share of tradables will over time be lower and the share of Non-tradables higher than otherwise similar countries but with lower D. The more distorted is the local market and institutional environment, the more likely is development progeria relative to another country otherwise identical. The hypothesis corresponding to the above follows:

Hypothesis 1: The better is the governance quality in the economy, the higher the share of the Tradable goods sector in total gdp.

3. The exchange rate and development progeria

Yet another factor that possibly contributes to development progeria is the value of the domestic currency. Rodrik [2008] has shown that overvaluation is bad for the growth of low-income countries. Though we will not directly tackle overvaluation in this paper, we are interested in the impact of the value of the domestic currency on the industry shares. The production frontier of this simple open economy given D* is

T(N; D*) = A(D*)fL – g-1[N/B(D*)]

where g-1(.) is the inverse function of g(.) assumed to exist and is convex and increasing (that is, concave and increasing g(.) is assumed to be a one-to-one map). The first derivative of T(N; D*) with respect to N is

TN(N; D*)= A(D*)f '[–g-1'/B(D*)] < 0.

The production frontier is well-behaved, i.e., downward sloping and strictly concave to the origin. Being small and open, the economy is best off producing at the point in the frontier characterized by

–TN(N; D*) = EP

T/P

N

where PT is the price of a unit of tradables in foreign currency determined in the

world, and PN is the price of a unit of non-tradables in domestic currency (say,

the Philippine peso) determined in the domestic market, and E is the amount of domestic currency that exchanges for a unit of foreign currency. EP

T/P

N is known

as the real exchange rate and is effectively the price of tradables in terms of non-tradables in domestic currency units. The higher is (EP

T/P

N), the lower is the non-

tradable sector output in the output mix of the economy or the larger is T’s share in total gross domestic product. The opposite result happens when (EP

T/P

N) is

lowered ( a real appreciation of the domestic currency).

Hypothesis 2: The higher is (EPT/P

N), the higher is the share of T in gdp.


4. Estimation results

We employ Blundell and Bond’s [1998] and Windmeijer’s [2005] two-step system-generalized- method-of-moments (system-gmm) procedure to estimate the equation below:

yit = βχχit + βZZit + μit, (1)

where yit is the ratio of the value-added of the tradable sector to gdp in country i in period t. χit is a vector containing predetermined and endogenous regressors, which may include the lagged values of the dependent variable; Zit is a vector of strictly exogenous regressors; and μit is the error term containing the fixed-individual effects.

The main advantage of the two-step system-gmm procedure is that it enables one to account for the endogeneity of the regressors by allowing one to use the lagged values of both the dependent and independent variables as instruments. It also treats the Nickell bias, which is ubiquitous in macro-panel datasets with large n (cross-section length) and small t (number of periods). Moreover, Windmeijer’s two-step correction procedure generates more precise and more efficient estimates, mitigating the finite-sample bias. Furthermore, we favor the two-step system-gmm procedure over Arellano and Bond’s [1991] difference-gmm model, as the former is more appropriate for dealing with variables that are or close to “random walk,” which most, if not all macroeconomic variables are purported to be. Lastly, information loss due to differencing in unbalanced panel datasets is less severe under the two-step system-gmm model than under the difference-gmm model.

To estimate Equation 1, we employ the ratio of annual manufacturing value added to gdp as a proxy for the ratio to gdp of the tradable goods sector, considering the fact that most manufactured goods are tradable and that manufacturing value added is more readily available in cross-country datasets.

The XX vector comprises the following: lags of manufacturing growth; the annual growth of the services sector value-added, which proxies for the non-tradable sector; the International Country Risk Guide (icrg) index, which is a measure of the quality of institutions; the ratio of purchasing power parity conversion factor to the market exchange rate, which substitutes for the real exchange rate, with a higher ratio implying a decline in the competitiveness of locally produced goods, thereby hurting the tradable goods sector1; the average

1 The ratio of purchasing power parity conversion factor to market exchange rate measures the amount of us dollars required in the local economy to purchase the same basket of goods and services that a dollar can purchase in the United States. Thus, a higher ratio signifies that the same basket of goods and services becomes relatively more expensive in the domestic economy. We prefer this series over the real effective exchange rate, as the latter has considerably less observations.


tariff rate on manufactured goods, which is a measure of distortion in the tradable goods sector; and fixed capital formation as a percentage of gdp, which is purported in theory as the main driver of output growth. We also include the “power of N,” which is analogous to the power of T and is computed as follows:

ΔLN / ΔL = [(ΔL

N / L

N) / (ΔL / L)] (L

N / L), (2)

which is the ratio of the growth rate of employment in the services sector (the non-tradable sector) to the growth rate of total employment multiplied by the ratio of employment in services to total employment. As with the power of T, the power of N is a measure of the absorptive capacity of the services sector or of its ability to effectively employ the increase in labor supply. As this is negatively related to the power of T, we then expect it to negatively affect the manufacturing to gdp ratio.

The Z vector consists of a measure of tropical land area to account for country-specific characteristics and of period dummies. This is in line with the “Tropical Paradox,” which is based on the observation that tropical countries closer to the tropical zone have lower per capita incomes than countries in more temperate climates. The period dummies are included to control for time-related shocks.

Except for the icrg index, the rest of the data are downloaded from the World Development Indicators website. The dataset is an unbalanced panel of 51 developing countries2 with real GNIs per capita of at most US$10,000, spanning six periods from 1984 to 2013.3 Each period is an average of five years, as is typical in cross-country growth regressions, to minimize the impact of short-run fluctuations on the estimated parameters.4

Table 1 below presents the regression results and the instrumentation details.5 With 51 instruments created using the second to fifth lags of the dependent and the appropriate independent variables, the estimated model passes the requisite diagnostic tests, i.e., the Arellano-Bond test of ar(2) in first differences and the Hansen J-tests of over-identifying restrictions. As a further test of the model’s validity, we also checked that the estimated coefficient of the lagged value of manufacturing growth lies between the Ordinary Least Squares regression and fixed-effects estimates. Indeed, the Ordinary Least Squares estimate is 1.19, which is greater than the two-step system-gmm estimate of 1.18, which is, in turn, greater than the fixed-effects estimate of 0.60 (see Table A2 in the appendix).

2 See the appendix for a list of countries included.3 We begin with 1984 due to data limitations.4 Due to the unbalanced nature of the panel, the effective number of periods is only 2.47 on average.5 The summary statistics of the regression variables are in Table A1 in the appendix.


TABLE 1. Two-step system GMM results

Dependent variable: Manufacturing-to-GDP ratio

Variable Coefficient Standard error

t-statistic p-value 95% confidence interval

Manufacturing-to-GDP ratio (-1) 1.18 0.03 35.47 0.00 1.11 1.25

Manufacturing-to-GDP ratio (-2) -0.33 0.03 -10.15 0.00 -0.39 -0.26

Power of N (-1) -0.06 0.01 -4.71 0.00 -0.09 -0.04

Services growth rate -0.13 0.04 -3.34 0.00 -0.20 -0.05

Fixed capital formation as percentage of GDP 0.09 0.03 3.18 0.00 0.03 0.14

ICRG index 0.05 0.02 2.74 0.01 0.01 0.08

Exchange rate (purchasing power parity-adjusted) -0.08 0.01 -7.49 0.00 -0.10 -0.06

Tropical area -0.51 0.33 -1.58 0.12 -1.17 0.14

Period 3 1.31 1.03 1.27 0.21 -0.76 3.39

Period 4 1.52 0.91 1.67 0.10 -0.31 3.35

Period 5 0.92 0.92 1.01 0.32 -0.92 2.77

Period 6 1.04 0.85 1.22 0.23 -0.68 2.76

Number of observations: 119

Number of groups: 51

Number of instruments: 51

Diagnostic tests

Arellano-Bond test for AR(1) in first differences: Prob > z = 0.006

Arellano-Bond test for AR(2) in first differences: Prob > z = 0.185

Hansen J-test of over-identifying restrictions: Prob > chi2 = 0.766

Difference-in-Hansen tests of exogeneity of instrument subsets:

GMM instruments for levels

Hansen test excluding group: Prob > chi2 = 0.527

Difference (null H = exogenous): Prob > chi2 = 0.824

GMM instruments: lagged manufacturing-GDP ratio, 2nd-5th lags



GMM instruments: lagged power of N, lagged services growth, exchange rate, 2nd lag



GMM instruments: fixed capital formation (percentage of GDP), lagged ICRG, 2nd-5th lags



IV instruments: Period 1, Period 2, Period 3, Period 4, Period 5, Period 6, eq(level)



The results in Table 1 provide evidence for the genesis of development progeria: the tradable sector, i.e., manufacturing, is adversely affected by an expansion of the non-tradable sector, i.e., services. The power of N is statistically significant and has the expected negative effect on manufacturing growth. Moreover, a higher growth rate of the services sector stunts the manufacturing sector, as is observed in development progeriacs.


In line with the theoretical results, icrg, the real exchange rate proxy and the average tariff rate on manufactured goods have the expected effects on manufacturing growth. icrg has a positive significant coefficient, which is consistent with the theoretical result that good institutional quality has a strong impact of spurring the tradable sector. The coefficient of the real exchange rate proxy also has a negative significant effect: A higher ratio of the purchasing power parity conversion factor to market exchange rate implies that a given basket of goods and services becomes more expensive in the domestic economy than in the United States, undermining the competitiveness of the tradable sector.

However, the coefficient of “tropical area” is not significant, although it has the requisite negative sign. This is in line with Rodrik et al. [2002], who find that the effect of geography on incomes per capita in a sample of 140 countries tends to vanish once the quality of institutions is controlled for.

As a robustness check, we employ an alternative measure of the power of N: the ratio of employment in services to total employment in the services and manufacturing sectors. The expected sign is still negative, as relative employment expansion in the services sector, which stands in for the non-tradable sector, would be to the detriment of manufacturing’s own expansion prospects.

Indeed, Table 2 below shows that all the regressors are still significant and have the expected signs, as in the previous model.6 In particular, the coefficient of the alternative definition of the power of N is still negative and significant, albeit much higher in magnitude than that of the previous definition. Moreover, the model passes all the diagnostic checks even with 56 instruments generated using the 2nd to 4th lagged values of the pertinent dependent variables.7

TABLE 2. Two-step system GMM results with alternative definition of the power of N


Variable

Co

effi

cien

t

Sta

ndar

d

erro

r

t-st

atis

tic

p-v

alue

95%

co

nfid

ence

inte

rval

Manufacturing-to-GDP ratio (-1) 0.84 0.03 30.20 0.00 0.79 0.90Alternative power of N -0.13 0.03 -3.78 0.00 -0.20 -0.06Alternative power of N (-1) -0.17 0.03 -5.53 0.00 -0.23 -0.11Alternative power of N (-2) 0.20 0.03 5.79 0.00 0.13 0.26Services growth rate -0.06 0.02 -2.46 0.02 -0.11 -0.01Fixed capital formation as percentage of GDP 0.08 0.01 6.25 0.00 0.06 0.11ICRG index 0.03 0.01 2.87 0.01 0.01 0.04

6 The summary statistics of the regression variables are in Table A3 in the appendix.7 A caveat, however, is that the coefficient of lagged manufacturing-to-GDP ratio does not fall within the interval bounded by the Ordinary Least Squares regression and fixed estimates. The Ordinary Least Squares estimate is 0.78, while the fixed effects estimate is 0.38.



Variable

Co

effi

cien

t

Sta

ndar

d

erro

r

t-st

atis

tic

p-v

alue

95%

co

nfid

ence

inte

rval

Exchange rate (purchasing power parity-adjusted) -0.04 0.01 -2.97 0.01 -0.06 -0.01Average manufacturing tariff -0.03 0.01 -2.55 0.01 -0.05 -0.01Period 3 7.70 2.78 2.77 0.01 2.11 13.28Period 4 9.15 2.71 3.37 0.00 3.69 14.60Period 5 7.75 2.88 2.69 0.01 1.97 13.54Period 6 7.20 2.90 2.49 0.02 1.37 13.02Number of observations: 127Number of groups: 48Number of instruments: 56Diagnostic testsArellano-Bond test for AR(1) in first differences: Prob > z = 0.021Arellano-Bond test for AR(2) in first differences: Prob > z = 0.132Hansen J-test of overid. restrictions: Prob > chi2 = 0.774Difference-in-Hansen tests of exogeneity of instrument subsets:GMM instruments for levelsHansen test excluding group: Prob > chi2 = 0.319Difference (null H = exogenous): Prob > chi2 = 0.943GMM instruments: manuf’g-GDP ratio, alternative power of N, 3rd-4th lagsHansen test excluding group: Prob > chi2 = 0.548Difference (null H = exogenous): Prob > chi2 = 0.909GMM instruments: services growth rate, 3rd-4th lagsHansen test excluding group: Prob > chi2 = 0.573Difference (null H = exogenous): Prob > chi2 = 0.896GMM instruments: lagged ICRG, fixed capital formation (percentage of GDP), average manufacturing tariff, 2nd-3rd lagsHansen test excluding group: Prob > chi2 = 0.573Difference (null H = exogenous): Prob > chi2 = 0.896IV instruments: Period 1, Period 2, Period 3, Period 4, Period 5, Period 6, eq(level)Hansen test excluding group: Prob > chi2 = 0.663Difference (null H = exogenous): Prob > chi2 = 0.906

5. Summary

In this paper, we first define development progeria as a phenomenon where the industrial share dynamics in low-income countries mimic the industrial share dynamics in mature high-income economies. This means that the Non-traded goods sector share is growing faster than the Traded Sector share in poor countries. By contrast, the industrial share dynamics of economies in the catch-up or convergent trajectory has the Traded goods sector share first outstripping the Non-traded sector share before shifting to the share dynamics of the mature high-income economies once a high income level is attained.

Development progeria entails slow overall growth for these economies and a thin chance of a catching up. We argue that the Philippines seems to be so


afflicted in the 25 years after the fall of the Marcos regime. We then inquire into the possible contributors to development progeria. We focus on two factors: the market and institutional distortions in developing countries which, according to Rodrik [2008] pull down the Traded goods sector more than they do the Non-traded goods sector growth and shares; and the impact of the value of the local currency which, when weak, tend to favor the Tradable goods sector and, when strong, the Non-traded goods sector. We use icrg as proxy for the first and the purchasing power parity conversion factor to market exchange rate as proxy of the second.

We show, using the cross-country data from the World Bank Development Index for countries with per capita income us$10,000 or less and using Manufacturing as proxy for the Tradable sector and Services for the Non-traded goods sector, that those two factors after controlling for other factors cannot be rejected as contributors to development progeria. The better the governance environment and the more favorable the exchange rate to market rate, the higher is the share of tradable goods sector share. The variable “power of N” which is the capacity of the Non-tradable sector to absorb labor and the growth of the Non-tradable Sector have each a negative and significant effect on the share of the tradable sector. Likewise, the higher is the investment rate (proxied by the ratio of Gross Fixed Capital Formation to gdp), the higher is the share of T in gdp. That the variable “tropical area” is not significant is consistent with the Rodrik et al. (2002) result that when institutions and policies are introduced, the equatorial paradox loses its bite.

*University of the Philippines School of Economics

References

Arellano, M. and S. Bond [1991] “Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations”, The Review of Economic Studies 58 (April): 277-297.

Blundell, R. and S. Bond [1998] “Initial conditions and moment restrictions in dynamic panel data models”, Journal of Econometrics 87: 115-143.

Fabella, R. [2013] “Development progeria: malady and remedy”, Transactions 35(1).

Fabella, R. [2015] “Development progeria: the algebra”, unpublished working paper. University of the Philippines School of Economics.

Fabella, R. and M.C. Fabella [2012] “Development progeria?”, Chapter 4 in Looking Back, Moving Forward: 25 Years of BusinessWorld. Quezon City: BusinessWorld Publishing Corp.

Rodrik, D. [2008] “The real exchange rate and economic growth”, Brookings


Papers on Economic Activity. Fall 2008. Available at http://www.brookings.edu/~/media/projects/bpea/fall-2008/2008b_bpea_rodrik.pdf.

Rodrik, D., A. Subramanian, and F. Trebbi [2002] “Institutions rule: the primacy of institutions over geography and integration in economic development”, nber working paper 9305.

Roodman, D. [2006] “How to do xtabond2: An introduction to difference and system gmm in Stata”, Center for Global Development working paper 103.

Usui, N. [2011] “Transforming the Philippine economy: ‘walking on two legs’”, adb Economics Working Paper Series No. 252. Available at http://poseidon01.ssrn.com/delivery.php.

Windmeijer, F. [2005] “A finite sample correction for the variance of linear efficient two-step gmm estimators”, Journal of Econometrics 126 (May): 25-51.

APPENDIX: List of countries included in the regressions

51 countries in Table 1: Albania, Armenia, Azerbaijan, Burkina-Faso, Bangladesh, Belarus, Bolivia, Brazil, Botswana, Chile, China, Cameroon, Colombia, Costa Rica, Cuba, Dominican Republic, Ecuador, Egypt, El Salvador, Guatemala, Guyana, Honduras, Indonesia, India, Jamaica, Kazakhstan, Latvia, Madagascar, Malaysia, Mexico, Mongolia, Morocco, Namibia, Nicaragua, Pakistan, Panama, Peru, Senegal, Sierra Leone, South Africa, Sri Lanka, Suriname, Thailand, Trinidad and Tobago, Turkey, Uganda, Uruguay, Venezuela, Vietnam, Zambia, Zimbabwe

48 countries in Table 2: Albania, Armenia, Azerbaijan, Burkina-Faso, Bangladesh, Belarus, Bolivia, Brazil, Botswana, Chile, China, Cameroon, Colombia, Costa Rica, Cuba, Dominican Republic, Ecuador, Egypt, El Salvador, Guatemala, Honduras, Indonesia, India, Jamaica, Jordan, Kazakhstan, Latvia, Madagascar, Malaysia, Mexico, Mongolia, Morocco, Namibia, Nicaragua, Pakistan, Panama, Peru, Paraguay, South Africa, Sri Lanka, Suriname, Thailand, Trinidad and Tobago, Turkey, Uganda, Uruguay, Venezuela, Vietnam


TABLE A1. Summary statistics of regression variables in Table 1

Variable Number of observations

Mean Standard deviation

Minimum Maximum

Manufacturing-to-GDP ratio 119 16.81 6.32 4.85 34.93

Manufacturing-to-GDP ratio (-1) 119 17.69 6.28 6.20 34.93


Power of N (-1) 119 -0.28 3.79 -38.46 9.73

Services growth rate 119 5.20 3.12 -2.03 18.86

Fixed capital formation as percentage of GDP 119 23.00 5.56 10.81 39.37

ICRG index 119 64.99 14.92 -2.83 82.97

Exchange rate (purchasing power parity-adjusted)

119 42.32 14.12 17.17 81.47

Tropical area 119 0.66 0.44 0.00 1.00

Period 1 119 0.00 0.00 0.00 0.00

Period 2 119 0.00 0.00 0.00 0.00

Period 3 119 0.17 0.38 0.00 1.00

Period 4 119 0.23 0.42 0.00 1.00

Period 5 119 0.29 0.45 0.00 1.00

Period 6 119 0.32 0.47 0.00 1.00

TABLE A2. Ordinary Least Squares and fixed effects estimation results

Ordinary Least Squares Fixed effects Coefficient p-value Coefficient p-value

Manufacturing-to-GDP ratio (-1) 1.19 0.00 0.60 0.00Manufacturing-to-GDP ratio (-2) -0.34 0.00 -0.33 0.00Alternative power of N (-1) -0.02 0.70 -0.01 0.80Services growth rate -0.17 0.01 0.11 0.18Fixed capital formation as percentage of GDP (-1) 0.08 0.03 -0.08 0.17ICRG index (-1) 0.02 0.18 -0.10 0.19Exchange rate (purchasing power parity-adjusted) -0.05 0.00 -0.09 0.03Tropical area 0.03 0.95Average manufacturing tariff

Period 1 0.00 0.00Period 2 0.00 0.00Period 3 0.23 0.67 0.33 0.45Period 4 0.86 0.10 0.00Period 5 0.30 0.50 -0.73 0.16Period 6 0.00 -0.62 0.33Constant 1.66 0.29 24.09 0.00R-squared:

Within 0.41 0.62Between 0.97 0.43Overall 0.93 0.43N 119 119


TABLE A3. Summary statistics of regression variables in Table 2

Variable Number of observations

Mean Standard deviation

Minimum Maximum

Manufacturing-to-GDP ratio 127 17.41 6.38 4.85 34.93


Alternative power of N 127 70.28 6.61 52.16 82.56

Alternative power of N (-1) 127 69.06 6.99 47.12 83.87

Alternative power of N (-2) 127 67.91 6.80 44.57 82.14

Services growth rate 127 5.09 3.01 -2.03 18.86

Fixed capital formation as percentage of GDP (-1)

127 22.76 6.19 9.67 45.16

ICRG (-1) 127 65.43 14.51 -2.83 82.97

Exchange rate (purchasing power parity-adjusted)

127 41.74 14.14 17.17 81.47

Average manufacturing tariff 127 9.02 9.33 0.56 86.87

Period 1 127 0.00 0.00 0.00 0.00

Period 2 127 0.00 0.00 0.00 0.00

Period 3 127 0.18 0.39 0.00 1.00

Period 4 127 0.23 0.42 0.00 1.00

Period 5 127 0.31 0.46 0.00 1.00

Period 6 127 0.28 0.45 0.00 1.00


PRE

Malaysia-China trade: old and new routes

Tham Siew Yean* and Andrew Kam Jia Yi*

The opening up of China and its subsequent rapid integration into the global economy provided opportunities for Malaysia to increase trade with China through their respective participation in global value chain activities. This paper aims to examine the nature of Malaysia-China trade and the contributory factors to this bilateral trade pattern and to assess the prospects and challenges for this bilateral trade, with the new Silk Road initiatives.

The main findings of this paper show increased information and communications technology and palm oil trade between the two countries. But the trade pattern also indicates Malaysia’s trade in electronics and palm oil with China is facing either stagnating or declining revealed comparative advantage. Instead, China’s rise has also raised severe competitive pressures as each country tries to shift up the global value chain. Internal weaknesses also contributed to the weakening competitiveness of Malaysia’s electrical and electronics sector so that the share of this sector’s manufacturing and exports fell substantially, while the share of resource based manufacturing activities gained ground.

Malaysia is trying to tap the promises of the new Silk Road initiative for enhancing its trade and investment relations with China. However, realizing this potential faces several hurdles, such as the viability of this initiative and internal challenges within China and Malaysia. The most urgent of the three challenges is for Malaysia to overcome its domestic challenges, since without the requisite domestic changes, no investor, including China, will venture into the country.

JEL classification: F10, F14, N75 Keywords: Malaysia, China, trade, Silk Road, electronics, palm oil


1. Introduction

Malaysia’s strategic geographic location in Southeast Asia has facilitated its trade between the East and West since the time of the Malacca Sultanate in the 14th and 15th century a.d. Malacca’s importance as a center for entrepôt trade along this old trade route led to close relations between the Malacca Sultanate and several countries, including China. This in turn paved the way for the establishment of Malacca-China diplomatic relations and the arrival of traders from China to Malacca. The subsequent waves of migration of Chinese workers to Malaysia and a growing Chinese population led to a significant share of Chinese people in Malaysia’s total population (38 percent)1 in 1957.

Although bilateral relations between Malaysia and China waxed and waned over time, Malaysia became the first member country of the Association of Southeast Asian Nations (asean) to normalize relationships with China in 1974, thereby laying the foundation for cordial bilateral relations, including trade relations, between the two countries. Not surprisingly, Malaysia’s first free trade agreement is the asean-China Free Trade Agreement, which was signed in 2002. The growing importance of China as a trade partner of Malaysia is manifested in the upward trend of exports and imports between the two countries since 1990, with the exception of a dip during the years of the 2008-2009 global financial crisis. In fact, China became Malaysia’s largest trading partner since 2010. Within asean, Malaysia has been China’s largest trading partner since 2008. It is also China’s third-largest trading partner in Asia, after Japan and South Korea, despite the relatively small size of Malaysia’s economy. Both sides have pledged to aim for a trade volume of us$160 billion by 2017.2

Since China’s reforms in 1978 and the opening up of its economy in the 1990s, China has grown to become the second-largest economy in the world and has joined the upper middle income economies in 2012. Demand from China, including the demand from its rising middle-income group, has and will undoubtedly continue to influence global and regional trade. China’s increasing economic size and grand plans, as manifested in the new Silk Road or the “One Belt, One Road” development strategy and the Asian Infrastructure Investment Bank in 2014, and its trade ambition may offer a window of opportunity for Malaysia to deepen its economic relations with China in the coming years. This is especially important at this juncture of Malaysia’s economic development as domestic growth is faltering in the midst of restructuring problems, political unrest, and global uncertainties.

The objectives of this paper are twofold. First, we seek to examine the nature of Malaysia-China trade and the contributory factors to this bilateral trade pattern.

1 This figure is an estimate by Hirschman [1980]. 2 This figure is an estimate by Sta Maria [2014].

102 Tham Siew and Kam: Malaysia-China trade: old and new routes

Second, we assess the prospects and challenges for Malaysia-China trade, given the new Silk Road initiatives. The paper is organized as follows. In the second section, we examine two important sectors in the bilateral trade between Malaysia and China. In the third section, we look into the prospects and challenges of the new routes. In the final section, we summarize the main findings of this paper.

2. Two major sectors

Despite the importance of Malaysia in China’s trade, the bilateral relations between these two countries are not well explored, except as part of asean’s trade with China (see, for example, Mendoza et al. [2015]). The only recent study that examined the bilateral trade relations is that of Lee [2014] which examined it from China’s perspectives, namely within the context China’s economic engagement with Southeast Asia. Lee’s focus is on the electrical and electronics sector where he finds increasing competition between the two countries due to weaknesses in Malaysia’s economy and China’s move up the manufacturing value chain.

While the bilateral trade relation is very much focused on electrical and electronics, palm oil constitutes another important export from Malaysia to China. China is Malaysia’s largest export destination for palm oil products, taking up 20.4 percent of Malaysia’s global exports. It is therefore pertinent to include this sector in analyzing the trade links between Malaysia and China.

The paper will focus on information and communications technology (ict) and palm oil trade between Malaysia and China, given the importance of these two sectors in the bilateral trade as well as Malaysia’s policy focus in these two sectors.3

2.1. Information and communications technology

Within manufacturing, the ict4 sector stands out as the main contributor to the bilateral trade between the two countries. Table 1 shows that ict goods (on average) contribute approximately half of Malaysia’s total manufacturing exports and import to/from China from 2000 to 2013. The bulk of Malaysia’s exports of ict goods are in electronic components. In 2013, electronic components comprise 48 percent of total manufacturing exports to China. The second-largest share in ict exports is contributed by computer and peripherals, averaging at around 19 percent for the period shown. However, its export share has fallen by more than half from a peak of 28 percent in 2001 to 11 percent in 2013. Imports from China are mainly in computer and peripherals (average of 21 percent for the period shown in Table 1) while imports of electronics component contribute

3 See Tham and Kam [2015] for a summary on the information communications technology policies in Malaysia as a driver for the economy since the Seventh Malaysia Plan (1996-2000) onwards.

4 ict product classification is based on oecd [2011].


an average of 14 percent for the period shown. By 2013, the share of imports of electronics components and computer and peripherals are 20 percent and 10 percent, respectively.

Table 1 also indicates a diverging pattern between the import and export of Malaysia’s ict goods with China. The share of ict imports in total manufacturing imports can be seen to decrease steadily from 2003 to 2013, excluding the year of the 2009 global financial crisis. On the other hand, exports show a converse pattern with the share of ict exports to China falling before the global financial crisis from 2001 to 2007 and increasing after the global financial crisis to 67 percent in 2010 before falling slightly annually to 61 percent in 2013.

The share of electronics exports fell from 19 percent in 2000 to 12 percent in 2003, after which it fluctuated from 11 percent in 2004 to 14 percent in 2007. However, its share increased steadily from 37 percent in 2010 to 48 percent in 2013. This is accompanied by an increasing share of electronics components in total manufacturing imports from 2010 to 2013. Notably, although Malaysia has an overall trade deficit in manufacturing exports with China, Table 1 shows a surplus in Malaysia’s trade in overall ict and in the electronics sub-sector after 2008.

TABLE 1. Information communications technology (ict) import and export of Malaysia to China, 2000-2013 (share of manufacturing export, percentage)

Exports Imports

Co

mm

unic

atio

n eq

uip

men

t

Co

mp

uter

and

p

erip

hera

ls

Co

nsum

er e

lect

roni

cs

Ele

ctro

nics

co

mp

one

nts

Mis

cella

neo

us

Tota

l IC

T a

s sh

are

of

man

ufac

turi

ng e

xpo

rt

Trad

e b

alan

ce, o

vera

ll IC

T (U

S$

bill

ions

) (X

-M)

Co

mm

unic

atio

n eq

uip

men

t

Co

mp

uter

and

p

erip

hera

ls

Co

nsum

er e

lect

roni

cs

Ele

ctro

nics

co

mp

one

nts

Mis

cella

neo

us

Tota

l IC

T a

s sh

are

of

man

ufac

turi

ng im

po

rt

Trad

e b

alan

ce,

elec

tro

nics

(US

$ b

illio

ns) (

X-M

)

2000 7.0 22.0 2.0 19.0 5.0 55.0 0.03 10.0 7.0 7.0 10.0 7.0 41.0 -0.022001 3.0 28.0 1.0 16.0 4.0 54.0 -0.13 12.0 15.0 6.0 14.0 5.0 52.0 -0.132002 1.0 21.0 1.0 18.0 2.0 44.0 -1.25 4.0 30.0 4.0 14.0 4.0 55.0 -0.202003 0.0 15.0 1.0 12.0 3.0 32.0 -2.18 2.0 34.0 4.0 13.0 4.0 57.0 -0.432004 1.0 17.0 1.0 11.0 4.0 34.0 -3.37 3.0 30.0 3.0 14.0 6.0 56.0 -0.912005 2.0 16.0 1.0 14.0 3.0 35.0 -4.30 3.0 30.0 2.0 16.0 5.0 56.0 -1.202006 2.0 18.0 1.0 11.0 2.0 34.0 -4.40 3.0 27.0 2.0 14.0 3.0 50.0 -1.362007 2.0 19.0 1.0 13.0 1.0 36.0 -3.96 2.0 25.0 2.0 13.0 3.0 46.0 -1.092008 3.0 29.0 1.0 12.0 1.0 46.0 -3.24 3.0 23.0 2.0 10.0 3.0 41.0 -0.992009 1.0 16.0 1.0 52.0 1.0 71.0 2.06 5.0 24.0 2.0 16.0 3.0 50.0 4.602010 1.0 26.0 1.0 37.0 1.0 67.0 4.46 5.0 15.0 2.0 12.0 6.0 41.0 4.182011 1.0 18.0 1.0 44.0 1.0 65.0 4.60 5.0 12.0 1.0 12.0 6.0 37.0 5.782012 1.0 15.0 1.0 44.0 1.0 62.0 2.01 5.0 10.0 1.0 18.0 4.0 38.0 3.802013 1.0 11.0 1.0 48.0 0.0 61.0 0.51 5.0 10.0 1.0 20.0 4.0 39.0 3.38Average 2.0 19.0 1.0 25.0 2.0 50.0 5.0 21.0 3.0 14.0 4.0 47.0

Source: Authors’ calculations using OECD 2011 classifications. See Appendix 1.


The concentration of electronics exports to China can be attributed to China’s rapid integration into the global value chain in this sector since opening its economy to foreign direct investment in the 1990s (Lee [2014]; Kong et al. [2015]). Both Malaysia and China used foreign direct investment policies to attract foreign multinational companies in the electrical and electronics sector for the development of their respective manufacturing sector (Xing [2013]; Tham and Kam [2015]) and are linked to each other’s production through the global value chains in this sector.

It can be seen that Malaysia’s exports of electronics goods is more concentrated in parts and components, rather than finished goods (Figure 1). However, although exports of parts and components increased rapidly before the global financial crisis, its increase has slowed down after the crisis. For example, in the four years before the global financial crisis, exports of parts and components to China grew by 156.6 percent but subsequently grew by only 18 percent from 2009 to 2013, due in part to the global slowdown after the global financial crisis.

Note: Parts and components and finished goods are aggregated based on Aldaba [2015].Source: Calculated from COMTRADE

FIGURE 1. Malaysia’s electronics exports to China by parts and components and finished goods, us$ billions

Bilateral revealed comparative advantage5 in Figure 2 shows Malaysia having a comparative disadvantage in overall electronics exports to China prior to 2005. After 2005, the revealed comparative advantage for overall electronics increased steadily to a value that is above one, barring the year of the global financial crisis.

5 Bilateral rca = (Exports of product Xi from Malaysia to China/All manufacturing exports of Malaysia to China) / (All exports of Xi from Malaysia/All manufacturing exports of Malaysia). A value above one represents comparative advantage of product Xi in China’s manufacturing sector.

12

10

8

6

4

2

0

Parts and components

Finished goods

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013


The revealed comparative advantage for all electronics components is above one, while the converse holds true for finished goods. This suggests that the electronics sector in Malaysia may have undergone some form of industrial upgrading in order to compete in the Chinese manufacturing market.

The emergence of wafer fabrication and r&d centers is an indicator of functional upgrading in a country. According to Shan and Rasiah [2015], foreign multinational companies have established four chip design centers, one r&d support facility, five wafer fabrication plants, and 28 assembly and test plants, while national firms have two wafer fabrication plants and seven assembly and test plants (Table 2). They attribute the “shift to wafer fabrication” to the investment grants offered to foreign firms. Therefore, a few firms such as Infineon, osram, and on Semiconductor have embarked on wafer fabrication in Malaysia since 2005. China is also rapidly upgrading, as shown in Table 2. The race to move up the global value chain in both countries has led to a slight stagnation in the revealed comparative advantage for parts and components electronics observed after the global financial crisis (Figure 2).

Source: Calculated from COMTRADE

FIGURE 2. Revealed comparative advantage for electronics exports to China, 2000-2013

2000

1.8

1.6

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0

2001

2002

2003

2004

2005

2006

2007

2008

2009

2012

2013

2010

2011

Overall electronics

All parts and components electronics

All finished electronics


TABLE 2. Integrated circuit firms facing upgrading in Malaysia, China, and Vietnam, 2011

Countries

National companies Foreign companies

Total

R&

D

Chi

p d

esig

n

Sup

po

rt R

&D

Waf

er

fab

rica

tio

n

Ass

emb

ly a

nd

test

R&

D

Chi

p d

esig

n

Sup

po

rt R

&D

Waf

er

fab

rica

tio

n

Ass

emb

ly a

nd

test

China 1 3* 2# 25 34 0 11 8 6 58 148*#

Malaysia 0 0 0 2 6 0 4 1 5 25 43

Vietnam 0 1 0 1 0 0 3 1 0 2 8

Notes: Firms are defined by the registration status, and hence, some subsidiaries of the same firm are counted more than once.* Denotes firms having both R&D and chip design in the same registration premises;# Denotes firms having both R&D and other supportive R&D in the same premises.Source: Shan and Rasiah [2015]

2.2. Palm oil

There are two main sub-sectors dominating palm oil exports to China, reports the Malaysian Palm Oil Board: palm oil, palm oil kernel, and finished products; and oleochemicals and biodiesel. In Figure 3a, it can be seen that the largest contributor to exports is palm oil, palm oil kernel, and finished products. Figure 3a also shows that exports of both products have increased substantially since 2000. This is essentially contributed by the increasing demand from China, which does not produce any palm oil. The nation uses the oil in its food processing industry, especially in the production of instant noodles, as palm oil is relatively cheaper than soybean oil and rapeseed oil. Figure 3b shows the trend in the exports of palm oil products—excluding palm oil, palm oil kernel, and finished products—and oleochemicals and biodiesel due to the significant difference in value for the latter two products. Soaps and surfactants show increasing trade value since 2008, as shown in Figure 3b.

In terms of imports, oleochemicals and biodiesel comprise the largest import value from China (Figure 3c). Figure 3c further shows that oleochemical and biodiesel imports have increased sharply since 2012, amounting to over US$770 million in 2014. Other palm oil imports from China are also increasing gradually over the years as well (Figure 3d).

One important distinction between manufacturing and the palm oil exports is the trade balance. While Malaysia has a trade deficit in the manufacturing sector, trade in palm oil products has always been at a surplus since 2000. Figure 4a shows the surplus from Malaysia’s palm oil trade with China increasing steadily until 2011, after which the surplus narrowed slightly. ict and electronics trade also show trade surpluses only from 2008 to 2013. Figure 4b shows that all


subsectors of palm oil have generated trade surpluses for Malaysia since 2000. This surplus is led by palm oil, palm oil kernel, and finished products, followed by oleochemicals and biodiesel.

Source: Computed from COMTRADE

FIGURE 3. Palm oil trade between Malaysia and China, US$ millions

c) Imports from China, US$ millions

900

2000

2002

2004

2006

2008

2010

2012

2014

Palm kernel cakePalm oil, palm oil kernel, and finished products

Soaps and surfactantOleochemicals and biodiesel

800

700

600

500

400

300

200

100

0

d) Imports from China, US$ millions20

2000

2002

2004

2006

2008

2010

2012

2014

Palm kernel cake

Palm oil, palm oil kernel, and finished products

Other palm products

Soaps and surfactant

18

16

1412

10

8

6

4

0

2

2001

2003

2005

2007

2009

2011

2013

b) Exports to China, US$ millions120

100

80

60

40

20

0

2000

2002

2004

2006

2008

2010

2012

2014

2001

2003

2005

2007

2009

2011

2013

Palm kernel cake

Other palm products


a) Exports to China, US$ millions5000

4500

3500

4000

2500

3000

1500

2000

1000

0

500

2000

2002

2004

2006

2008

2010

2012

2014

Palm kernel cakePalm oil, palm oil kernel, and finished products

Soaps and surfactantOleochemicals and biodiesel


Note: * “Electronics” is based on consumer electronic equipment and electronic components in Appendix 1.Source: Computed from COMTRADE

FIGURE 4. Trade surpluses/deficits, US$ millions

However, Figure 5 provides some cautionary notes on the performance of Malaysia’s palm oil exports to China. Since 2000, the revealed comparative advantage for overall palm oil has been declining, where in 2013, it is considered as a product with marginal comparative advantage in China (slightly above one). The declining trend is striking as it holds true for all the product groups—except for palm kernel cake, which does not show a comparative advantage for the period shown. Even though the main palm oil products (palm oil, palm oil kernel, and

a) Trade surplus across industries, US$ millions

8000

2000

2002

2004

2006

2008

2010

2012

ICT

Electronics

Manufacturing

Palm oil

6000

4000

2000

0

-2000

-4000

-6000

-8000

-12000

-10000

2001

2003

2005

2007

2009

2011

2013

b) Trade surplus by palm oil products, US$ millions

5000

4500

3500

4000

2500

3000

1500

2000

1000

0

500

Palm kernel cake



Oleochemicals and biodiesel

Other palm products20

0020

0220

0420

0620

0820

1020

1220

1420

0120

0320

0520

0720

0920

1120

13


finished products) still have a comparative advantage in the Chinese market, the second most traded product (oleochemicals and biodiesel) has lost its advantage since 2005. Other palm products have lost their comparative advantage since 2007, while soaps and surfactants became comparatively disadvantaged in 2013.

Source: Computed from COMTRADE

FIGURE 5. Revealed comparative advantage for palm oil products, 2000-2013

The drop in comparative advantage for Malaysia’s palm oil exports may be due to excess supply as the rise in global demand and high commodity price had encouraged production to escalate. The average annual growth in palm oil production from 1993-2012 was 7.3 percent; the same figure for palm kernel oil production in the same period was 6.3 percent [Basiron 2013]. Total world supply outstripped world demand for every year shown in Table 3, except for 2009. In 2013, for example, the growth of palm oil production in both Indonesia and Malaysia in 2013 surpassed palm oil demand and resulted in a high inventory and a price decline.

TABLE 3. Palm oil supply and demand, 2005-2012 (million tonnes)

2005 2006 2007 2008 2009 2010 2011 2012

Supply 33.4 36.0 37.7 42.6 45.3 45.9 50.1 51.2

Demand 32.7 35.4 37.4 41.3 45.4 46.5 48.7 51.0

Source: European Delegation to Malaysia, 2012

4.00

3.50

2.50

3.00

1.50

2.00

0.50

1.00

0.00

Palm kernel cake



Oleochemicals and biodiesel

2000

2002

2004

2006

2008

2010

2012

2001

2003

2005

2007

2009

2011

2013

Overall

Other palm products


In addition, there was also increasing competition from Indonesia as it changed its export duty structure to favor production and export of processed palm oil [European Delegation to Malaysia 2012]. Hence its export duty on crude palm oil increased while the export duty for processed palm oil products went down from 25 percent to 13 percent. This posed considerable competition to Malaysia as 68.8 percent of its export revenue is from the processed product segment.

3. Prospects and challenges

The new Silk Road was announced by the Chinese Premier in 2013 during a visit to Kazakhstan. This was subsequently renamed “One Belt, One Road” as it encompasses two routes: a land trade route linking China with Central Asia, Russia, and Europe; and a second route or the Maritime Silk Route, stretching from Fujian on China’s coast, through the Malacca Straits, around the horn of Africa, and ending in Venice. This second route is of particular relevance to the maritime countries in Southeast Asia, including Malaysia.

The “One Belt, One Road” initiative promises to enhance trade links along the area covered by this initiative. This is estimated to include about 50 percent of the world’s gdp and approximately the same for global trade [Bloomberg Brief 2015]. The use of the Asian Infrastructure Investment Bank, also proposed by China, to fund the infrastructure needs along these routes envisages a reduction in transportation costs that should stimulate more trade. This is backed by a us$40-billion Silk Road Fund and a promise of us$50 billion for the Asian Infrastructure Investment Bank and the proposed new brics New Development Bank by the Chinese government.

Prior to the announcement of the “One Belt, One Road” initiative, the Malaysian government has been actively courting closer economic cooperation with China. In 2012, the China-Malaysia Qinzhou Industrial Park was established. Then followed the establishment of the Malaysia-China Kuantan Industrial Park in 2013, which started operations in the same year. Sta Maria, the Secretary General of the Ministry of International Trade and Industry in her keynote address at the Symposium on the 40th Anniversary of China Malaysia Diplomatic Relations on May 22, 2014, noted that this is the first time a twin park model is introduced anywhere in the world, serving complementary roles with the same principles. The Chinese Central Government has approved a rmb2.4 billion development fund for Qinzhou Industrial Park, over and above the rmb1 billion pledged by the Government of Guangxi, where Qinzhou Industrial Park is located. In turn, the Malaysian Federal and State Government have allocated rm700 million to Malaysia-China Kuantan Industrial Park [Sta Maria 2014]. While Qinzhou Industrial Park will focus on food processing, biotechnology, and information technology, Malaysia-China Kuantan Industrial Park will host businesses engaged in steel manufacturing, aluminium processing, and palm oil refinery. This initiative is in line with Malaysia’s ambitions to use this park


to increase Chinese investment in Malaysia [Khor 2013]. China has also been upgrading the Kuantan port with Beibu Gulf Holding (Hong Kong) Co. Ltd given a 38 percent equity share in a consortium that received a 30-year concession to manage, operate and develop the port as part of its move to secure more seaports [Pitlo III 2015].

Port Klang and Singapore’s port are key ports in the Maritime Silk Route since the Straits of Malacca is a strategic passageway from China to the West. Port Klang is reportedly moving towards establishing sister-port relationships with seven ports in China—namely, Shanghai, Ningbo, Tianjin, Guangzhou, Xiamen, Dalian, and Fuzho—in order to enhance its status in the new Silk Road.

Malaysia also signed the Five Year Programme for Economic and Trade Cooperation with China in October 2013. The agreement aims to enhance bilateral cooperation in agriculture, energy, and mineral resources, information and telecommunications, manufacturing, infrastructure and engineering, tourism, logistics, and retailing. This program aims to engage the private sector to stimulate investment and trade, thereby signaling again Malaysia’s intentions to attract more Chinese investment into the country.

Economic cooperation has recently expanded into the financial sector. Last year, the two countries’ central banks agreed to establish a yuan clearing bank in Kuala Lumpur [Chew 2014]. This is part of China’s quest to internationalize the renminbi, and Malaysia is the second country in Southeast Asia, after Singapore, to host a yuan clearing bank.

Apart from these bilateral initiatives, Malaysia is also part of the asean-China Free Trade Agreement as well as the ongoing negotiations for a Regional Comprehensive Economic Partnership that includes the free trade agreement partners of asean, including China.

All these initiatives indicate Malaysia’s strategic interests to tap China’s ambitious plans for catalyzing its own economic performance. This is especially important for Malaysia in view of the sharp fall in private investment after the Asian financial crisis and the moderation in its growth performance. Nevertheless, the prospects of tapping the economic potential of the “One Belt, One Road” initiative face numerous challenges, including domestic challenges within China and Malaysia. Ghosh [2015] points out the need for cooperation and partnership of the countries along the route for the maritime silk route to succeed, which is not likely to be forthcoming since opaqueness of the plan in general tends to breed suspicion rather than trust. More importantly, it remains to be seen as to how the proposed route can coexist with the existing system of maritime cooperation.

3.1. China shifts to slower growth

China has been growing at an average rate of 10 percent per annum for the last three decades. Clearly, a slowdown in growth, estimated to be seven percent


for 2015, is inevitable. Since the rapid growth in the past has generated excess production capacity, dissipated its low-wage advantage, increased resource constraints, and led to considerable environmental damage, China will have to manage an internal transition to a slower—but more sustainable growth rate. Specifically, China needs to rebalance its economy from export-led to domestic consumption led. This will shift China from being an export platform to a market provider so that its trade surplus will also decrease. Although the most recent International Monetary Fund review indicates an increasing role in domestic consumption driving growth in China and a reduction in the current account surplus, it remains to be seen if this rebalancing can be sustained in the longer term [imf 2015].

Given that past overinvestment has contributed to the rising demand of China for commodities and the subsequent commodity price boom over the last decade, the shift to lower growth and reduced demand for commodities is expected to have significant spillovers on primary commodity-producing economies like Malaysia. Palm oil prices are trending downwards, and this can affect Malaysia’s future trade surplus, which has also been trending downwards after the global financial crisis.

3.2. Malaysia restructures its economy

Economic restructuring is the key economic problem in Malaysia. Within manufacturing, resource-based industries have been the largest growth drivers of this sector from 2002-2012 due to the high commodity prices during this period [Rizwan et al. 2014]. The sector grew at 6.2 percent on a compounded annualized growth rate during this period, compared to the 1.5 percent growth in the electrical and electronics sector. It became the largest manufacturing sub-sector after 2005. Likewise, exports of resource-based industries grew at 10.3 percent on a compounded annualized growth rate basis during the same period. Its share in total manufacturing exports grew from 30 percent in 2002 to 51 percent in 2012, while exports of electrical and electronics fell from 70 percent to 49 percent over the same period.

Both electrical and electronics and palm oil industries face multiple challenges in their respective structural transformation. Deepening the electrical and electronics sector from low value added to high valued-added activities has been emphasized since the Second Industrial Master Plan (1996-2005). The plan advocated the use of a cluster concept for propelling manufacturing development forward to achieve economies of agglomeration. Unfortunately, the plan was deemed to be poorly executed [Danaraj 2011]. Since then, there are further efforts to shift the electrical and electronics sector towards higher value-added activities, but as noted in Tham [2015], a critical mass of local firms in the electronics sector has yet to be achieved (see Table 2). Although small, incremental changes have taken place within this sector, including the emergence of a few local design


houses as indicated earlier, they lack the critical mass to be catalytic in the desired shift. Thus, Malaysia has yet to make critical leap up the value chain, as observed in the Taiwan and South Korea’s respective electrical and electronics sector. Instead, Malaysia’s electrical and electronics sector continues to be dogged with a shortage of talents that is exacerbated by brain drain and inflows of less skilled workers. Weak industry-university linkages and inadequate r&d spending of 1.13 percent of its gdp, compared to China’s 1.98 percent of its gdp, in 2012 are other contributory factors to the slow deepening of the electrical and electronics sector.

The palm oil sector is currently more concentrated in upstream and midstream activities rather than downstream. Similarly, palm oil export is highly biased toward the upstream and midstream segments, with a total contribution to the overall industry at 74 percent, while downstream contribution came in low at 17 percent in 2013 [Khoo 2014]. The industry faces multi-faceted challenges such as limited land, stagnating yields, dependency on foreign labor, trade wars from producers of other oils, increasing competition from Indonesia, and environmental concerns. Therefore, upstream activities need to enhance their productivity. The government is pushing for a shift to the more lucrative downstream segment, but this will require more private investment that is holding off largely due to the negative perceptions associated with numerous unresolved financial scandals that have emerged of late. In addition, there are ongoing leadership struggles within parties and nationally, while public acceptance of a one-party dominant political system continues to weaken [Alagappa 2015]. Fundamentally, there is a critical need for the country to politically shift from the “well-trodden but increasingly contested bases of race and religion” [Algappa 2015:72] discourse that is emphasized by the dominant party of the country.

While the government is also pushing for a shift towards services, exports in the country are still dependent on manufacturing. A shift to services will also face the same talent shortage challenges as in the manufacturing sector, if knowledge-intensive services are the focus. It will also face the same lack of investor confidence due to the ongoing political deadlock. Economic restructuring will be further derailed if the current political challenges are not surmounted in the near future.

4. Conclusion

The opening up of China and its subsequent rapid integration into the global economy provided opportunities for other emerging countries to have increased trade with the nation through their respective participation in global value chain activities. This is best reflected in Malaysia’s ict trade with China. Similarly, China’s rise increased the demand for natural resources like palm oil, thereby contributing to the high commodity prices in the last decade and the growing importance of palm oil trade between Malaysia and China. But the bilateral trade patterns shown in this paper indicate that Malaysia is facing stagnating or


declining revealed comparative advantage in its exports of electronics and palm oil to China. This does not augur well for the future of Malaysia-China trade relations, from Malaysia’s perspective.

Instead, China’s rise also poses severe challenges to Malaysia through competitive pressures as each country tries to shift up the global value chain. Internal weaknesses also contributed to the weakening competitiveness of Malaysia’s electrical and electronics sector so that the share of this sector in manufacturing and exports fell substantially, while the share of resource based manufacturing activities gained ground.

Malaysia is trying to tap the promises of the “One Belt, One Road” initiative for enhancing its trade and investment relations with China. Realizing this potential faces several hurdles, including the viability of this initiative and internal challenges within China and Malaysia. The most urgent of the three challenges for Malaysia is to overcome its domestic challenges; without the requisite domestic changes, no investor, whether domestic or foreign, will venture into the country. Overcoming domestic economic challenges, however, will require Malaysia’s current political development to move beyond its racial and religious paradigms towards a more inclusive paradigm that embraces multicultural diversity and religious tolerance.

*Institute of Malaysian and International Studies, Universiti Kebangsaan Malaysia

The authors thank Professor Hal Hill for his many years of mentoring that has helped each of them to grow as professional economists. They hope Prof. Hill will enjoy reading this paper on one of his pet issues.

References

Alagappa, M. [2015] “Asia’s next great challenge: to balance growth and political development”, Global Asia 10(3), Fall: 68-78.

Aldaba, R.M. [2015] “The Philippines in the electronics global value chain,” paper presented at the Second eia Workshop on “Leveling up Southeast Asia’s value added”, April 24 to 26, 2015, Okura Hotel, Japan.

Basiron, Y. [2013] “Malaysian supply and demand of palm oil: challenges and opportunities until 2020”, http://www.mpoc.org.my/upload/Oil-World-Outlook-Conference-2013-Hamburg.pdf. Accessed 15 October 2015.

Bloomberg Brief [2015] “One Belt, One Road: assessing the economic impact of China’s new Silk Road”, http://www.bloombergbriefs.com/content/uploads/sites/2/2015/07/SC_062615-obor.pdf. Accessed 15 October 2015.

Chew, Elffie [2014] “Malaysia’s central bank works with pboc to start yuan clearing”, http://www.bloomberg.com/news/articles/2014-11-10/malaysia-s-


central-bank-works-with-pboc-to-start-yuan-clearing. Accessed 15 October 2015.

Danaraj, N. [2011] “Technology policy: a critical appraisal”, in Malaysia: policies & issues in economic development. Kuala Lumpur: Institute of Strategic and International Studies.

European Delegation to Malaysia [2012] “Malaysia – the Malaysian palm oil sector overview”, http://www.ice.gov.it/paesi/asia/malaysia/upload/173/Palm%20Oil_overview_2012.pdf. Accessed 15 October 2015.

Ghosh, P.K. [2015] “Daunting realities: territorial disputes and shipping challenges to China’s maritime Silk Road”, Global Asia 10(3), Fall: 48-52.

International Monetary Fund (imf) [2015] People’s Republic of China 2015 Article IV consultation. imf Country Report No. 15/234. https://www.imf.org/external/pubs/ft/scr/2015/cr15234.pdf. Accessed 15 October 2015.

Khoo, D. [2014] “Palm oil industry needs to invest more in downstream activities”, http://www.thestar.com.my/Business/Business-News/2014/06/21/Call-to-venture-downstream-Palm-oil-industry-needs-to-invest-more-in-downstream-activities-to-reduc/?style=biz. Accessed 15 October 2015.

Khor, Y.L. [2013] “The significance of China-Malaysia industrial parks”, Institute of South East Asian Studies Perspective No. 317. Singapore: iseas.

Kong, X.X., M. Zhang, and S.C. Ramu [2015] “China’s semiconductor industry in global value chains”, Economic Research Institute for asean and East Asia (ERIA) Discussion Paper No. 2015-15. Jakarta.

Lee, J. [2014] “China’s economic engagement with Southeast Asia: Malaysia”, Trends in Southeast Asia, No.1. iseas: Singapore.

Hirschman, C. [1980] “Demographic trends in peninsular Malaysia: 1947-1975”, Population and Development Review 6(1): 103-125.

Malaysian Palm Oil Board (mpob) [2013] Revision of customs tariff codes for oil palm products & schedule of tariff rates. Kuala Lumpur.

Mendoza, R., K. Chua, and M. Melchor [2015] “An analysis of the evolving asean-China trade linkages”, Rizalino S. Navarro Policy Center for Competitiveness Working Paper 15-002. Manila: Asian Institute of Management Policy Center.

Organisation for Economic Co-operation and Development (oecd) [2011] oecd guide to measuring the information society. Paris.

Pitlo III, L.B. [2015] “asean’s connectivity and China’s ‘One Belt, One Road’”, http://thediplomat.com/2015/03/asean-connectivity-and-chinas-one-belt-one-road/. Accessed 15 October 2015.

Rizwan, M.H.R., K.T.F. Wong, and R.S.L. Cho [2014] “Agglomeration in practice: the Malaysian experience in diversifying manufacturing”, http://www.bnm.gov.my/documents/conference_vol/2014_Econs_Research/Paper_6_The_Diversification_of_Manufacturing_in_Malaysia.pdf. Accessed 15 October 2015.


Shan, X.S. and R. Rasiah [2015] “Sticky spots on slippery slopes: the development of the integrated circuits industry in emerging East Asia”, http://www.globelics.org/publication/sticky-spots-on-slippery-slopes-the-development-of-the-integrated-circuits-industry-in-emerging-east-asia/. Accessed 15 October 2015.

Sta Maria, R. [2014] “Looking into the future of China-Malaysia relationship”, http://my.china-embassy.org/eng/sbgx/t1158491.htm. Accessed 15 October 2015.

Tham, S.Y. and A.J.Y. Kam [2015] “Trade in value-added: case of Malaysia”, www.eria.org/eria-dp-2015-58.pdf. Accessed 15 October 2015.

Tham, S.Y. [2015] “Diversification and industrial policies in Malaysia”, Development and modern industrial policy in practice: issues and country experiences. Cheltenham: Asian Development Bank and Edward Elgar. Chapter 12 in J. Felipe ed..

The Star [2015] “Malaysia-China in sister port pact”, http://www.thestar.com.my/News/Nation/2015/08/01/Msia-and-China-in-sister-port-pact-pka-chief-Tie-up-will-boost-trade-and-tourism/. Accessed 15 October 2015.

Xing, Y. [2013] “Rethinking the success of China’s high-tech exports”, in W. Gungwu and Z. Yongnian, eds., China: development and governance. Singapore: World Scientific. Chapter 23.


APPENDIX 1. Information communications technology codes (HS 1996)

Computers and peripheral

equipment

Communication equipment

Miscellaneous Consumer electronic equipment

Electronic components

844351 851711 852390 851810 852190 852330

847050 851719 852410 851821 852210 852460

847110 851730 852491 851822 852290 853400

847130 851750 852499 851829 852530 854011

847141 851780 852910 851830 852540 854012

847149 851790 852990 851840 852712 854020

847150 852510 854381 851850 852713 854040

847160 852520 901320 851890 852719 854050

847170 852790 851910 852721 854060

847180 853110 851921 852729 854071

847190 851929 852731 854072

847220 851931 852732 854079

847290 851939 852739 854081

847330 851940 852812 854089

847350 851992 852813 854091

851721 851993 852821 854099

851722 851999 852822 854110

900911 852010 852830 854121

900912 852020 950410 854129

852032 854130

852033 854140

852039 854150

852090 854160

852110 854190

852190 854212

852210 854213

854214

854219

854230

854240

854290

854890

Source: Organisation for Economic Co-operation and Development, 2011


PRE

Class participation in politics in Southeast Asia

Joseph J. Capuno

Using the dataset from the Asian Barometer Survey conducted in 2010-2012, we determine the variations in political participation across socioeconomic classes in five Southeast Asian countries and apply Oaxaca decomposition method to explain the variations. In general, we find high rates of voting participation across classes in Indonesia, Malaysia, the Philippines, and Thailand. In Singapore, barely half of the sample voted in previous elections. In Indonesia, Thailand, and Malaysia, the middle classes distinguish themselves from the rest by participating in other political activities, including contacting officials or the news media, joining others to voice out or directly address their common concerns, and attending rallies or demonstrations. Furthermore, interclass differences in political participation is due more to divergence in mean characteristics in Thailand, but they are due more to the heterogeneous effects of these characteristics in Indonesia and Malaysia. Relative to the middle classes in these three countries, those in the Philippines and Singapore each appears politically disengaged.

JEL classification: D72, Z10, 057Keywords: political participation, middle class, Oaxaca decomposition, Southeast Asia

1. Introduction

Are the middle classes more politically active or engaged than other socioeconomic classes in democratic Southeast Asian countries? There is theoretical basis to expect them to be so, which some casual observations seem to confirm. Aristotle first advanced the idea that the middle class is crucially linked with democracy, i.e., a large prosperous middle class mediate between the rich and the poor, and thereby creates the structural foundation for democratic processes [Glassman 1995]. Building on this idea, Lipset [1959] hypothesized a country’s


democratization—more evidently, perhaps, among the middle classes—as it develops. Also, Moore [1967] famously claimed that there can be no democracy without the middle class. Enunciating the casual link between middle class and democratization, Acemoglu and Robinson [2006] posit that “… a strong and large middle class may aid democratization because it is less in favor of radical policies than the poor. Hence, if the rich are convinced that democracy is controlled by the interest of the middle class agents, they have less to fear from democracy and are less inclined to use repression to avoid it.” This idea finds support in a study where democratic institutional reforms are found likelier to be in countries with sizable middle classes [Loayza, Rigolini, and Llorente 2012].1 Arguably, the middle classes help bring about such reforms by espousing democratic views and values and then acting on them when needed.

The last two decades or so in Southeast Asia witnessed a range of participation of the middle classes in major events. In Thailand, the Bangkok middle classes have participated in political rallies that led to the ouster of the popularly elected Thaksin Shinawatra as prime minister in 2006. Thaksin’s sister, Yingluck Shinawatra, while also popularly elected as prime minister in 2011, nonetheless also faced massive anti-government protests in 2013 that culminated in her removal from office in 2014 by the Constitutional Court on corruption charges. Even in post-Soeharto Indonesia, corruption in the highest level of government continues to agitate various groups, including the middle classes that, together with the poor population, elected Joko Widodo for president in 2014. In the Philippines, the urban middle classes organized the so-called Million People March in August 2013 that brought in more than 100,000 Filipinos to Luneta Park to rally against corruption and pork barrel funds. Before that, they joined in the street rallies in edsa that culminated in the removal from office of President Joseph Estrada in early 2001. In Singapore and Malaysia, the same political parties continue to remain in power, although Malaysia’s United Malays National Organization is facing a growing opposition led by its former member Anwar Ibrahim, whose supporters include the youth, the ethnic Chinese, and the urban residents.

Notwithstanding the importance of these political activities, do they really distinguish the middle classes from the rest? Several previous studies on Southeast Asia (e.g., Robison, David, and Goodman [1996]; Embong [2001a]; Kimura [2003]) noted that even in earlier periods, the middle classes—loosely referring to those living in major urban areas, are professionals, are white-collar workers2, or are college educated—engaged in political rallies, protest demonstrations, and

1 For recent studies linking the middle classes, economic development and democracy, see, for example, Gerring et al. [2005], Banerjee and Duflo [2008], Chun, Hasan and Ulubasoglu [2011], and Birdsall [2015].2 The studies in Embong [2001a] use the Erikson-Goldthorpe scheme in classifying people into social classes on employment status and occupational classifications.

120 Capuno: Class participation in politics in Southeast Asia

election debates and campaigns to demand reforms or change in government. Rather than leading in these political activities, the middle classes, however, were merely following, and they were more sporadic than regular in their engagements. Their participation was prodded more by a general or popular sentiment (say against a repressive regime) and less by pure class consciousness or interest.

Two reasons are suggested for their apparently restrained role. Firstly, they also benefitted from the economic policies promulgated by the very regimes that they wanted to reform or change. Secondly, they continue to espouse traditional values that emphasize respect for authority and primacy of group welfare, just like the other Asian societies. Indeed, it was Singapore’s Lee Kwan Yew who famously argued that Asians can achieve economic growth without giving up traditional values, unlike Western societies (Emerson [1995]; Fukuyama [1995]).

Yet, it is hard to determine from the aforementioned studies whether the middle classes as a group are not unlike the rest. One the one hand, the socioeconomic classes are not well identified; much of the findings is based on the observed actions of known members of the middle class who may be extremist themselves. Second, there are other forms of political activities—like voting in elections or contacting officials that are less public or dramatic than others—wherein the socioeconomic classes may differentiate themselves. We thus re-investigate the issue of class participation in politics in five Southeast Asian countries.

Indonesia, Malaysia, the Philippines, Singapore, and Thailand are selected for two reasons. Firstly, all five have the survey data required for the method adopted here. Of the countries in Southeast Asia, only these five, Vietnam, and Cambodia were included in the third wave of the Asian Barometer Survey conducted in 2010-2012. Being in the same region, they all share a common geography, socio-cultural heritage, and history, which facilitate comparison. Secondly, they vary enough in terms of economic development and politics that can be explored to answer our main research question. The World Bank classifies Indonesia and the Philippines as lower-middle countries and Thailand and Malaysia as upper-middle income countries. Like most Asian developing countries, the four developing Southeast Asian countries also experienced in recent years reductions in poverty rates and expansion of the middle-income sub-population [Asian Development Bank 2010]. All four have positive polity iv scores in 2010, indicating they are more democratic than autocratic.3 Of the five, only Singapore is considered a high-income country, but it also has a negative polity iv score and, perhaps, a more entrenched Confucian culture. Arguably, Singapore may provide the benchmark for the group.

Following the lead of previous studies on the Asian values thesis that used survey data (e.g., Chang, Zhu, and Pak [2007]; Chang, Chu, and Diamond

3 polity iv is developed by the polity iv Project. http://www.systemspeace.org/polity/polity4htm. Accessed 30 July 2015.


[2012]), we use for this paper the dataset from the third round of the Asian Barometer Survey. Instead of using the self-reported income status or occupation, we construct wealth/asset indices and then delineate the nationally representative sample into three socioeconomic classes. We then compare the three classes in terms of proportions that participate in certain political activities. The differences in proportions are further decomposed using the Oaxaca method into two parts: one part that is due to differences in the observed characteristics (like age, education, urbanity, occupation) of the classes; and the other part that is due to heterogeneous effects of the same characteristics across classes.4 Unlike in previous studies, our approach here allows us to measure and decompose the class differences in their political participation and to compare the results across countries.

In general, we find high rates of voting participation across classes in Indonesia, Malaysia, the Philippines, and Thailand. In Singapore, barely half of the sample voted in previous elections. In Indonesia, Thailand, and Malaysia, the middle classes distinguish themselves from the rest by participating in other political activities, including contacting officials or the news media, joining others to voice out or directly address their common concerns, and attending rallies or demonstrations. Interclass differences in political participation is due more to divergence in mean characteristics in Thailand, but they are due more to the heterogeneous effects of these characteristics in Indonesia and Malaysia. Relative to the middle classes in these three countries, those in the Philippines and Singapore each appears politically disengaged.

The rest of the paper is organized as follows. Section 2 discusses the data used in the analysis.5 Section 3 details the methods used in the analysis. The results are then presented in Section 4. The concluding section discusses the implications of the findings.

2. Data

2.1. Survey data

The data used here is obtained from the third wave of the Asian Barometer Survey6 jointly conducted by Taiwan-based Academia Sinica and the National Taiwan University. The third-wave surveys were conducted in 2010-2012 in nine East Asian and Southeast Asian countries, including Indonesia, Malaysia, the Philippines, Singapore, and Thailand. Local institutions administered the country-level surveys and used the same survey instrument and sampling design to collect comparable data. In each country, a nationally representative, random sample of

4 Dow [2009] used the same method to account for the differences in political views between genders.5 Sections 2 and 3 draw heavily from Capuno [2015]. 6 The website is www.asianbarometer.org.


voting-age population served as interview respondents. The sample consisted of 1,550 for Indonesia, 1,214 for Malaysia, 1,200 for the Philippines, 1,000 for Singapore, and 1,512 for Thailand.7

The survey questionnaire used has several modules designed to capture, among others, the respondent’s participation in elections and other political activities and his or her demographic and socioeconomic characteristics (e.g., family composition, self-reported household income level, and social status). Additionally, the enumerator also records his or her own observations about the respondent and the community characteristics in the interview sites.

2.2. Indicators of political participation

We constructed ten binary indicators of political participation. The first 4 indicators pertain to electoral participation. These indicators are voted_last election (= 1 if the respondent voted in the most recent national election, parliamentary, or presidential; 0 otherwise), voted_most elections (= 1 if voted in every or most past elections, 0 otherwise), member_political organization (= 1 if member of a political party or of an organization that supports candidates, 0 otherwise), and support_candidate (= 1 if tried to persuade others to vote for a certain candidate or party or do anything else to help out or work for a party or candidate running in the election, 0 otherwise). Note that the valid samples for the above indicators exclude those who were not eligible to vote or to participate in the past elections, possibly because they were too young then.

The next three indicators pertain to personal initiatives to draw the attention of authorities presumably to their concerns. The indicators of personal engagement are contacted_official (= 1 if contacted elected officials or legislative representatives or officials at higher level at least once in the past three years, 0 otherwise), contacted_other leaders (= 1 if contacted traditional leaders/community leaders or other influential people outside the government at least once in the past three years, 0 otherwise), and contacted_media (= 1 if contacted news media more than once in the past three years, 0 otherwise). Note that it is possible that those who responded did the implied political activities together with others, but that is not asked in the survey.

The last three indicators pertain to collective actions other than electoral participations to directly address a common concern or to call public attention. The three indicators are participated_with others (= 1 if got together with others to try to resolve local problems or to raise an issue or sign a petition at least once in the past three years, 0 otherwise), attended_rally (= 1 if attended a campaign

7 The sample sizes were set to allow for a minimum confidence interval of ±3 percent at the 95 percent probability. The sample sizes are standard in similar surveys (e.g., World Values Survey, EuroBarometer Survey).


meeting or rally in the last national election, 0 otherwise), and attended_demonstration (= 1 if attended a demonstration or protest march more than once in the past three years, 0 otherwise). Note that the samples for attended_rally exclude those who responded “not applicable” to the relevant survey question, possibly again because they were too young.

No single indicator is a definitive measure of political engagement, but altogether they will indicate the extent. Our participation indicators are similar to those in Wu and Lee [2012] and Verba, Nie, and Kim [1978]. Table 1 presents the variables used in this paper and their definitions.

TABLE 1. Variable definitions

Variable name DefinitionPolitical participationVoted_last election = 1 if voted in the last election, 0 otherwiseVoted_most elections = 1 if voted in all or most past elections, 0 otherwise Member_political organization

= 1 if member of a political party or an organization that supports a candidate, 0 otherwise

Supported_candidate = 1 if tried to persuade others to vote for a certain candidate or party or do anything else to help out or work for a party or candidate running in the election, 0 otherwise

Contacted_officials = 1 if contacted elected officials or legislative representatives or officials at higher level at least once in the past three years, 0 otherwise

Contacted_other leaders = 1 if contacted traditional leaders/community leaders or other influential people outside the government at least once in the past three years, 0 otherwise

Contacted_media = 1 if contacted news media more than once in the past three years, 0 otherwise

Participated_with others = 1 if got together with others to try to resolve local problems or to raise an issue or sign a petition at least once in the past three years, 0 otherwise

Attended_rally = 1 if attended a campaign meeting or rally in the last national election, 0 otherwise

Attended_demonstration = 1 if attended a demonstration or protest march more than once in the past three years, 0 otherwise

Economic classLower class = 1 if belongs to wealth quintile 1 or 2, 0 otherwiseMiddle class = 1 if belongs to wealth quintile 3 or 4, 0 otherwiseUpper class = 1 if belongs to wealth quintile 5, 0 otherwiseOther characteristicsMale = 1 if male, 0 otherwiseAge Age in yearsAge2 Square of ageSome college = 1 if has at least some college education, 0 otherwiseFinished college = 1 if finished college, 0 otherwiseEmployed = 1 if employed, 0 otherwise Catholic = 1 if religion is Roman Catholic, 0 otherwise


Variable name DefinitionBuddhist = 1 if religion is Buddhism, 0 otherwiseIslam = 1 if religion is Islam, 0 otherwiseIn union = 1 if married or living in with a partner, 0 otherwise Household size Number of household membersUrban = 1 if residing in urban area, 0 otherwiseMega_majorcity = 1 if living in capital city, megacity, regional center, or major city,

0 otherwiseAccess index = 100 x {[paved road + public transport + post office + school +

police + sewerage + clinic + cellular signal + recreation facility + church + town hall + market stall]/12}

2.3. Socioeconomic classes

Following the method in Filmer and Pritchett [2001], we classified into economic classes the samples in each country based on wealth indices constructed, which are based on the reported household amenities and assets. The lists of amenities and assets vary slightly across countries. First, we developed binary indicators of the household’s connection to electricity and piped water and ownership of motor vehicle (car or jeep), tractor, television set, cable television, motorcycle, telephone, mobile phone, bicycle, radio, pumping set, refrigerator, camera, and livestock (goat and cow). Next, we took the mean and standard deviation of each indicator, and then we derived the first principal component of the vector containing these indicators. These statistics are then used to generate a score for each household asset or amenity. The asset scores are then added up to derive the household’s overall wealth index. Ranked from lowest to highest wealth index, the households are then grouped into quintile.

The first two quintiles together are classified as the lower class, the third and fourth quintiles together as the middle class, and the fifth quintile as the upper class. By construction, the lower class may include households with incomes above the poverty threshold but who may become poor due to sudden unemployment, illness, calamity, or other adverse shocks. As well, the middle class possibly includes some wealthy households whose fortunes are still modest to be considered part of the economic elite. By construction, the last group is the wealthiest and the smallest in size (only 10 percent of the sample). This is consistent with the observation that the rich are few and rarely participate in such surveys.

For the Philippines, Thailand, or Indonesia, the first principal component accounts for 21 percent of the within-country variations in the household amenities and assets (and livestock). The corresponding proportions are 18 percent in Singapore and 15 percent in Malaysia. To further validate the wealth quintiles, we correlated them with the self-declared household income levels. In the surveys, the respondents were asked to identify where they are at in a five-rung income ladder, where the first rung is the “lowest level” and the fifth and last


rung is the “highest level”. The resulting correlation coefficients are 63 percent for Thailand, 55 percent for Indonesia, 47 percent for the Philippines, 46 percent for Malaysia, and 25 percent for Singapore.8

Finally, we also compared the proportions of white-collar workers in the lower class, middle class, and upper class in each country. The expectation is that most, if not all, white-collar workers would belong in the middle and upper classes.9 The proportions of white-collar workers among the employed are as follows: in the Philippines, 34 percent in the lower class, 57 percent in the middle class, and 66 percent in the upper class; in Thailand, 12 percent in the lower class, 29 percent in the middle class, and 72 percent in the upper class; in Indonesia, 17 percent in the lower class, 41 percent in the middle class, and 66 percent in the upper class; in Singapore: 76 percent in the lower class, 82 percent in the middle class, and 90 percent in the upper class; and in Malaysia, 40 percent in the lower class, 63 percent in the middle class, and 70 percent in the upper class.

Overall, our wealth quintiles and the economic classes derived from them correlate fairly well with other indicators of social classes. The correlation is relatively lower in Singapore, arguably because of the limited list of amenities and assets contained in the Asian Barometer Survey questionnaire. Furthermore, the choices are too commonplace to sharply distinguish Singaporeans by wealth status. Since most Singaporeans are well off, and all live in urban areas, they are likely to own all the assets in the list, save perhaps the livestock and farm equipment.

2.4. Covariates

Table 1 shows the variables used to further characterize the sample respondents. The individual-level characteristics are gender (male = 1, 0 otherwise), age (in years), attainment of at least some tertiary-level of formal schooling (at least some college), employment status (employed), married or living-in with partner (in-union), and religion (Catholic, Islam, Buddhist). The household-level characteristics include household size and residence in an urban area, capital city, megacity, or regional center. Both age and its squared term (age2) and household size are continuous variables; the rest of the aforementioned variables are dichotomous variables.

8 The correlation estimates are for samples that reported their household income level and excludes those with missing responses or replied not to know their income levels.9 The classification of the respondents into “white-collar” worker or otherwise is based on the scheme proposed by Erikson and Goldthorpe [1992]. In this paper, “white-collar” workers include “professionals, large enterprise employers, managers (with 1-10 or more than 10 subordinates), and routine clerical and sales workers.” However this classification can be applied only to respondents who, at the time of the survey, were currently employed and declared their main occupation.


We also constructed an access index based on the reported presence of certain public facilities or amenities within the respondent’s community or immediate environment. In developing the index, first we developed binary indicators of the presence of paved roads leading to the respondent’s abode, and of access to a public transportation, post office, school, police station, public sewerage facility, clinic, cellular phone signal, recreation facility, church, town hall, and marketplace. Each indicator has a value of 1 if the reference public amenity or facility is present and 0 if otherwise. Then we summed up the values of these indicators, divided the total by 12, and finally multiplied the resulting quotient by 100 to derive the household’s access index score. The access index score ranges from 0 (lowest) to 100 (highest). A high score implies that the household has fairly easy access to sources of information or facilities for mobility or social interactions.

Annex 1 shows the mean and standard deviations of the covariates for each country. In each country, about half of the sample is male, and the mean age is between 41 to 47 years. Between 9 percent in Indonesia and 48 percent in Singapore attained some college education. In the Philippines, about 84 percent are Catholic. In Indonesia, 87 percent practice Islam, while the figure is 57 percent in Malaysia. The proportion of Buddhists is 94 percent in Thailand and 40 percent in Singapore. In each country, at least 48 percent claims to be employed and at least 68 percent are married or living-in with a partner (in union). The mean household size ranges from around 4 (Singapore) to nearly 5.7 (Malaysia). Less than half of the respondents in Indonesia, Malaysia, and Thailand live in urban areas. A huge majority of the sample in the Philippines resides in megacities or major cities, while less than 25 percent of the samples in Indonesia or Malaysia do so. (Note that this indicator cannot be constructed with the survey data for Thailand and Singapore.) Finally, the mean access index is lowest in the Philippines at 65.10 percent and highest in Singapore at 98.08 percent. As expected, around forty percent of the sample in each country belongs to the lower class and another forty percent to the middle class; the rest belongs to the upper class.

3. Methods

3.1. Test of proportions

For each country sample, we apply tests of proportions to determine if the middle class differs from other economic classes in terms of political participation. Basically, the test ascertains if the middle class and another class have statistically equal proportions of members who, say, voted in the last election. If the proportions are not equal, we then proceed to decompose the difference using the Oaxaca method.


3.2. Oaxaca decomposition method

First introduced in Blinder [1973] and Oaxaca [1973], the original Oaxaca decomposition method (also called the Blinder-Oaxaca decomposition method) was used to decompose the sources of wage differentials between sexes, where wages are a linear function of gender and other control variables. In this paper, we follow the extension of the original method to non-linear functions made by Sinning, Hanh, and Bauer [2008].

To fix ideas, consider an outcome Y being a nonlinear function of the characteristics of members of group g. Y could be a dichotomous indicator of political participation (e.g., 1 = Yes, 0 otherwise) and is related by a logistic function, say, to a vector of observable characteristics X (that contains, for example, gender, employment, and educational attainment). Let the average outcome for group M be denoted by Y

M. Let the average outcome for group L

be denoted by YL. The difference in the average outcomes can be decomposed as

follows:

YM - Y

L = {EβL

(YiM

| XiM

) - EβL (Y

iL| X

iL) } + {EβM

(YiL| X

iL) - EβL

(YiL| X

iL)}

+ {[EβM(Y

iM| X

iM) - EβL

(YiM

| XiM

)] - [EβM (Y

iL| X

iL) - EβL

(YiL| X

iL)]},

where Eβg (Y

ig|X

ig) refers to the conditional expectation of Y

ig, and Eβg

(Yig|X

ih) refers

to the conditional expectation of Yih evaluated at the parameter vector β

g, with

g,h= (M,L) and g ≠ h. To elaborate, the difference in the mean outcomes of groups M and L can be

decomposed into three parts, as represented by the three terms in parentheses in the right-hand side of the equation above.

The first part, {EβL (Y

iM| X

iM) - EβL

(YiL| X

iL) }, is due to the differences in the

average characteristics of the two groups. Specifically, it denotes the change in the mean outcome of group L if its members attain the same characteristics as those in group M, but multiplied by the coefficients of group L.

The second part, {EβM(Y

iL| X

iL) - EβL

(YiL| X

iL)}, is due to differences in the

coefficients of the two groups. It denotes the change in the mean outcome of group L if its coefficients are set equal to that of group M, but evaluated at the same average characteristics of group L.

The last term, {[EβM(Y

iM| X

iM) - EβL

(YiM

| XiM

)] - [EβM (Y

iL| X

iL) - EβL

(YiL| X

iL)]},

derives from the interaction of the differences in the mean characteristics and in the coefficients of the two groups. In our empirical implementation of this decomposition, the reference group is the middle class (M) and the comparison group (L) could be other households not in the middle class (“other class” means low class and upper class), those that are in the low class alone, or those that are in the upper class alone.


To illustrate the insight that can be gained with this method, consider one characteristic, say, college education, and one outcome, say, voted in the last election. Suppose the middle class has a higher proportion of members than the lower class who voted. The decomposition allows us to determine how much of the difference in proportions would be reduced if the lower class achieved the same average level of education as the middle class. This is important since if by doing so the differences are completely or substantially eliminated then the case for subsidized college education of the underprivileged is strengthened, in this case the higher education will inculcate in them the virtue and value of exercising their rights and duties to vote.

However, if the coefficients are significantly different, the decomposition method will also indicate that the same college education provided to the poor will not make them exercise their voting rights as much as the middle class, which renders such an education policy inadequate for the purpose. The differences in coefficients reflect the heterogeneous effects of the same intervention targeted to the two groups. These may arise out of idiosyncrasies in personal (or class) experiences or expectation, or these may be conditioned by the social context or culture to which members of the same class belong. The decomposition method used here provides an estimate and the statistical significance of the heterogeneous effects, but it does not identify the reasons for the differential impacts. Notwithstanding this limitation, distinguishing the differences in mean characteristics from the differences in their effects to explain the variations in political participation across classes is still relevant to policy. Since at least some observable characteristics (like education and employment) are directly influenced by government intervention, the decomposition method can shed light on the effectiveness in terms of shaping a national political consensus of government policies and programs, on the one hand, and of possible (yet still unidentified) psychosocial and cultural factors, on the other.

Since the dependent variables are all dichotomous variables, we use probit regression models to estimate the differences in the probability of an outcome (measuring political participation) conditional on a list of covariates (characteristics). To carry out non-linear Oaxaca decomposition analysis, we use the Stata module nldecompose [Sinning, Hahn, and Bauer 2008]. Note that in both the tests of proportions and Oaxaca decompositions, the reference group is the middle class. Thus, a negative (positive) difference in proportions indicates a lower (higher) proportion of the middle class than either the lower class or upper.


4. Results

4.1. Test of proportions

Table 2 shows the results of the tests of proportions for Indonesia, the Philippines, and Thailand. In the left panel, we find that, overall, a high proportion (90 percent) of Indonesians exercised their right to vote, but a significantly lower percentage had other forms of political engagements. Only a fourth of them reported to have joined a rally or a campaign meeting, and about one in ten joined forces with others to address a collective issue.

Nonetheless, there are indications that the Indonesian middle class is different from the other classes. When compared to the lower class, they appear to be more active in being members of political organizations, in supporting political candidates or parties, in contacting local officials or the news media, or in joining others to address a common concern. When compared to the rich, however, they appear less inclined to contact officials or the news media or to team up with others to resolve a common concern.

In the middle panel, we find that nearly 80 percent of the Filipinos voted in past elections, and about 23 percent of them reported to have joined political rallies. Less than five percent reported to have participated in other types of political activities. Generally, members of the Filipino middle class do not appear to be more or less politically active or engaged than the lower class or upper class. They appear to be distinct from the lower class only in that a lower percentage of them have joined forces with others to address a common problem.

Like Indonesians, most Thais exercised their right to vote (right panel). Unlike in Indonesia and the Philippines, nearly half of the Thais claimed to have joined political campaigns or rallies during the last national election. Less than 10 percent of Thais reported to have participated in other forms of political activities. When compared to the lower class, members of the middle class are less likely to vote or to attend a rally, but they are more likely to contact the news media or participate with others. When compared to the rich, the middle class is likelier to vote, to support a political party or a candidate, or to attend a rally.


TAB

LE 2

. Tes

t of p

ropo

rtio

ns: I

ndon

esia

, the

Phi

lippi

nes,

and

Tha

iland

Ind

icat

or

Ind

one

sia

Phi

lipp

ines

Tha

iland

Ove

rall

pro

po

rtio

n

Diff

eren

ce in

pro

po

rtio

ns

Ove

rall

pro

po

rtio

n

Diff

eren

ce in

p

rop

ort

ions

Ove

rall

pro

po

rtio

n

Diff

eren

ce in

pro

po

rtio

ns

Mid

dle

cl

ass

vs.

oth

ers

Mid

dle

cl

ass

vs.

low

er

clas

s

Mid

dle

cl

ass

vs.

upp

er

clas

s

Mid

dle

cl

ass

vs.

oth

ers

Mid

dle

cl

ass

vs.

low

er

clas

s

Mid

dle

cl

ass

vs.

upp

er

clas

s

Mid

dle

cl

ass

vs.

oth

ers

Mid

dle

cl

ass

vs.

low

er

clas

s

Mid

dle

cl

ass

vs.

upp

er

clas

sVo

ted

_las

t el

ectio

n0.

903

-0.0

05-0

.020

0.02

60.

798

0.02

40.

027

0.01

80.

913

0.00

3-0

.033

**0.

072*

**[0

.029

6](0

.015

)(0

.016

)(0

.023

)[0

.401

](0

.025

)(0

.028

)(0

.035

)[0

.283

](0

.015

)(0

.015

)(0

.024

)N

=15

50

N=

1032

N

=15

09Vo

ted

_mos

t el

ectio

ns0.

923

-0.0

08-0

.015

0.00

60.

786

0.01

40.

025

-0.0

110.

921

0.00

3-0

.033

**0.

076*

**[0

.266

](0

.014

)(0

.015

)(0

.020

)[0

.410

](0

.025

)(0

.027

)(0

.033

)[0

.268

](0

.014

)(0

.014

)(0

.023

)N

=15

28

N=

1127

N

=14

98M

emb

er_p

oliti

cal

orga

niza

tion

0.07

00.

030*

*0.

035*

*0.

019

0.02

20.

003

0.00

8-0

.008

0.00

6-0

.002

-0.0

01-0

.008

[0.2

56]

(0.0

14)

(0.0

15)

(0.0

19)

[0.1

46]

(0.0

09)

(0.0

09)

(0.0

14)

[0.0

77]

(0.0

04)

(0.0

04)

(0.0

07)

N=

1550

N

=12

00

N=

1512

Sup

por

ted

_ca

ndid

ate

0.08

40.

016

0.02

6*-0

.006

0.11

90.

013

0.02

3-0

.009

0.08

10.

014

-0.0

010.

037*

*[0

.277

](0

.015

)(0

.015

)(0

.021

)[0

.324

](0

.021

)(0

.022

)(0

.030

)[0

.273

](0

.015

)(0

.016

)(0

.017

)N

=15

50

N=

1032

N

=15

03C

onta

cted

_of

ficia

ls0.

015

0.00

10.

013*

*-0

.023

**0.

026

-0.0

11-0

.008

-0.0

170.

034

0.01

40.

016

0.00

7[0

.124

](0

.006

)(0

.006

)(0

.012

)[0

.159

](0

.009

)(0

.009

)(0

.014

)[0

.181

](0

.010

)(0

.010

)(0

.013

)N

=15

50

N=

1200

N

=15

09C

onta

cted

_oth

er

lead

ers

0.04

60.

006

0.01

8-0

.020

0.02

80.

006

0.00

50.

010

0.04

0-0

.002

0.00

3-0

.016

[0.2

09]

(0.0

11)

(0.0

11)

(0.0

17)

[0.1

66]

(0.0

10)

(0.0

11)

(0.0

13)

[0.2

00]

(0.0

10)

(0.0

11)

(0.0

15)

N=

1550

N

=12

00

N=

1509


Ind

icat

or

Ind

one

sia

Phi

lipp

ines

Tha

iland

Ove

rall

pro

po

rtio

n

Diff

eren

ce in

pro

po

rtio

ns

Ove

rall

pro

po

rtio

n

Diff

eren

ce in

p

rop

ort

ions

Ove

rall

pro

po

rtio

n

Diff

eren

ce in

pro

po

rtio

ns

Mid

dle

cl

ass

vs.

oth

ers

Mid

dle

cl

ass

vs.

low

er

clas

s

Mid

dle

cl

ass

vs.

upp

er

clas

s

Mid

dle

cl

ass

vs.

oth

ers

Mid

dle

cl

ass

vs.

low

er

clas

s

Mid

dle

cl

ass

vs.

upp

er

clas

s

Mid

dle

cl

ass

vs.

oth

ers

Mid

dle

cl

ass

vs.

low

er

clas

s

Mid

dle

cl

ass

vs.

upp

er

clas

sC

onta

cted

_m

edia

0.01

9-0

.002

0.01

3**

-0.0

35**

*0.

030

0.01

00.

009

0.01

40.

043

0.01

30.

020*

-0.0

04[0

.138

](0

.007

)(0

.006

)(0

.014

)[0

.171

](0

.010

)(0

.011

)(0

.013

)[0

.202

](0

.011

)(0

.011

)(0

.016

)N

=15

50

N=

1200

N

=15

03P

artic

ipat

ed_w

ith

othe

rs0.

126

0.00

80.

034*

-0.0

48*

0.06

3-0

.023

-0.0

29*

-0.0

090.

082

0.01

40.

034*

*-0

.022

[0.3

32]

(0.0

17)

(0.0

18)

(0.0

26)

[0.2

44]

(0.0

14)

(0.0

16)

(0.0

19)

[0.2

75]

(0.0

15)

(0.0

15)

(0.0

21)

N=

1550

N

=12

00

N=

1510

Att

end

ed_r

ally

0.24

8-0

.004

-0.0

250.

041

0.22

60.

010

0.02

3-0

.019

0.49

9-0

.078

***

-0.1

75**

*0.

112*

**[0

.432

](0

.022

)(0

.025

)(0

.029

)[0

.418

](0

.027

)(0

.029

)(0

.038

)[0

.500

](0

.026

)(0

.028

)(0

.034

)N

=15

50

N=

1032

N

=15

07A

tten

ded

_d

emon

stra

tion

0.02

10.

004

0.00

7-0

.004

0.02

9-0

.009

-0.0

08-0

.013

0.05

10.

002

-0.0

020.

004

[0.1

42]

(0.0

08)

(0.0

08)

(0.0

11)

[0.1

68]

(0.0

10)

(0.0

10)

(0.0

14)

[0.2

20]

(0.0

13)

(0.0

14)

(0.0

17)

N=

1550

N

=12

00

N=

1509

Not

es: F

igur

es in

bra

cket

s ar

e st

and

ard

dev

iatio

ns. F

igur

es in

par

enth

eses

are

sta

ndar

d e

rror

s. N

mea

ns n

umb

er o

f ob

serv

atio

ns.

*p<

0.10

**p

<0.

05**

*p<

0.01

TAB

LE 2

. Tes

t of p

ropo

rtio

ns: I

ndon

esia

, the

Phi

lippi

nes,

and

Tha

iland


Table 3 shows the corresponding results for Singapore and Malaysia. Relative to Indonesians, Filipinos and Thais, Singaporeans in general appear to be less politically engaged. Less than half of them voted in most of the past elections, and only 56 percent of them voted in the last election. Less than eight percent of them joined in rallies, and less than four percent joined any other political activity. Moreover, a lower proportion of the middle class than the lower class voted in most elections or contacted any government official. Also, a lower percentage of them than the upper class supported a political party or candidate.

The overall pattern of political participation of Malaysians is closer to that of the Indonesians, Filipinos, or Thais than to Singaporeans. About three in every four Malaysians exercised their right to vote. About one in ten was a member of political parties or organizations or supported political candidates. Nearly two in ten contacted government officials or joined others to address a common concern. About three in ten attended political rallies. When compared to the lower class, the middle class has a greater percentage of members who belong to political organizations or contacted officials, but less of them attended demonstrations. When compared to the upper class, the middle class appears less likely to contact traditional leaders or community leaders or to contact news media.

TABLE 3. Test of proportions: Singapore and Malaysia

Indicator

Singapore Malaysia

Overall proportion

Difference in proportions

Overall proportion


Middle class vs. others

Middle class vs. lower class

Middle class vs. upper class




Voted_last election 0.554 -0.022 -0.017 -0.035 0.770 0.008 0.014 0.004[0.497] (0.034) (0.038) (0.047) [0.421] (0.027) (0.029) (0.037)N=855 N=1026

Voted_most elections

0.484 -0.059* -0.070* -0.034 0.746 -0.013 -0.001 -0.033[0.500] (0.034) (0.037) (0.048) [0.436] (0.028) (0.030) (0.037)N=855 N=1026

Member_political organization

0.012 0.0003 0.002 -0.004 0.100 0.032* 0.042** 0.012[0.109] (0.007) (0.007) (0.011) [0.300] (0.019) (0.019) (0.025)N=1000 N=1214

Supported_candidate

0.012 -0.013* -0.011 -0.016* 0.105 -0.015 -0.012 -0.019[0.110] (0.006) (0.007) (0.011) [0.306] (0.018) (0.019) (0.024)N=979 N=1214

Contacted_officials 0.02 -0.013 -0.018* -0.004 0.288 0.023 0.062** -0.054[0.140] (0.008) (0.01) (0.011) [0.453] (0.027) (0.028) (0.037)N=1000 N=1214

Contacted_other leaders

0.008 -0.005 -0.008 -0.0004 0.181 -0.024 0.002 -0.075**[0.089] (0.005) (0.007) (0.006) [0.385] (0.022) (0.024) (0.032)N=1000 N=1214

Contacted_media 0.022 0 0.002 -0.004 0.060 -0.013 -0.001 -0.038**[0.147] (0.009) (0.010) (0.014) [0.238] (0.014) (0.014) (0.021)N=1000 N=1214


Indicator

Singapore Malaysia

Overall proportion


Overall proportion








Participated_with others

0.038 -0.019 -0.016 -0.026 0.167 -0.015 -0.014 -0.014[0.191] (0.012) (0.013) (0.018) [0.373] (0.022) (0.024) (0.029)N=1000 N=1214

Attended_rally 0.076 0.014 0.021 -0.002 0.316 -0.003 -0.029 0.052[0.265] (0.017) (0.019) (0.025) [0.465] (0.027) (0.030) (0.035)N=973 N=1214

Attended_demonstration

0.01 -0.004 -0.0002 -0.0008 0.030 -0.030*** -0.036*** -0.016[0.010] (0.006) (0.007) (0.009) [0.172] (0.009) (0.011) (0.012)N=1000 N=1214

Notes: Figures in brackets are standard deviations. Figures in parentheses are standard errors. N means number of observations.*p<0.10**p<0.05***p<0.01

In summary, there is generally a high level of participation in elections among Southeast Asian people. A small but significant minority in most countries attends rallies, contacts officials or the news media, or joins others to voice out or directly address their common concerns. In these political activities, the middle classes distinguish themselves from the rest. The factors that make them distinct are identified below.

4.2. Oaxaca decomposition10

Table 4 shows for Indonesia, the Philippines, and Thailand the results of the Oaxaca decomposition of the sources of the class differences in political participation. Note that the decomposition is done only for those participation indicators where class differences are statistically significant. In the left panel, we see that, in Indonesia, the differences between the middle class and the lower class are not due to the dissimilarities in their average characteristics but rather to the relatively bigger effects of the same characteristics on the middle class. The results show that the differences in coefficients account for at least three percentage points of the total gap in proportions of the middle class and lower class that reported to be members of political organizations or supported political candidates. Furthermore, the differences in coefficients

10 The results reported are based on a set of 32 unweighted regression runs, one pair for each indicator of political participation where proportions of the middle classes and another class (lower classes, upper classes) differ. Sampling weights are not used to make the results comparable with those obtained in the tests of proportions, which do not allow weights. The detailed regression results are not reported here to save on space, but they are available from the author upon request.


account for at least a percentage point of the total gap in proportions that contacted officials or the news media. The decomposition reveals, however, that the differences between the middle and upper classes are only apparent. Neither the differences in average characteristics, their differential impact on the likelihood of political participation, or the interaction of these two factors appears to be statistically significant.11

TABLE 4. Oaxaca Decomposition of political participation: Indonesia, the Philippines, and Thailand

Indicator

Indonesia Philippines Thailand







Middle class vs. other



Voted_last election

Characteristics -0.038*** 0.038*(0.009) (0.020)

Coefficients 0.042* -0.002(0.025) (0.020)

Interaction -0.036* 0.039 (0.020) (0.026)Voted_most elections

Characteristics -0.039*** 0.068***(0.009) (0.023)

Coefficients -0.005 0.028(0.022) (0.026)

Interaction 0.007 -0.014(0.016) (0.03)


Characteristics -0.009 -0.005(0.008) (0.012)

Coefficients 0.027* 0.030*(0.015) (0.016)

Interaction 0.012 0.010(0.009) (0.015)

11 Note that the decomposition method applied here works only when the two classes have the same set of characteristics. When a characteristic is unique to or highly prevalent in a class, it is dropped by the statistical method used here. This is the reason why a number of decomposition results in Tables 4 and 5 are statistically insignificant. It is possible the omitted variable may be the one that drives the class differences.


Indicator

Indonesia Philippines Thailand







Middle class vs. other



Supported_candidate

Characteristics 0.008 0.005 (0.009) (0.017)

Coefficients 0.036** 0.005 (0.017) (0.026)

Interaction -0.017 0.021

(0.011) (0.024)Contacted_officials

Characteristics -0.001 -0.008 (0.004) (0.010)

Coefficients 0.013** -0.014 (0.006) (0.014)

Interaction 0.001 -0.001 (0.004) (0.011)Contacted_media

Characteristics 0.002 -0.015 0.016*** (0.002) (0.010) (0.005)

Coefficients 0.012*** -0.019 -0.006 (0.005) (0.012) (0.024)

Interaction -0.001 -0.001 0.008 (0.003) (0.012) (0.019)Participated_with others

Characteristics -0.010 0.015 -0.001 -0.002 (0.011) (0.011) (0.007) (0.009)

Coefficients 0.032 -0.040 -0.043*** 0.030* (0.020) (0.030) (0.016) (0.017)

Interaction 0.013 -0.022 0.015 0 (0.014) (0.023) (0.012) (0.013)Attended_rally Characteristics 0.004 -0.046*** 0.103***

(0.01) (0.013) (0.031) Coefficients -0.100*** -0.131*** 0.002

(0.026) (0.027) (0.042) Interaction 0.009 -0.011 0.004 (0.01) (0.02) (0.036)

Figures in parentheses are bootstrap standard errors estimated with 100 replications.*p<0.10**p<0.05***p<0.001


In the case of the Philippines, we also find that it is the heterogeneous effects rather than the dissimilarities in average characteristics that distinguish the middle class from the lower class in terms of likelihood to team up with others to address a common concern. Unlike in the Indonesia, however, the effects are lower for the middle class than the lower class.

In the case of Thailand, it is the dissimilarities in mean characteristics that make the middle class less likely than the lower class to vote in elections and more likely than the upper class to vote or to attend rallies. We also find that the interaction of the differences in the average characteristics and in the coefficients explains why the middle class is less likely than the lower class to vote in the previous election. The heterogeneous effects of average characteristics also explain why the middle class appears more likely than the lower class to team up with others to solve a common problem but less likely than the lower class to attend rallies. Finally, the apparent discrepancies in proportions of middle class and upper class that supported political candidates are neither due to divergence in mean characteristics or in their differential effects.

Table 5 shows the results of the decomposition analysis for Singapore and Malaysia. In the case of Singapore, it is the differential impacts of the characteristics that make the middle class less likely than the rest to support a candidate or than the lower class to contact any government official. In addition, we find that the interaction of the differences in average characteristics and the differences in coefficients also explain why the middle class is likelier than the lower class to contact government officials The decomposition results indicate that it is not the differences in mean characteristics, the differences in the impacts of the characteristics, or the interaction of these two differences that account for why the middle class and the lower class are not equally likely to vote, or why the middle class and the upper class are not equally likely to support a candidate.

TABLE 5. Oaxaca Decomposition of political participation: Singapore and Malaysia

Singapore MalaysiaIndicator Middle

class vs. others






Voted_most elections Characteristics -0.012 -0.028

(0.017) (0.020) Coefficients -0.052 -0.037

(0.032) (0.034) Interaction 0.005 -0.005 (0.017) (0.020)


Singapore MalaysiaIndicator Middle

class vs. others







Characteristics -0.002 -0.010(0.009) (0.011)

Coefficients 0.035** 0.054***(0.018) (0.017)

Interaction -0.001 -0.003(0.009) (0.014)

Supported_candidate Characteristics -0.002 -0.004

(0.005) (0.009) Coefficients -0.012** -0.011

(0.005) (0.008) Interaction 0.001 -0.001 (0.005) (0.009)Contacted_officials Characteristics -0.001 0.017

(0.003) (0.020) Coefficients -0.032*** 0.059*

(0.012) (0.033) Interaction 0.015** -0.013 (0.007) (0.026)Contacted_other leaders Characteristics -0.005

(0.011) Coefficients -0.065*

(0.035) Interaction -0.006 (0.021)Contacted_media Characteristics 0.006

(0.007) Coefficients -0.052**

(0.023) Interaction 0.008 (0.015)Attended_demonstration Characteristics 0.001 0.001

(0.004) (0.003) Coefficients -0.029*** -0.043***

(0.008) (0.013) Interaction -0.001 0.006 (0.006) (0.008)

Figures in parentheses are bootstrap standard errors estimated with 100 replications.*p<0.10**p<0.05***p<0.001


Finally, in the case of Malaysia, we find that it is more the heterogeneous effects of characteristics that distinguish the middle class from the other classes. In particular, this divergence makes the middle class more likely than the lower class to become members of political organizations or to contact officials. However, the dissimilarities in coefficients make the middle class less probable than the upper class to contact traditional or community leaders or less likely than the lower class to attend demonstrations.

In sum, we find evidence that the differences in class participation in politics are explained by differences in mean characteristics and in the effects of these characteristics. However, it is not always the case that raising the average characteristics of the lower class to match that of the middle class will make them as politically engaged as the latter; in fact, there are instances when doing so will make them less engaged. Part of the reason is that the same characteristics may lead to lower or negative incremental effects on their probability of political participation.

5. Discussion and conclusion

Here we find evidence in five Southeast Asian countries that, generally, the middle classes are like other socioeconomic classes in terms of voting participation, but they are unlike the rest in other forms of political engagement. We also find that the middle classes are not always more politically active than other classes. Moreover, if the characteristics that seem to make them active were adopted by or transplanted in others, they may yield a different, even negative, effect on others. We also find the middle-class activism varies in both extent and form across countries.

In democracies, elections are an important institutional mechanism by which the people hold their leaders accountable. As the pivotal group of voters, the middle class is also expected to participate more in this activity than others. Our results show that voting participation is equally high across socioeconomic classes in Indonesia, the Philippines, Thailand, and Malaysia. In Thailand, the voting participation rate of the middle class is slightly lower than that of the lower class, but it is higher than that of the upper class. However, the middle classes are distinct in their involvement in other election-related activities, such as membership in political organizations in Indonesia and Malaysia or supporting a political candidate or party in Indonesia and Thailand.

Democracy also thrives in other forms of political activism. In Indonesia, Thailand, and Malaysia, the middle classes further distinguish themselves, in terms of contacting government officials or news media, participating with others, and attending rallies and demonstration. In Indonesia and Thailand, the differential participation rates across classes is explained more by the heterogeneous effects of average characteristics, whereas in Malaysia it is more due to the inequalities


in mean characteristics. These results suggest political activism may be more directly influenced by policies in Indonesia and Thailand than in Malaysia.

Relative to the three countries, the Philippines and Singapore each appear distinct. In contrast to the Indonesian, Thai, or Malaysian middle classes, the Filipino middle class seem unremarkable. Its rates of participation in all election-related activities or in most of the other political activities are not different from any other class. Where it diverges is only in terms of participating with others, showing an even lower proportion than the lower class. Even its high voting rate does not distinguish it from other classes, although the overall voting rates among the poor and, perhaps, even of the middle class in the Philippines could be driven by clientilist relations. That is, perhaps a large segment of the population votes for politicians in exchange for money or personal favors.12

In Singapore, barely half of the sample reported to have voted in the past elections. Moreover, the level of engagement in other political activities of the middle class is generally lower than either the lower class or the upper class. Their lower participation rates could imply that they are resigned to any attempt to change or reform the government. Alternatively, it could mean that they are “co-opted” through the government programs (like public housing, for example). Investigating the issue further will have implications on the claim made that Singapore follows a different Asian style of democracy.13

Our results suggest two directions for future research. One direction is to apply the same method on the same or similar data but to explore other interclass differences in their political or social orientations or activities. For example, it may be investigated if the middle class eschews traditional values and adopts modern values more than the lower class or upper class. The results of this investigation will shed light on the role of the middle classes in the democratization of Asian countries. The other direction is extending the decomposition method to identify the critical characteristics that matter more and in developing methods to identify the cognitive, social, or contextual factors that lead to heterogeneous effects.

University of the Philippines School of Economics

The author gratefully acknowledges the Freiburg Institute for Advanced Studies - Albert Lüdwig University, which hosted the author and funded initial work on this research, with support from the People Programme (Marie Curie Actions) of the European Union’s Seventh Framework Programme (FP7/2007-2013) under REA grant agreement no. [609305]. He also appreciates the assistance and the data from the Asian Barometer Project Office. The views expressed in this paper are the author’s own.

12 Thanks to Prof. Corina Gochoco-Bautista for this interpretation of the results.13 In another study [Capuno 2015], the Singaporean middle class, however, appears more pro-democracy in their views and values than other socioeconomic classes.


References

Acemoglu, D. and J.A. Robinson [2006] Economic origins of dictatorships and democracy. Cambridge, uk: Cambridge University Press.

Asian Development Bank (adb) [2010] “The rise of Asia’s middle class”, in Key indicators for Asia and the Pacific 2010. Mandaluyong City, Philippines: Asian Development Bank.

Banerjee, A.V. and E. Duflo [2008] “What is middle class about the middle classes around the world?”, Journal of Economic Perspectives 22(2): 3-28.

Birdsall, N. [2015] “Does the rise of the middle class lock in good government in the developing world?”, http://www.cgdev.org/publication/does-rise-middle-class-lock-good-governemnt-developing-world. Accessed 30 April 2015.

Blinder, A.S. [1973] “Wage discrimination: reduced form and structural estimates”, Journal of Human Resources 8: 895-909.

Capuno, J.J. [2015] “An Oaxaca-decomposition analysis of popular support for democracy in Southeast Asian countries: are the middle classes different from the rest?”, draft paper. University of the Philippines School of Economics, Diliman, Quezon City.

Chang, Y., Y. Zhu, and C. Pak [2007] “Authoritarian nostalgia in Asia”, Journal of Democracy 18(3): 66-80.

Chang, Y., Y. Chu, and L. Diamond [2012] “A longitudinal and comparative analysis of citizens’ orientations toward democracy and their evaluation of the overall performance of the democratic regime in East Asia”, Asian Barometer Working Paper Series No. 53. National Taiwan University and Academia Sinica, Taipei, Taiwan.

Chun, N., R. Hasan, and M. Ulubasoglu [2011] “The role of the middle class in economic development: what do cross-country data show?”, adb Economics Working Paper Series No. 245. Mandaluyong City, Philippines: Asian Development Bank.

Dow, J.K. [2009] “Gender differences in political knowledge: distinguishing characteristics-based and returns-based differences”, Political Behavior 31: 117-136.

Embong, A.R. [2001a] Southeast Asian middle classes: prospects for social change and democratisation. Bangi, Malaysia: Penerbit Universiti Kebangsaan Malaysia.

Emerson, D. [1995] “Singapore and the ‘Asian values’ debate”, Journal of Democracy 6: 95-105.

Erikson, R. and J.H. Goldthorpe [1992] The constant flux: a study of class mobility in industrial societies. Oxford: Clarendon Press.

Filmer, D. and L. Pritchett [2001] “Estimating wealth effects without expenditure data—or tears: an application to educational enrollment in states of India”, Demography 38(1): 155-132.

Fukuyama, F.D. [1995] “Confucianism and democracy”, Journal of


Democracy 6(2): 20-33.Gerring, J., P. Bond, W. Barndt, and C. Moreno [2005] “Democracy and

growth: A historical perspective”, World Politics 57(3): 323-364.Glassman, R.M. [1995] The middle class and democracy in socio-historical

perspective. Leiden, The Netherlands: Brill Publishers. Hewison, K. [1999] “Political space in Southeast Asia: ‘Asian style’ and other

democracies”, Democratization 6(1): 224-245. Kimura, M. [2003] “The emergence of the middle classes and political change

in the Philippines”, The Developing Economies 41(2): 264-84.Lipset, S.M. [1959] “Some social requisites of democracy: economic

development and political legitimacy”, American Political Science Review 53(1): 69-105.

Loayza, N., J. Rigolini, and G. Llorente [2012] “Do middle classes bring about institutional reforms?”, Economics Letters 116: 440-444.

Moore, B. [1967] Social origins of dictatorship and democracy. Harnondsworth: Allen Lane, The Penguin Press.

Oaxaca, R. [1973] “Male-female wage differentials in urban labor markets”, International Economic Review 14: 693-709.

Robison, R. and D.S.G. Goodman [1996] The new rich in Asia: mobile phones, McDonald’s, and middle-class revolution. New York: Routledge.

Verba, S., N. Nie, and J.O. Kim [1978] Participation and political equality. New York: Comparative Politics.

Welzel, C. [2011] “The Asian values thesis revisited: evidence from the World Values Surveys”, Japanese Journal of Political Science 12(1): 1-31.

Wu, C. and F. Lee [2012] “A comparative analysis of the wealth divide and the issue of political inclusion”, Asian Barometer Working Paper Series No. 60, Asian Barometer Project Office, National Taiwan University, Taipei, Taiwan.


AN

NE

X 1

. Mea

ns a

nd s

tand

ard

dev

iati

ons

of

the

exp

lana

tory

var

iab

les

Vari

able

s

Ind

one

sia

Phi

lipp

ines

Tha

iland

Sin

gap

ore

Mal

aysi

a(N

=15

50)

(N=

1200

)(N

=15

12)

(N=

1000

)(N

=12

14)

Mea

nS

tand

ard

d

evia

tio

nM

ean

Sta

ndar

d

dev

iati

on

Mea

nS

tand

ard

d

evia

tio

nM

ean

Sta

ndar

d

dev

iati

on

Mea

nS

tand

ard

d

evia

tio

nM

ale

0.50

0.50

0.50

0.50

0.48

0.50

0.50

0.50

0.50

0.50

Age

41.8

14.1

40.9

15.4

46.9

14.7

41.5

14.5

41.4

15.2

Age

219

46.6

1293

.519

06.8

1404

.524

18.5

1466

.219

31.8

1370

1945

.113

62.9

Som

e co

llege

0.09

0.29

0.27

0.45

0.13

0.34

0.36

0.48

0.16

0.36

Em

plo

yed

0.65

0.48

0.49

0.50

0.78

0.42

0.61

0.49

0.53

0.50

Cat

holic

0.84

0.36

Bud

dhi

st0.

940.

240.

400.

49Is

lam

0.87

0.34

0.57

0.49

In u

nion

0.84

0.37

0.74

0.44

0.75

0.44

0.68

0.47

0.70

0.46

Hou

seho

ld s

ize

4.7

2.03

4.91

2.27

4.64

2.16

4.12

1.43

5.65

3.16

Urb

an0.

300.

460.

680.

470.

250.

430.

440.

50M

ega_

maj

orci

ty0.

220.

410.

840.

370.

140.

34A

cces

s in

dex

68.9

20.2

65.1

21.9

76.7

814

.998

.08

7.30

85.6

918

.12

Low

er c

lass

0.41

0.49

0.43

0.49

0.40

0.49

0.40

0.49

0.40

0.49

Mid

dle

cla

ss0.

390.

490.

390.

490.

390.

490.

410.

490.

400.

49U

pp

er c

lass

0.19

0.40

0.19

0.39

0.19

0.39

0.19

0.39

0.20

0.40

The effect of trade policy on firm productivity in Thai manufacturing

Archanun Kohpaiboon* and Juthathip Jongwanich*

The paper examines the effect of trade policy on firm productivity using two recent industrial censuses of Thai manufacturing (i.e., 2006 and 2011). Trade policy and global participation are treated as two different variables in our analysis. Controlling for firms’ global participation, which is defined as export-sale ratio and the extent to which raw materials are imported, our study finds that trade liberalization could induce firms to commit to activities that improve productivity. The effective rate of protection, where output and input tariffs are taken into consideration together, matters in improving firm productivity. Thus, it would be risky to continue tariff reform by focusing solely on a reduction in input tariffs while leaving output tariffs untouched. In fact, both input and output tariffs must be taken into consideration to neutralize incentives in trade policy reform.

JEL classification: F10, N75 Keywords: trade policy, Thai manufacturing, effective rate of protection, productivity

and developing countries

1. Introduction

While tariffs in general went down substantially in the past two decades after the establishment of the World Trade Organization in 1995, much remains to be done in developing countries. In particular, exceptions of liberalization schedules (i.e., tariff peaks) are often found together with an escalating tariff structure, where tariff rates are escalating from raw materials to finished products. This results in nominal protection underestimating actual/effective protection. It is done with a hope that maintaining such cross-border measures would give more time for firms to improve their international competitiveness and survive in the more intense competitive environment.


PRE

144 Kohpaiboon and Jongwanich: The effect of trade policy on firm productivity in Thai manufacturing

Whether the hope materializes or not is an empirical research question. Until the new millennium, there were a number of empirical studies examining the relationship between trade liberalization and firm performance. These studies can be categorized into two main groups. The first examines the relationship between trade openness and output growth through cross-countries econometric analyses at the aggregate level.1 The second is based on case studies of specific countries.2 The main finding is in favor of the hypothesis that trade protection retards firm performance and, eventually, medium- to long-term growth.

However, both groups of studies are subject to shortcomings. In the first group, the link between trade policy and firm performance is weak and, to a certain extent, treated as a black box. They are subject to serious econometric problems in terms of endogeneity and measurement errors. As argued in Levine and Renelt [1992] and Sala-i-Martin [1997], cross-country econometric analyses are tenuous at best. In the second group, there are policy insights, but their case studies include only a handful of countries and their analytical tools vary significantly across studies. This makes it difficult to generalize the findings of these studies.

From the late 1990s, the proliferation of plant-level data available in many countries allows researchers to re-visit and mitigate some of the above shortcomings. Where the effect of trade policy on firm performance is concerned, there are at least two aspects relevant in policy circles of developing countries.

The first aspect is related to self-selection hypothesis.3 The positive relationship found between trade liberalization and performance could be due to a self-selection process, in which firms that enter export markets are already more productive than non-exporters before they ever actually begin to export. The self-selection hypothesis suggests that the nature of trade policy and firm market orientation must be treated as two separate explanatory variables in the analysis. While both could be important to productivity, they are two different things. The former refers to the policy environment, while the latter is a firm decision.

The second aspect of research in this field is to examine channels through which trade policy affects firm performance (Amiti and Konings [2007]; Melitz and Redding [2012]). In particular, Amiti and Konings [2007] argue possible different effects of input and output tariffs due to different operating channels. Lower output tariffs can increase productivity by inducing tougher competition, whereas cheaper imported inputs can raise productivity via learning, variety, and quality effects.

1 See the literature reviews of this group of studies in Edwards [1993]. 2 Two influential works include the Bhagwati-Krueger project for the National Bureau of Economic Research in the 1970s and the Papageorgious-Michalely-Choksi study for the World Bank in the 1980s.3 See the literature review in Lopez [2005].


However, a number of case studies4 point to the fact that introduction of new technologies does not guarantee that productivity increases instantaneously. Most likely, a long process of learning and mastery of skills may be required to reach high levels of productivity. Actually, a long-term commitment and real resources are required to have substantial effects on productivity. This can be influenced by the policy environment such as trade policy. As argued in these case studies, firms receiving the effective rate of protection (erp) tend to be “unresponsive to improved technological capability”. It does not matter whether effective protection is mainly driven by output or input tariffs. Hence, a decomposition of the erp into output and input tariffs might not be appropriate.

Against this backdrop, the paper aims to examine the determinants of firm productivity with emphasis on the effects of trade policy. The 2006 and 2011 industrial censuses are used in our analysis. Our proposed study has at least three contributions to the existing literature: • First, we carefully distinguish the possible effects between trade (export and

import) and trade policy (e.g. cross-border protection) on productivity. The latter has important policy implications for trade policy reform.

• Second, we examine whether the effect of trade policy varies across firm types by introducing the interaction term between firms’ specific and trade policy variable. The former focuses market orientation (whether firms participate in the global market) and input sourcing (whether firms import raw materials or intermediates from abroad).

• Third, the effect of output and input tariffs are re-examined as opposed to a case when using the erp (both input and output tariffs combined). Whether the effects of output tariffs are actually less than that of input tariffs could have a substantial policy impact as policymakers in developing countries prefer a reduction in tariffs on inputs rather than on output.

Thailand is chosen for this study since trade policy reform remains a challenging issue for policymakers. In particular, efforts to streamline tariff rates to 3 rates (0-1 percent for raw materials, 5 percent for intermediates, and 10 percent for finished products) are at best far from complete. There is almost a fifth of tariff lines subject to a 20 percent tariff rate or higher. In addition, by design, the tariff structure in Thailand is cascading, so that nominal protection tends to underestimate the effective rate. Hence, effective protection seems to vary across industries. This allows us to test the effect of protection on firm productivity.

The remainder of this paper is organized as follows. Section 2 presents the analytical framework. Section 3 briefly discusses trade policy and firm

4 See the cases of Keesing [1983], Keesing and Lall [1992], Westphal et al. [1979, 1984], Aw and Batra [1998], Wortzel and Wortzel [1981], Hobday [1995], Pietrobelli [1998], Pack and Saggi [1997], and Nelson and Pack [1999].


productivity in Thai manufacturing. Section 4 discusses the model, while section 5 presents the data set used in this study. Section 6 empirically assesses the effects of trade policy on firm productivity. The final section provides some conclusions and policy implications.

2. The analytical framework

Gains from trade in terms of output growth/productivity have been studied for several decades. Beginning with the standard neoclassical trade model, Ricardian comparative advantage model, and/or Hecksher-Ohlin-based comparative advantage, gains from trade are derived from resource reallocation from sectors the country has a comparative disadvantage in to those sectors in which it has comparative advantage. A country’s comparative advantage is driven by technology, or resource endowment, or both. Note that under these model settings, trade could lead to output expansion, but it does not have an impact on rate of economic growth. In the 1980s, the source of gains from trade shifted to intra-industry trade. In this model setting, consumers enjoy a variety of products that are close but are not perfect substitutes. As a result, the market is fragmented in response to consumer wants and needs, and differentiated goods are produced under increasing return to scale and traded (Krugman [1979]; Helpman and Krugman [1985]). Fragmentation of the market results in a struggle by firms to attain adequate production volumes to cover fixed costs so that trade enlarges market size.

Note that all the models mentioned above are still based on a representative (homogenous) firm. They cannot explain well why in a given industry only some and not all firms export [Greenaway et al. 2004]. So the assumption of a representative firm is relaxed and the literature on firm heterogeneity has grown (Melitz [2003]; Bernard et al. [2003]). The firm heterogeneity literature argues that even within a narrowly defined industry, some firms are much larger, more productive, and more profitable than the others [Melitz and Trefler 2012]. International trade makes better-performing firms expand their products into larger markets, while resources are re-allocated from less productive firms into productive ones. This, therefore, leads to the improvement of industry efficiency.

Gains from trade are also examined in the economic growth literature. For example, in extensions of neoclassical models, including the Solow-Swan model and the Ramsey growth (optimal-saving) model, trade liberalization increases output level in the economy but not the rate of growth. The rate of growth in these models depends on growth of input and the rate of technological progress (Baldwin [1992]; Srinivasan and Bhagwati [1980]). As argued in the Harrod-Domar model, trade could generate positive growth effects under a circumstance, where the marginal product of capital is bounded by some positive number (Srinivasan [1999]; Lopez [2005]).


The effect of international trade on productivity and on economic growth has been highlighted in the new or endogenous growth theory (Young [1991]; Rivera-Batiz and Romer [1991]; Pack [1994]). Nonetheless, the effect of trade liberalization on productivity and growth in developing countries is still unclear. On the one hand, Rivera-Batiz and Romer [1991] argue that international integration helps to improve technological progress by expanding the size of the market, which helps to expand innovation activity. In addition, cross-border technological spillovers could be created from integration so that economic growth/productivity is promoted. This mechanism occurs in both developing and developed countries. On the other hand, Young [1991] argues that the positive effect of trade on growth would be far less in developing countries. While trade could positively affect labor productivity through learning by doing process, the gains would be larger for developed countries. These countries produce and trade products that still need learning by doing activity. But the products that are usually manufactured in developing countries require less skill and are not much involved with the learning process.

During the 1990s, studies examining the effect of trade on output growth were based on the endogenous growth framework. They can be categorized into two main groups. The first examines the relationship between trade openness and output growth through cross-countries econometric analyses at the aggregate level. The second is based on case studies of specific countries.

In the first group, (real) output growth, especially that at the aggregate level, is used to represent firm performance alongside different measures of trade openness (Edwards [1993]; Sachs and Warner [1995]; Lopez [2005]). In some studies, trade measures are instrumented to redress a possible endogeneity problem (Frankel and Romer [1999]; Alcala and Ciccone [2004]; Noguer and Siscart [2005]).5 Most of the empirical studies find a positive relationship between trade openness and economic growth.

The second group analyzes the relationship based on case studies of specific countries. Either a survey or an econometric analysis at the aggregate level is used (World Bank [1991]; Kohpaiboon [2003]) . As in the cross-countries analysis, most studies find a positive relationship between trade openness and growth. However, both groups are subject to shortcomings, which make the link between trade policy and firm performance remain a black box to a certain extent.6 In

5 For example, Frankel and Romer [1999], Rodriguez and Rodrik [2001], and Irwin and Tervio [2002] use geography as an instrumental variable in examining the relationship between trade liberalization and economic growth.6 Note that some scholars (Edwards [1993]; Srinivasan and Bhagwati [2001] ) argue that case studies such as those done by the National Bureau of Economic Research [1978] or World Bank [1991] could provide good evidence about the effect of trade on growth. However, drawing conclusions about the role of trade from case studies is still difficult, since they include only a handful of countries while differences in firm/plant characteristics are not taken into account [Lopez 2005].


particular, by using aggregate data either cross-country or within a country, heterogeneity at the firm and industry levels is not taken into account. This tends to mix countries and industries with very different characteristics (Berry [1992]; Tybout [1996]; Lopez [2005]). As mentioned earlier, allocation within an industry is crucial in generating efficiency within an industry and in contributing to a country’s economic growth.

As plant/firm-level panel data become increasingly available in many countries, researchers have re-visited the trade-growth nexus. Shortcomings, especially econometric problems, were better addressed. Using the micro data set, a new hypothesis is formed relating to the effects of trade liberalization on firms’ performance. The positive relationship found between trade liberalization and performance is explained by two competing theses.

First is the self-selection thesis, in which firms entering the export markets are already more productive than non-exporters. Hence the positive relationship would not be directly related to trade liberalization. Also, the nature of trade policy and firms’ market orientation must be treated as two separate explanatory variables in the analysis. They are two different things. The former refers to the policy environment, whereas the latter is a firm’s decision.

The second thesis is the learning-by-exporting thesis. In this thesis, firms participating in foreign markets are more likely to experience productivity gains as opposed to non-exporters, as the former would receive new information about technological progress, product designs, and quality of goods from their foreign exposure (Grossman and Helpman [1991]; Aw and Hwang [1995]; Bernard and Wagner [1997]).7

When policy implications are concerned, both these groups of studies argue in favor of liberalization of trading partners instead of home countries. Reducing trade barriers between or among trading partners and policies affecting the profitability of being an exporter in the home country would induce more productivity improvement in the home country. Interestingly, Melitz and Trefler [2012] argue for another possible channel where liberalization in the home country could yield productivity gains that may help exporters to import inputs/technology at cheaper costs, which induce higher productivity improvements. Trefler [2004] and Lileeva and Trefler [2010] also argue that liberalization at home could improve the country’s overall productivity through better resource reallocation.

Empirical studies (Amiti and Konings [2007]; Melitz and Redding [2012])further analyze the channels through which trade policy affects firm performance. In particular, Amiti and Konings [2007] argue possible different effects of input and output tariffs due to differences in operating channels. Two possible

7 See possible explanations of self-selection—for example, sunk costs, imported technology, and increased R&D—in Keesing and Lall [1992], Bernard and Jensen [2004], and Lopez [2005].


explanations are provided. First, it occurs when locally manufactured and imported intermediates are not close substitutes. Any change in input tariffs would have a significant effect only on the firms that actually import them. For those who use locally manufactured intermediates, such change would not have any significant effects on their behavior. Second, as shown in the firm heterogeneity literature, switching market orientation between the domestic market and exports is costly. Hence, changes in output tariffs might not have any significant impact on those who already export. They just continue in business due to the presence of sunk and fixed costs within the export business.

Both circumstances above are often observed in a country that is long engaged with the world and is pursuing a dual-objective trade policy, where being reluctant to lower tariffs is associated with the introduction of input tariff exemption schemes to promote export-oriented activities. Interestingly, based on Indonesian plant level data from 1991 to 2001, Amiti and Konings [2007] show that the effects of input tariffs are more favorable than that of output tariffs in generating firms’ productivity improvement. However, some studies8 argue that firms’ productivity improvement is a long process of learning and mastery of skills so that both input and output tariffs facing by a firm should be considered together in affecting firms’ performance.

3. Trade policy and firm productivity in Thai manufacturing

In Thailand, a tariff is the core measure in conducting trade policy. Non-tariff measures have been occasionally used in a narrow range of products, mainly in certain sensitive agricultural products such as soybean, palm seed, silk, and milk. Like other developing countries, a high tariff level associated with an escalating tariff structure was used to promote industrialization from the 1960s to the mid-1980s. From 1983 to 1995, tariff levels remained virtually unchanged with few exceptions, whereas the effort to promote Thailand as an export platform for multinationals was done through the introduction of various tariff exemption schemes (Kohpaiboon [2006]; Kohpaiboon and Jongwanich [2007]).

As part of its commitments under the World Trade Organization, a comprehensive plan for tariff reduction and rationalization was proposed in 1990 and implemented in 1995 and 1997. Maximum tariffs were reduced from 100 percent to 30 percent. Tariffs were significantly lowered on some 4,000 items (at the 6-digit HS level) or 75 percent of total tariff lines. By the end of the 1990s, the tariff bands were reduced from 39 to six (0, 1, 5, 10, 20, and 30 percent). Nonetheless, there were numerous exceptions whose tariff rates exceed 30 percent. Tariff restructuring has received renewed emphasis as an essential part of the overall economic reforms aimed at strengthening efficiency and

8 See details in footnote 6.


competitiveness (Warr [2000]; World Trade Organization [1999]) . The Thai government introduced tariff cuts, commencing in June 2003 (implemented in October 2003), followed by a four-year period of tariff reduction from 2004 to 2008. In 2010, there are around 900 items involved in tariff reduction process, covering a wide range of manufacturing intermediates such as rubber and articles thereof (HS40), glass and glassware (HS70), knitted fabrics (HS60), other base metals (HS81), woven fabrics (HS58), articles of stone (HS68), man-made staple fiber (HS55), wadding yarns (HS56), cotton (HS52), and miscellaneous vegetable preparations (HS21) [Kohpaiboon and Jongwanich 2007].

Table 1 presents the average of most-favored-nation tariffs of Thailand and selected Asian economies in 2010. It is clear that average tariffs in Thailand are relatively high compared with other middle-income countries in the region. Interestingly, the weighted average was lower than the unweighted one, implying that tariffs imposed on certain products are redundant. In general, agricultural products are subject to higher tariffs than manufacturing products.

TABLE 1. Weighted average of most-favored-nation tariff rate of RCEP members during 2010-2012

Country (year) Unweighted Weighted Agricultural products

Non-agricultural products

Thailand (2011) 8.7 5.0 9.0 4.9

Viet Nam (2010) 9.8 12.2 24.4 10.7

Singapore (2011) 0 0 0 0

Philippines (2011) 6.2 12.2 23.2 10.4

Myanmar (2011) 5.6 6.6 12.6 4.9

Malaysia (2012) 5.3 6.7 8.7 6.5

Indonesia (2012) 6.6 9.8 1.8 11.1

Laos (2008) 9.7 13.6 19.3 12.6

Brunei (2011) 2.5 1.7 0 2.6

Cambodia (2012) 10.9 12.0 14.7 11.1

Australia (2011) 2.8 3.8 1.6 3.9

New Zealand (2011) 2.0 2.7 1.9 2.8

China (2010) 9.9 8.6 21.5 7.4

India (2012) 13.3 9.4 48.6 7.7

Japan (2011) 3.0 2.1 7.0 1.3

South Korea (2011) 11.2 9.6 34.1 5.6

Source: Authors’ calculations using most-favored-nation tariff rates from the World Trade Organization

Table 2 presents the distribution of tariff lines in Thailand over the past two decades. Clearly, the distribution changed as a result of the comprehensive tariff reform in the mid-1990s. During the pre-1997 period, more than a quarter of total HS6 tariff lines had tariff rates in the category of more than 30 percent. The bracket of 15-30 percent also accounted for 30 percent of total tariff lines. After 1997, there was a dramatic shift of tariff lines to lower brackets. For example, more than 50 percent of tariff lines are in the 0-10 percent bracket. This is followed


by 3.9 percent in the 10.1-15 percent bracket and 21.4 percent in the 15.1-20 percent brackets. The share of tariffs above 20 percent dropped from 40 percent in the pre-1997 period to around 20 percent during the post-1997 period. The 2003 tariff reduction plan marginally changed the distribution of tariff lines. Tariff reductions in 2003 have basically involved shifting the tariff lines from the 16-20 percent bracket to a lower bracket, with little impact on those belonging to the above 20-percent brackets. The proposed changes from the next two years seem to follow the same pattern, while changes proposed for 2006-2008 seem negligible.

TABLE 2. Share of 4-digit HS categories of applied tariff rates in Thailand, 1989–2008

Tariff bands 1989 1995 2002 2003 2004 to 2008

0 2.5 2.6 5.6 5.7 6.0

0.1 - 5 14.4 17.3 33.3 37.7 48.8

5.1 - 10 14.2 17.6 14.1 14.2 14.8

10.1 - 15 12.7 3.2 3.9 4.5 3.6

15.1 - 20 15.4 16.4 21.4 17.9 8.4

20.1 - 30 15.8 16 13.8 14.3 12.7

30 - 100 25 26.8 7.8 5.8 5.7

Source: Data for 1989 and 1995 from World Trade Organization. Data for 2002 to 2008 are from the authors’ compilation from Official Document provided by Ministry of Finance.

Table 3 presents statistical indicators of how firms engage globalization, i.e., export-sales ratio and ratio of raw material imports to total materials used, both measured in percent. They are reported in terms of average, maximum, and minimum. They are classified into 5 categories across the erp figures. Patterns observed in Table 3 tend to be in line with the theoretical postulations of the firm heterogeneity literature. Engaging exports incurs fixed and sunk costs. Therefore, in a given industry, regardless the erp figure, only some firms export. There is a vast difference between maximum and minimum values in all erp categories. In theory, industries subject to high and positive erp tend to sell their products locally in order to reap economic rents induced by protection. This pattern is, to a certain extent, found in Table 3. The mean value of the export-sales ratio of firms located in industries with around zero erp and negative erp is higher than those experiencing positive erp. Interestingly, such a pattern is not clearly observed when using a raw material import criterion. The mean value of percent of raw material imports to total used swings up and down across the erp categories. Industries subject to around zero erp exhibit the highest raw material import ratio, followed by those in highly positive erp. Firms in highly negative erp and moderate positive erp have the same figure of raw material import ratio. The unclear pattern is due to the extent to which domestic and imported raw materials are substituted varies across industries instead of protection.


TABLE 3. Market orientation and raw material sourcing behavior of Thai manufacturing firms in 2011

Export-sale ratio (percent of total sales) Mean Max Min

Highly Negative ERP (<-10%) 9.1 88.8 0.0

Moderately Negative ERP (-2% to 10%) 8.3 85.5 0.0

Around Zero ERP (-2% to 2%) 9.0 81.3 0.0

Moderate Positive ERP (2% to 10%) 7.1 79.6 0.0

Highly Positive ERP (>10%) 8.8 85.9 0.0

Raw material import as a percentage of total raw materials used

Highly Negative ERP (<-10%) 8.6 90.3 0.0

Moderately Negative ERP (-2% to 10%) 6.8 81.3 0.0

Around Zero ERP (-2% to 2%) 9.5 87.9 0.0

Moderate Positive ERP (2% to 10%) 8.6 86.4 0.0

Highly Positive ERP (>10%) 9.3 83.6 0.3

Source: Authors’ compilation from the 2011 industrial census

4. The model

The model used here starts with trans-log production function of the firm. The plant’s value added is a function of two primary inputs (labor and capital), their squared terms, and their interaction. Labor is further disaggregated into production (PL

ij) and nonproduction (NL

ij) workers to capture their difference in

contributing to firm productivity. Blue-collar workers are regarded as the former and white-collar ones are the latter. Over and above, a set of firm- and industry-specifics as well as the trade policy variables are included as controlling variables as expressed in Equation 1.

ln VAij = β0 + β1 ln K

ij + β2 (ln K

ij)2 + β3 ln PL

ij + β4 ln NL

ij + β5 (ln PL

ij)2

+ β6 (ln NLij)2 + β7 ln PL

ij * β8 ln NL

ij * ln K

ij + γ1 FS

ij + γ2 ISij

+ γ3 tradepolicyj + ε

ij (1)

where VAij = Value added of firm i in industry j

Kij = Capital used by firm i in industry j

PLij = Production workers employed by firm i in industry j

NLij = Non-production workers employed by firm i in industry j

FSij = Firm-specific characteristics of firm i in industry j

ISij = Industry-specific characteristics of industry j

tradepolicyij = The nature of trade policy of industry j


Firm-specific characteristics (FSij) include market orientation, foreign

ownership (ownij), and R&D investment (RD

ij). In this study, two aspects of

market orientation are captured: the export-sales ratio; and how much finished products are exported as a percentage of total sales (mktij). Both are introduced in the model. International competition would make firms alert to any productivity improvement and eventually enhance firm productivity so that the coefficient associated with mktij is expected to be positive. The second aspect of market orientation is the extent to which imported raw materials are used as a percentage of total raw materials (rawm

ij). Firms that import raw materials would benefit from

technology embodied in them, thus improving their productivity. The coefficient associated rawmij is also expected to be positive.

The consensus in the foreign direct investment literature [Caves 2009] suggests that foreign firms (ownij) are generally more productive than indigenous counterparts so own

ij is expected to be positive. own

ij is measured by firms’ foreign

equity (percent) share. RDij, measured here by the firm’s research, planning, and

development expenditure to total sales, would raise firm productivity so that the coefficient associated is expected to be positive.

Three industry-specific factors are controlled in our analysis. The first is the extent to which an industry engages into a global production network. This can have an implication on productivity as reviewed in part 2, under gains from intra-industry trade. Ideally, details at the firm level (e.g., whether firms are actually engaged in multinational enterprises' (mnes) production sharing and whether they import tailor-made raw materials for specific customers, etc.) are needed. Unfortunately, such details at the firm level are not available within the Thai dataset. To overcome the unavailability of perfect measures of global production sharing, two alternative proxies are used in this study. The first two proxies are shares of parts and component in total imports (gpn1

j) and total trade (gpn2

j) as

reflected in Equations 3 and 4:

(gpn1j) = P&C Imports

j / Total Imports

j (3)

(gpn2j) = P&C trade (import + export) / Total Trade (4)

The higher the share, the more important the global production sharing is to the industry. The parts list is the result of a careful disaggregation of trade data based on the Revision 3 of the Standard International Trade Classification extracted from the un Comtrade database. It is important to note that the un Comtrade database does not provide for the construction of data series covering the entire range of fragmentation-based trade. The parts list used here is from that developed in Athukorala and Kohpaiboon [2009].9 To convert Standard International Trade

9 The use of lists of parts in the Board Economics Classification 42 and 53 is a point of departure. Note


Classification to International Standard Industrial Classification (isic) , standard concordance is applied.

The second industry-specific factor is producer concentration (CRj). Its effect

on productivity is ambiguous. Industries with high barriers to entry are likely to be concentrated and are often capital- and/or skill-intensive. Hence, this could make firms less responsive to any technological improvement so it negatively affects productivity (negative sign). On the other hand, as argued in the well-known creative destruction thesis by Schumpeter, a highly concentrated industry would give firms incentive to innovate. If so, the coefficient associated with producer concentration could be positive. Producer concentration is measured by the sum of the sales share of the top-4 firms in total.

tradepolicyj

is introduced to examine the study’s main hypothesis. Two alternatives of trade policy are used in this study. Effective rate of protection (ERP

j) is used as the first measure of trade policy due to its theoretical superiority.

In theory, policy-induced incentives from cross-border protection measures like tariffs would not be different, regardless of the incentives generated by either input or output tariffs. Hence, it would be theoretically superior to employ an effective rate of protection (ERP), instead of separating input and output tariffs.

ERPj = (t

j – �a

kj tk) / (1 – �a

kj) (6)

where t

j = Tariff on outputs on industry j

tk = Tariff on inputs k

akj

= A value share of inputs k used in finished products on industry j

To examine whether the effect of trade policy varies across firms, the interaction term between firm specific and trade policy variable is introduced; ERP

j*mkt

ij and ERP

j *rawn

ij are introduced. The former implies that giving

protection to an industry of interest, the effects could vary according to the extent to which firms export their products to the world. Similarly, in the latter, the effect of protection on firm productivity could depend on how much a firm is integrated globally through importing raw materials and intermediates. In addition, the interaction term between the erp and ownership (ERP

j * own

ij) is

introduced because a foreign firm might behave differently under different trade

that the parts in Board Economics Classification 211 are not included as they are primary products that are usually classified as traditional, rather than fragmented-intermediates. The additional lists of parts are included based on firm interviews reported in Kohpaiboon [2010]. Data on trade in parts are separately listed under the commodity classes of machinery and transport equipment (Standard International Trade Classification 7) and miscellaneous manufacturing (Standard International Trade Classification 8) and are based on firm interviews elaborated in Kohpaiboon [2010]. The list of parts and components is available on request.

n

k = 1

n

k = 1


policy environments (known as Bhagwati’s hypothesis).10 Trade liberalization could provide an incentive for foreign firms to behave productively. By contrast, rent-seeking behavior of foreign firms, which is more likely to occur under trade restriction, could retard overall productivity improvement.

As discussed in Section 2, input and output tariffs could have different impacts on firm productivity improvement, hence, inputtariff

i and outputtariff

i, are

separately introduced as alternative measures of trade policy here. That is,

ERPj + ERP

j * mkt

ij + ERP

j * rawn

ij ;

�tj – �a

kj tk�

+ �t

j – �a

kj tk�

* mkt

ij + �t

j – �a

kj tk�* rawn

ij (7)

where outputtariffj = Tariff on outputs of industry j (t

j)

inputtariffj = The weighted average of input tariff from k = 1,…,n. ��a

kj tk�

Note that to mitigate any possible endogeneity problem from these industry-specific factors, all of them are lagged.

All in all, the empirical model to be estimated is as follows:

ln VAij = β0 + β1 ln K

ij + β2 (ln K

ij)2 + β3 ln PL

ij + β4 ln NL

ij + β5 (ln PL

ij)2

+ β6 (ln NLij)2 + β7 ln PL

ij * β8 ln NL

ij * ln K

ij + γ1 own

ij + γ2 R&D

ij

+ γ3 rawnij + γ4 mkt

ij + λ1 CR

j, t–j + λ2 gpnj, t–j

+ λ4 tradepolicyj, t–j

tradepolicyj * mkt

ij + φ2tradepolicy

j, t–j * rawn

ij + ε

ij (8)

where ln VAij = Value added of firm i in industry j (in natural log)

ln Kij = Capital used by firm i in industry j (in natural log)

PLij = Production workers employed by firm i in industry j

NLij = Non-production workers employed by firm i in industry j

tradepolicyj, t–j

= Lag variable of trade policy measured alternatively by 1. Effective rate of protection (erp) 2.Outputtariff

i (t

j) and Inputtariff

i (t

j) ��a

kj tk�

mktij = Market orientation of firm i of industry j measured by a percentage

of export to total sales rawm

ij = Input sourcing of firm i of industry j measured by a percentage of

imported raw materials and intermediates to total inputs own

ij = Ownership of firm i of industry j measured by a share of

foreign owners in total capital

10 See the discussion in Kohpaiboon [2006].

n

k = 1

n

k = 1

n

k = 1

n

k = 1

n

k = 1


R&Dij = A share of R&D expenditure as a percent of total sales of firm i of

industry j CR

j, t–j = Producer concentration ratio of industry j at time t-j

gpnj, t–j

= The degree of industry involved in the global production networks of industry j at time t-j, measured by two alternatives:gpn1

j, t–j = The share of parts and components imports to total import

at the 4-digit isic gpn2

j, t–j = The share of parts and components trade (export+import)

to total trade at the 4-digit isic

5. Data set and cleaning procedure

The data set suitable for the current purpose is a long-panel data of establishments in Thai manufacturing, covering before and after major trade reform. Unfortunately, such a data set is not available in the country. So far Thailand has three industrial censuses—1996, 2006, and 2011—all of which are cross-sectional in nature. These three censuses are not able to formulate as a panel data set as the identification number used in each census is assigned differently. In particular, a given identification number of two different censuses does not necessarily refer to the same firm.

The latest census (2011) contains 98,482 observations. Out of the total, 71,387 observations are self-employed (zero record of paid workers) or micro-enterprises (less than or equal to 10 workers). Given the current research focus, we exclude these self-employed and micro-enterprises. Hence, the remaining observations number 27,095. Similar to what occurred in the censuses in 1996 and 2006, there are many duplicate samples in which at least two observations report the same value in most of variables. To identify the duplicated observations, the criterion is if samples report identical values of 7 key variables, they are treated as duplicated samples. The 7 key variables include total workers, female workers, initial fixed asset, ending fixed asset, registered capital, sale value, and input values. In this case, we count only one firm. According to this criterion, there are 4,418 duplicated samples to be removed. The remaining observations number 22,677.

Next, we drop observations reporting unrealistic values of the key variables. They include negative value added, low value added (less than 10,000 baht), and low fixed assets (less than 10,000 baht). Finally, 8 industries that either serve niches in the domestic market (e.g. processing of nuclear fuel, manufacture of weapons and ammunition), in the service sector (e.g. building and repairing of ships, manufacture of aircraft and spacecraft, and recycling), or are explicitly reserved for local enterprises (e.g. manufacture of ovens, furnaces and furnace burners, manufacture of coke oven products) are excluded. All in all, 13,593 observations remain. Summary statistics and correlation of variables are shown in Tables 4 and 5.


TABLE 4. Data summary statistics

Variables Mean Std. Dev. Min Max

VAij 15.88 2.40 9.21 25.26

PLij 0.99 1.39 0 8.88

NLij 3.67 1.13 0 9.50

Kij 15.84 2.32 9.21 26.32

ownij 4.19 17.21 0 100

mktij 7.430 21.90 0 100

rawmij 6.27 18.53 0 100

R&Dij -11.87 6.85 -13.82 20.51

ERPj 0.05 0.17 -0.58 0.60

cr4j,t-j 0.45 0.09 0.32 0.65

GPN1 j,t-j 0.04 0.12 0 1

GPN2 j,t-j 0.03 0.11 0 1

Note: All variables are in logarithm, with the exception of ownership, market-oriented, imported raw material, trade policy, concentration ratio, and production network.Source: Authors’ calculations

6. Results

Initially, the equations are estimated using the ordinary least squares method while paying attention to the possible presence of heterogeneity and outliers. Due to the nature of cross-sectional data, it is likely that outliers could have an impact on the estimated parameters and lead to misleading inferences. Therefore, careful treatment of outliers is needed. Cook’s Distance13 is used to identify suspected outliers. The intra-class correlation or the clustered data, based on industry level, is tested (Table 6). The results show a low level of the correlation (0.267).

Tables 7 and 8 present estimation results where trade policy is measured by the erp and tariffs of output and inputs are separately introduced. Column A in both tables is based on gpn1, whereas Column B is based on gpn2. The overall results from both tables are largely similar. The estimation results are not sensitive to choices of gpn. Hence, the following result interpretation will be discussed, based on these two tables. Coefficients corresponding to the interaction term between nonproduction workers and capital as well as the squared terms of two types of workers are statistically significant, suggesting that the trans-log production function fits the data well, relative to the more restrictive Cobb-Douglas one. The statistical difference of coefficients associated with production and nonproduction workers supports the hypothesis that quality of labor matters in determining firm productivity. The higher the number of white collar workers employed by firms, the greater the productivity improvement expected, all other things remaining constant.


TAB

LE 5

. Co

rrel

atio

n m

atri

x

Vari

able

sVA

ijPL i

jNL i

jK

ijow

nij

mkt

ijrawm

ijR&D

ijERP

jO

utp

ut

tari

ffIn

put

ta

riff

cr4 j,t

-jGPN1

j,t-j

GP

N2 j,

t-j

VAij

1.00

PL i

j0.

611.

00

NL i

j0.

540.

461.

00

Kij

0.81

0.57

0.49

1.00

own

ij0.

230.

220.

230.

211.

00

mkt

ij0.

290.

330.

240.

250.

371.

00

rawm

ij0.

250.

220.

230.

230.

340.

361.

00

R&D

ij0.

230.

200.

190.

210.

050.

110.

111.

00

ERP

j-0

.01

0.00

30.

01-0

.01

0.08

0.02

0.05

-0.0

21.

00

Out

put

tar

iff0.

080.

060.

070.

080.

070.

090.

040.

020.

731.

00

Inp

ut t

ariff

0.14

0.10

0.10

0.12

0.04

0.06

0.00

0.03

-0.1

90.

321.

00

cr4 j,t

-j0.

02-0

.01

0.01

0.06

-0.0

040.

020.

02-0

.02

0.06

0.04

-0.0

51.

00

GPN1

j,t-j

0.14

0.12

0.10

0.10

0.11

0.08

0.07

0.03

0.14

0.17

0.24

0.13

1.00

GP

N2 j,

t-j

0.13

0.13

0.09

0.10

0.11

0.05

0.08

0.03

0.20

0.21

0.24

0.14

0.92

1.00

Sou

rce:

Aut

hors

’ cal

cula

tions


TABLE 6. Intra-group correlation

Number of obs = 13593R-squared = 0.26

Source SS Df MS F Prob>FBetween isic_obs 22120.752 60 368.679 80.13 0.000Within isic_obs 62257.048 13532 4.601Total 84377.8 13592 6.21

Intra-class correlation Asy. S.E. 95% Confidence Interval

0.267 0.056 0.16 0.37

Estimated SD of isic_obs effect 1.29

Estimated SD within isic_obs 2.14

Est. reliability of a isic_obs mean (evaluated at n = 217.59) 0.98

Source: Authors’ estimates

TABLE 7. Productivity determinants based on the ERP and 2011 census

Variables Column A Column BCoefficient t-statistics Coefficient t-statistics

Intercept 2.17* 4.27 2.12* 4.18PLij 2.27* 23.86 2.27* 23.85

NLij0.41* 3.51 0.41* 3.50

Kij 0.82* 12.18 0.82* 12.26PLij*Kij -0.12* -20.46 -0.12* -20.42NLij*Kij 0.0007 0.07 0.0007 0.08PLij

2 0.04* 6.36 0.04* 6.24NLij

2 0.003 0.28 0.003 0.30Kij

2 -0.003 -1.22 -0.004 -1.29ownij 0.003* 3.55 0.003* 3.65ownij* ERPj, t-j 0.001 0.18 0.001 0.18mktij 0.002* 3.89 0.002* 4.09rawmij 0.002* 3.05 0.002* 2.96R&Dij 0.013* 8.13 0.013* 8.16ERPj, t-j -0.30* -3.83 -0.29* -3.78ERPj, t-j *mktij -0.01** -1.97 -0.006** -2.13ERPj, t-j *rawmij 0.003 0.77 0.003 0.83cr4j, t-j -0.35* -3.05 -0.32* -2.79GPN1j, t-j 0.64* 6.95

GPN2j, t-j0.55* 5.08

# obs 13593 13593Ad-R 0.73 0.73F-stat 1717 (p-value = 0.00) 1711 (p-value = 0.00)

Notes: * statistically significant at 1 percent ** statistically significant at 5 percentAn increase in the ERP reflects higher trade protection.Source: Authors’ estimates


TABLE 8. Productivity determinants based on ERP decomposition and 2011 census

Variables Column A Column B

Coefficient t-statistics Coefficient t-statistics

Intercept 2.17* 4.26 2.12* 4.17

PLij 2.26* 23.74 2.26* 23.74

NLij 0.41* 3.53 0.42* 3.53

Kij 0.80* 11.86 0.80* 11.91

PLij*Kij -0.12* -20.37 -0.12* -20.34

NLij*Kij 0.00 0.00 0.00 0.01

PLij2 0.04* 6.39 0.04* 6.29

NLij2 0.004 0.39 0.004 0.39

Kij2 -0.003 -1.01 -0.003 -1.06

ownij 0.003* 4.12 0.003* 4.23

mktij 0.004* 2.53 0.004* 2.62

rawmij 0.006* 4.12 0.006* 4.05

R&Dij 0.013* 8.22 0.13* 8.25

outputtariffj -0.27 -1.02 -0.27 -1.00

inputtariffj 3.02* 4.90 3.26* 5.30

Outputtariffj*mktij -0.02* -2.16 -0.02* -2.33

Inputtariffj*mktij -0.01 -0.30 -0.007 -0.27

Outputtariffj*rawmij -0.004 -0.30 -0.003 -0.25

Inputtariffj*rawmij -0.09* -2.79 -0.09* -2.76

cr4j, t-j -0.32* -2.77 -0.29 -2.48

GPN1j, t-j 0.54* 5.78

GPN2j, t-j 0.43* 3.89

# obs 13593 13593

Ad-R 0.73 0.73

F-stat 1572 (p-value = 0.00) 1568 (p-value = 0.00)

Notes: * statistically significant at 1 percent ** statistically significant at 5 percent Source: Authors’ estimates

In both tables, the coefficients corresponding to ownij

, mkt

ij, and rawm

ij turn

out to be positive and significantly different from zero at 5 percent. This finding is in line with previous studies. That is, foreign firms tend to be more productive than indigenous ones, all other things remaining constant. Meanwhile, whether domestic or foreign, the firms that engaged in international business (either exporting their products, or importing raw materials, or both) tend to be more productive than those strictly engaged in local markets. Similarly, everything else being equal, the positive sign of R&D suggests firms spending more on R&D tend to have higher value added.


For industry-specific factors, our study finds a negative and statistically significance effect of CR4 at the 1 percent. The negative sign suggests that industries with high barriers to entry or are concentrated tend to make firms less responsive to any technological improvement. Both gpn1 and gpn2 are positive and significant at 1 percent, confirming the robustness of the finding that participating in global production networks could result in higher firm productivity improvement. This finding is consistently with Kohpaiboon and Jongwanich [2014] that Thai firms participating in global production networks go beyond simple assembly, and a statistically significant wage premium in the industries engaged in such networks is found.

Regarding the effects of trade policy, the coefficient corresponding to erp turns out to be negative and statistically significant (Table 7). All other things unchanged, firms operating under a regime of higher cross-border protection have lower productivity. In other words, protection can retard the process of productivity improvement. This finding is consistent with the findings of previous studies. The negative effect of the erp on productivity tends to be higher on exporting firms, as suggested by its statistical significance of the interaction term between the erp and mkt

ij. When the erp is decomposed into output and

input tariffs, the coefficient associated with Outputtariffi attains the theoretically

expected sign, but it is not statistically significant (Table 8). The positive and statistical significance of Inputtariff

i must be interpreted with

care. First, this finding is in line with that in Table 7, i.e., we found a negative effect of protection on productivity as discussed above. Consider the erp formula expressed in Equation 6 above The negative coefficient associated with the erp will result in a negative coefficient on output tariffs and a positive coefficient on input tariffs. This suggests that for a given level of output tariffs, lowering input tariffs would simply increase effective protection to producers. Thus, the productivity of firms would decline, when output tariff maintained at the same level. This finding would be highly relevant to policymakers in developing countries, where policymakers emphasize input tariff reduction while expressing reluctance to lower output tariffs.

As revealed intensively in a number of case studies as discussed above, the greater protection granted producers makes them unresponsive to any productivity improvement activities, including the long process of learning and mastery of skills, which require long-term commitment and real resources. This finding is in line with that in Kohpaiboon and Jongwanich [2007]. In particular, when firms lobby for protection, they consider both input and output tariffs together to attain the expected effective protection.

Second, the net effect on productivity remains ambiguous as the interaction term between Inputtariff

i and rawm

ij turns out to be negative and statistically

significant. When firms import raw materials at an amount greater than 33.6 percent, an increase in input tariffs would have a net negative effect of productivity,


all other things equal. In addition, domestically manufactured and imported intermediates in Thai manufacturing are not close substitutes. Any change in input tariffs will have a significant effect only on the firms that actually import. For those who use domestically manufactured intermediates, such a change would not impact those relying entirely on domestic raw materials.

With regards to output tariffs, the negative effect would occur only in exporting firms. The coefficient corresponding to the interaction term between and is negative and statistical significant. For a country like Thailand, which has been long engaged in international markets, firms already make their decisions taking market orientation and where to sell their products. In a case where output is subject to higher tariff rates than inputs are, and various tariff exemption schemes are available, firms choose to either export or sell domestically.11 These firms are operating in different environments. Exporting firms are more productive than domestic-oriented ones as the former usually face more intense competition from the world. Granted protection could keep allow domestic-oriented firms to remain in business, produce products serving local niches, and compete with the former on primary inputs like labor. This is especially true for Thailand, where the labor market has tightened in recent years. This would inflate wages to a certain extent and unevenly affect these two groups of firms. It is the exporting firms that are adversely affected by the inflated wage as their output price is given by the world. To a certain extent, domestic-oriented firms would pass inflated wages on to output prices.

A reduction in output tariffs could generate a tougher competitive environment in domestic markets. The less productive firms that are likely to be purely oriented to the domestic market may be forced out of business. For exporting firms, such a reduction in output tariffs would not have any direct effect as they sell at the world price. Instead, the reduction in output tariffs would lower the inflated wage and relocate workers from less productive and more domestically oriented firms to more productive and export-oriented ones (i.e., resource reallocation).

Another interesting finding is that interaction term between ownership and trade policy (both the erp and disaggregated one) is statistically insignificant (Tables 9 and 10, respectively). This would reflect the dominant role of export-oriented and efficiency-seeking foreign direct investment, which is motivated by strengthening global competitiveness. These foreign firms tend to be eligible for tariff exemption schemes so that their behavior would not be altered by granted protection.

11 It would be costly for a firm to sell to both domestic and foreign markets simultaneously as they must deal with administrative complications—such as how much output is to be sold locally, how to refund the portion of input tariffs paid—as well as cumbersome tariff exemption schemes. This is especially true for small and medium firms.


TABLE 9. Productivity determinants based on ERP decomposition, interaction with ownership and 2011 census



Intercept 2.16* 4.24 2.11* 4.15PLij 2.26* 23.79 2.26* 23.79NLij 0.42* 3.54 0.42* 3.55Kij 0.80* 11.89 0.81* 11.93PLij*Kij -0.12* -20.41 -0.12* -20.38NLij*Kij -0.0002 -0.02 -0.0001 -0.01PLij

2 0.04* 6.37 0.04* 6.26NLij

2 0.004 0.41 0.005 0.42Kij

2 -0.003 -1.02 -0.003 -1.07ownij 0.001 0.46 0.001 0.41ownij*outputtariffj -0.009 -0.57 -0.009 -0.56ownij *inputtariffj 0.06 1.58 0.06 1.68mktij 0.004* 2.78 0.005* 2.90rawmij 0.007* 4.32 0.007* 4.27R&Dij 0.013* 8.24 0.013* 8.28outputtariffj -0.27 -1.01 -0.27 -0.99inputtariffj -2.98* -4.83 -3.21* 5.22Outputtariffj*mktij -0.02** -1.99 -0.02** -2.16

Inputtariffj*mktij-0.02 -0.68 -0.02 -0.67

Outputtariffj*rawmij -0.003 -0.22 -0.002 -0.17Inputtariffj*rawmij -0.098* -3.06 -0.098* -3.06cr4j, t-j -0.33* -2.82 -0.29* -2.54GPN1j, t-j 0.54* 5.71GPN2j, t-j 0.42* 3.84# obs 13593 13593Ad-R 0.73 0.73F-stat 1447 (p-value = 0.00) 1444 (p-value = 0.00)


As a robustness check, the empirical model (Equation 8) is re-estimated by using the previous industrial census (2006). Tables 10 to 12 correspond to Tables 7 to 9, respectively, but they use the 2006 census. To a certain extent, the results are in line with what are found in the recent census, with a few exceptions of statistical insignificance in some coefficients. The main finding of Tables 10 to 12 supports the crucial role of trade liberalization on productivity improvement. Despite being smaller in magnitude, the coefficient associated with the erp is negative and statistically significant (Table 10). The difference is that the coefficient corresponding to the interaction term ERP

j*mkt

ij in Table 10 is not


statistically significant. It becomes statistically significant when the 2011 census is used. The difference in the results between the 2006 and 2011 industrial censuses could be due to the differences in labor market conditions. Another difference is that all interaction terms with input and output tariffs turn out to be statistically insignificant, although the coefficients corresponding to them attain theoretical expected sign when 2006 census data are used.

TABLE 10. Productivity determinants based on ERP and 2006 census



Intercept 6.24* 19.89 6.22* 19.79

PLij1.77* 29.84 1.77* 29.89

NLij 1.69* 34.75 1.69* 34.78

Kij 0.02 0.49 0.03 0.55

PLij*Kij -0.09* -23.78 -0.09* -23.82

NLij*Kij -0.10* -31.68 -0.10* -31.70

PLij2 0.05* 14.08 0.06* 14.13

NLij2 0.06* 18.44 0.06* 18.47

Kij2 0.03* 15.61 0.03* 15.56

ownij 0.002* 4.02 0.003* 4.24

ownij* ERPj, t-j 0.01 0.73 0.009 0.64

mktij 0.001** 1.82 0.001* 2.06

rawmij 0.004* 7.91 0.004* 7.87

R&Dij 0.01* 10.28 0.01* 10.22

ERPj, t-j -0.15* -2.17 -0.14* -2.20

ERPj, t-j *mktij -0.0001 -0.08 -0.0005 -0.23

ERPj, t-j *rawmij -0.45 -1.56 -0.42 -1.46

cr4j, t-j 0.19* 2.77 0.21* 2.97

GPN1j, t-j 0.52* 6.20

GPN2j, t-j 0.41* 3.93

# obs 15564 15564

Ad-R 0.76 0.76


Notes: * statistically significant at 1 percent ** statistically significant at 5 percent An increase in the ERP reflects higher trade protection.Source: Authors’ estimates


TABLE 11. Productivity determinants based on ERP decomposition and 2006 census



Intercept 6.31* 20.07 6.28* 19.98

PLij1.76* 29.70 1.77* 29.74

NLij 1.69* 34.56 1.69* 34.59

Kij -0.0004 -0.01 0.001 0.01

PLij*Kij -0.09* -23.74 -0.09* -23.76

NLij*Kij -0.10* -31.57 -0.10* -31.59

PLij2 0.06* 14.18 0.06* 14.21

NLij2 0.06* 18.49 0.07* 18.50

Kij2 0.03* 15.86 0.03* 15.84

ownij 0.003* 6.44 0.003* 6.64

mktij 0.003* 2.26 0.003* 2.43

rawmij 0.005* 4.31 0.005* 4.14

R&Dij 0.01* 10.15 0.01* 10.08

outputtariffj -0.26 -1.09 -0.26 -1.09

inputtariffj 3.29* 5.45 3.53* 5.86

Outputtariffj*mktij -0.0002 -0.03 -0.002 -0.23



Inputtariffj*rawmij -0.02 -0.91 -0.02 -0.77

cr4j, t-j 0.23* 3.35 0.25* 3.61

GPN1j, t-j 0.43* 5.02

GPN2j, t-j 0.30* 2.80

# obs 15564 15564

Ad-R 0.76 0.76


Notes: * statistically significant at 1 percent ** statistically significant at 5 percentSource: Authors’ estimates


TABLE 12. Productivity determinants based on ERP decomposition, interaction with ownership and 2006 census



Intercept 6.31* 20.07 6.28* 19.97

PLij 1.76* 29.70 1.77* 29.74

NLij 1.69* 34.56 1.69* 34.61

Kij -0.002 -0.03 -0.0004 -0.01

PLij*Kij -0.09* -23.71 -0.09* -23.74

NLij*Kij -0.10* -31.51 -0.10* -31.59

PLij2 0.06* 14.13 0.06* 14.17

NLij2 0.06* 18.42 0.07* 18.44

Kij2 0.03* 15.86 0.03* 15.84

ownij 0.004* 2.50 0.003* 2.42

ownij*outputtariffj 0.009 0.70 0.008 0.65

ownij *inputtariffj -0.02 -0.76 -0.01 -0.56

mktij 0.003* 2.09 0.003* 2.28

rawmij 0.005* 4.35 0.005* 4.21

R&Dij 0.01* 10.12 0.01* 10.05

outputtariffj -0.26 -1.06 -0.26 -1.09

inputtariffj -3.30* -5.46 3.53* 5.86

Outputtariffj*mktij -0.001 -0.19 -0.003 -0.38



Inputtariffj*rawmij) -0.02 -0.76 -0.02 -0.67

cr4j, t-j 0.23* 3.33 0.25* 3.59

GPN1j, t-j 0.43* 5.01

GPN2j, t-j 0.30* 2.77

# obs 15564 15564

Ad-R 0.76 0.76



7. Conclusions and policy recommendations

The paper examines the effect of trade policy on firm productivity by using two recent industrial censuses of Thai manufacturing (2006 and 2011). The translog-production function is employed to avoid imposing any coefficient restrictions. Trade policy and global participation are treated as two different variables in our analysis. Foreign firms tend to be more productive than indigenous ones, all other things remaining equal. The firms, whether domestic or foreign, that engaged in global markets tend to be more productive than those strictly engaged in local markets. As expected, firms spending more on R&D tend to have higher productivity. Participating in the global production network could result in firm productivity improvement.


While controlling for firms’ global participation, defined as export-sale ratio, and the extent to which raw materials are imported, our study finds that trade liberalization could induce firms to commit to productivity-improving activities. The key finding that is different from previous studies is that when it comes to the decision to commit to productivity improvement, the effective rate of protection, where output and input tariffs are taken into consideration together, matters. Focusing solely on lowering input tariffs while leaving output tariffs untouched could retard the overall productivity improvement. This finding would be highly relevant for policymakers in developing countries, where policymakers generally emphasize input tariff reductions while expressing reluctance to lower output tariff.

Two policy inferences can be made from our study. First, our study supports global integration as this could promote productivity enhancement. Second, it would be risky to continue tariff reform by focusing solely on input tariffs while leaving output tariffs untouched. In fact, both input and output tariffs must be taken into consideration in neutralizing incentives in trade policy reform.

*Thammasat University, Thailand

References

Alcala, F. and A. Ciccone [2004] “Trade and productivity”, Quarterly Journal of Economics 119: 613-46.

Amiti, M. and J. Konings [2007] “Trade liberalization, intermediate inputs, and productivity: evidence from Indonesia”, American Economic Review 97(5): 1611-1638.

Aw, B.Y. and A.R. Hwang [1995] “Productivity and the export market: a firm-level analysis”, Journal of Development Economics 47: 313-332.

Aw, B.Y. and G. Batra [1998] “Technological capability and firm efficiency in Taiwan (China)”, World Bank Economic Review 12: 59-79.

Athukorala, P. and A. Kohpaiboon [2009] “Intra-regional trade in East Asia: the decoupling fallacy, crisis, and policy challenge”, adbi Working Paper 177, Tokyo: Asian Development Bank Institute.

Baldwin, R. [1992] “Measurable dynamic gains from trade”, Journal of Political Economy 100: 162-74.

Barry, R.A. [1992] “Firm (or plant) size in the analysis of trade and development”, in G.K. Helleiner, ed., Trade policy, industrialization, and development: new perspectives. Oxford: Clarendon Press.

Bernard, A.B., E.J. Jonathan, J. Bradford, and S. Kortum [2003] “Plants and productivity in international trade”, American Economic Review 93(4): 1268-90.


Caves, R. [2009] Multinational enterprise and economic analysis. 3rd edition. Cambridge University Press: Cambridge.

Edwards, S. [1993] “Openness, trade liberalization, and growth in developing countries”, Journal of Economic Survey 31: 1358-1393.

Frankel, J. and D. Romer [1999] “Does trade cause growth?”, American Economic Review 89: 379-99.

Hobday, M. [1995] “East Asian latecomer firms: learning the technology of electronics”, World Development 23:1171-1193.

Helpman, E. and P.R. Krugman [1985] Market structure and foreign trade. mit Press.

Keesing, D.B. [1983] “Linking up to distant markets: south to north exports of manufactured consumer goods”, American Economic Review 73: 338-42.

Keesing, D.B. and S. Lall [1992] “Marketing manufactured exports from developing countries: learning sequences and public support”, in G.K. Helleiner, ed., Trade policy, industrialization, and development: new perspectives. Oxford: Clarendon Press.

Kohpaiboon, A. [2006] Multinational enterprises and industrial transformation: evidence from Thailand. Cheltenham, uk: Edward Elgar.

Kohpaiboon, A. [2010] Product fragmentation phenomenon, production networks of multinationals and implication on Thai manufacturing (in Thai). Bangkok: Misterkopy.

Kohpaiboon, A. and J. Jongwanich [2007) “Determinants of protection in Thai manufacturing”, Economic Papers 26(3): 276-94.

Kohpaiboon, A. and J. Jongwanich [2014] “Global production sharing and wage premiums: evidence from the Thai manufacturing sector”, Asian Development Review 31(2): 141-164.

Krugman, P. [1979] “Increasing returns, monopolistic competition, and international trade”, Journal of International Economics 9: 469-79.

Melitz, M.J. [2003] “The impact of trade on intra-industry reallocations and aggregate industry productivity”, Econometrica 71(6): 1695-1725.

Melitz, M.J. and S. J. Redding [2012] “Heterogenous firms and trade”, nber Working Paper 18652, National Bureau of Economic Research, Massachusetts.

Melitz, M.J. and D. Trefler [2012] “Gains from trade when firms matter”, Journal of Economic Perspectives 26(2): 91-118.

Noguer, M. and M. Siscart [2005] “Trade raises income: a precise and robust result”, Journal of International Economics 65: 447-60.

Levine, R. and D. Renelt [1992] “A sensitivity analysis of cross-country growth regressions”, American Economic Review 82: 942-63.

Lopez, R.A. [2005] “Trade and growth: reconciling the macroeconomic and microeconomic evidence”, Journal of Economic Surveys 19(4): 623-648.

Pack, H. [1994] “Endogenous growth theory: intellectual appeal and empirical shortcomings”, Journal of Economic Perspectives 8: 55-72.

Pack, H. and K. Saggi [1997] “Inflows of foreign technology and indigenous


technological development”, Review of Development Economics 1: 81-98. Pietrobelli, C. [1998] Industry competitiveness and technological capabilities in

Chile: a new tiger from Latin America? New York: St. Martin’s Press.Nelson, R. and H. Pack [1999] “The Asian miracle and modern growth theory”,

Economic Journal 109: 416-436.Rivera-Batiz, L. and P.M. Romer [1991] “Economic integration and endogenous

growth”, Quarterly Journal of Economics 106: 531-56. Sala-i-Martin, X. [1997] “I just ran two million regressions”, American Economic

Review 87: 178-183. Sachs, J.D. and A. Warner [1995] “Economic reform and the process of global

integration”, Brookings Papers on Economic Activity 1: 1-118.Srinivasan, T.N. [1999] “Trade orientation, trade liberalization and economic

growth”, in G.R. Saxonhouse and T.N. Srinivasan, eds., Development, duality, and the international economic regime: essays in honor of Gustav Ranis. Ann Arbor, Michigan: University of Michigan Press.

Srinivasan, T.N. and J.N. Bhagwati [1980] “Trade and welfare in a steady-state”, in J.S. Chipman and C.P. Kindleberger, eds., Flexible exchange rates and the balance of payments: essays in memory of Egon Sohmen. Amsterdam: North-Holland.

Srinivasan, T.N. and J.N. Bhagwati [2001] “Trade, development and growth: Graham memorial lecture”, Essays in International Economics 225: 1-32.

Tybout, J.R. [1996] “Heterogeneity and productivity growth: assessing the evidence”, in M.J. Roberts and J.R. Tybout, eds., Industrial evolution in developing countries: micro patterns of turnover, productivity, and market structure. New York: Oxford University Press.

Warr, P.G. [2000] “Thailand’s post-crisis trade policies: the 1999 wto review”, World Economy 23(9): 1215-1236.

Westphal, L., Y. Rhee, and G. Pursell [1979] “Foreign influences on Korean industrial development”, Oxford Bulletin of Economics and Statistics 41: 359-388.

Westphal, L., Y. Rhee, and G. Pursell [1984] “Sources of technological capability in South Korea”, in M. Fransman and K. King, eds., Technological capability in the Third World. London: Macmillan.

World Bank (WB) [1991] World development report: the challenge of development. New York: Oxford University Press.

Wortzel, L.H. and H.V. Wortzel [1981] “Export marketing strategies for nic and ldc-based firms”, Columbia Journal of World Business 16: 51-60.

World Trade Organization (wto) [1999] Thailand: trade policy review. Geneva.Young, A. [1991] “Learning by doing and the dynamic effects of international

trade”, Quarterly Journal of Economics 106: 369-405.


PRE

Households’ access to financial services: some evidence from survey data

Gilberto M. Llanto

Many studies look at financial inclusion from the supply side. The discussion in those studies revolves around the different types of financial services being developed to provide the excluded segment of the population with access to such services and the evolving regulatory frameworks supporting those innovative financial services.

This paper views financial inclusion from the perspective of households who use financial services and asks what factors determine access to financial services. It provides a quantitative estimation of the factors affecting household decision to participate in the formal financial markets and the impact of the utilization of financial services on household incomes. It uses micro-data from the Annual Poverty Indicators Survey in the estimation.

The empirical findings provide useful information for designing policies and interventions to foster inclusive finance. It points to financial education of households as a key intervention in financial inclusion strategies.

JEL classification: D14, G21Keywords: financial inclusion, inclusive finance, access to banking services, financial education

1. Introduction

Financial inclusion presently occupies center stage in global discussions of development interventions, and it has drawn the attention of policy makers, regulators, financial service providers, other stakeholders, and even the support of the nobility who are concerned with the negative impact on households of the


inaccessibility of financial services.1 The importance given to financial inclusion globally is motivated by the belief that financial inclusion is important for inclusive growth and poverty reduction. Empirical studies tend to provide some evidence of the beneficial impacts of financial inclusion at the macroeconomic, household, and firm level. For this reason, the Global Partnership for Financial Inclusion was established by the G20 as the main implementing mechanism of the G20 Financial Inclusion Action Plan. In a recent report on financial inclusion, the World Bank points out that at the country level about two-thirds of regulatory and supervisory agencies in many countries are now working on ways to enhance financial inclusion, while some 50 countries have set formal targets and goals for financial inclusion [World Bank 2014]. Financial inclusion is an important strategy for inclusive growth in the 2011-2016 Philippine Development Plan; the Bangko Sentral ng Pilipinas, taking the lead in expanding the accessibility of financial services, created the Inclusive Finance Advocacy Staff to work with various stakeholders in achieving the objectives of financial inclusion.

The Bangko Sentral ng Pilipinas (bsp) defines financial inclusion as “a state wherein there is effective access to a wide range of financial services for all Filipinos” [bsp 2013:1]. This follows the standard definition of financial inclusion in the literature. The Consultative Group to Assist the Poor [2011] defines financial inclusion as a state in which all working-age adults, including those currently excluded by the financial system, have effective access to a range of financial services provided by formal financial institutions: credit, savings (including current account), payments, and insurance. Effective access involves convenient and responsible service delivery at a cost affordable to the customer and sustainable for the provider, while “financially excluded” refers to those who do not have access to or are underserved by formal financial services [cgap 2011]. Effective access requires that financial services are appropriately designed, of good quality, relevant for actual use, and beneficial to the target market [Llanto 2015:1].

Access to financial services satisfies economic agents’ demand for consumption smoothing, productive investments, and ways to help them cope with exogenous shocks, e.g., catastrophic risk. However, a large segment of the global population, especially poor households and micro-enterprises in developing countries, has been financially excluded. According to the latest World Bank estimates, half of the world’s adult population—more than 2.5 billion people—do not have an account at a formal financial institution. Globally, about 50 percent of adults have one or more bank accounts, and a nearly equal

1 Queen Maxima of the Netherlands, un Secretary General’s Special Advocate for Inclusive Finance for Development, lent her presence and support to the launch of the Philippines’ National Strategy for Financial Inclusion on July 1, 2015 at the Philippine International Convention Center.

Llanto: Households’ Access to Financial Services: Some Evidence from Survey Data

172

share are unbanked. In 2011, adults who were banked included the 9 percent of adults who received loans and the 22 percent of adults who saved through financial institutions [World Bank 2014:1-2]. In the Philippines, the 2009 Consumer Finance Survey of the bsp found that 8 in 10 Filipino households did not have a deposit account, only 10.5 percent of adults in the country had a loan from a formal financial institution, and 93 percent of those without any deposit account said they did not have enough money for bank deposits [bsp 2012].

To address these issues, innovative financial services intended to address the problem of financial inclusion—e.g., mobile money, branchless banking, and e-money—are in varying stages of development and utilization in many developing countries. For example, in the Philippines low-income households are primarily using mobile money to send and receive domestic remittances: on average sending us$57 and receiving us$48 [Pickens 2009]. In Malawi, Opportunity Bank took two years to develop m-banking service, which was launched in 2010. Also in 2010, m-pesa and Equity Bank in Kenya announced the a low-cost, low-entry microsavings account called M-Kesho. The objective is to convert the majority of m-pesa’s 9.4 million users into account holders at Equity Bank. There are further plans to offer microinsurance and microloans to account holders [Kumar, McKay, and Rotman 2010]. Such innovative financial services will require supportive regulatory frameworks. In this area, Peru and Philippines have been cited as being more advanced than other developing countries in the development of such frameworks. According to the bsp [2015], the Philippines ranked first in Asia and top three in the world in 2014 in terms of having a conducive environment for financial inclusion based on the Economist Intelligence Unit maiden survey on financial inclusion environments globally.

The literature has documented a positive relationship between finance and economic development at the macro level. It has been pointed out that the lack of access to financial services could lead to a poverty trap and to an increase in the inequality gap (Cámara and Tuesta [2015]; Beck, Demirguc-Kunt, and Levnie [2007])2, that inequality decreases as financial markets deepen [Clarke, Xu and Zou 2006], and that, in the case of India, an all-inclusive financial system would facilitate the process of human development by addressing the basic distortions in the level of human development [Kuri and Laha 2011]. Among others, research at the household level revolves around the link between financial inclusion and reduction of poverty rates (Honohan [2008]; Park and Mercado [2015]) and improvements of household welfare with an important function assigned to financial services as a tool for consumption smoothing and social protection as in the case of micro-insurance.

2 The literature on the impact of financial inclusion at the macro and household level is well summarized in Cámara and Tuesta [2015]. A more copious literature is in World Bank [2014].


If financial exclusion could have deleterious effects as explained in the literature, it is important to understand why households, especially poor households and microenterprises, fail to access financial services. Many studies on financial inclusion look at it from the supply side. The discussion in those studies revolves around the different types of financial services being developed to provide the excluded segment of the population with access to such services and the evolving regulatory frameworks supporting those innovative financial services. Financial inclusion is not the same as providing access to financial services although certainly the first step towards the goal of financial inclusion is to make those financial services very accessible to the excluded. Financial products and services could be accessible to the population, but utilization of such financial services could be low. Thus, there would be a large segment of the population that will continue to be financially excluded despite the accessibility of financial services. Financial inclusion is about providing access to financial services and the excluded households’ and firms’ utilization of those services.

This paper takes it from the perspective of users of financial services, that is, the households. What prevents those households from accessing financial services? What influences their decision to use or not to use financial services? Several demand-side factors have effectively excluded poor households from accessing and using financial services. There is a range of factors that prevent access and utilization: socio-economic and cultural factors; the lack of formal identification needed to satisfy the “know your client” policy imposed on banks by the regulator; low levels of financial literacy in addition to the absence of appropriate consumer protection mechanisms (Alliance for Financial Inclusion [2010]; Llanto [2015]) and lack of awareness of available services; inappropriateness of certain services to the needs of the low-income sectors; and the risks of dealing with poor customers [escap 2014]. It is important to understand the socio-economic characteristics conditioning the use of financial services by households, which enable such households to smoothen income cycles generated by unexpected shocks or discontinuous income flows [Cámara and Tuesta 2015].

Using micro-data from the Annual Poverty Indicators Survey, the paper provides a quantitative estimation of factors affecting household decision to participate in the formal financial markets and the impact of utilization of financial services on household incomes. It also shows that vulnerable groups— comprised of women, rural dwellers, and young people—find it most difficult to access banking services.

The empirical findings provide useful information for designing policies and interventions to foster inclusive finance. It points to financial education of households as a key intervention in financial inclusion strategies.


174

The paper is organized as follows. Section 2 provides an overview of the Philippine financial sector and the state of financial inclusion in the country, with a focus on the critical role of an enabling environment in promoting inclusive finance. Section 3 discusses the methodology and data used in the empirical estimation. Section 4 analyzes the empirical findings, and the concluding section provides some recommendations for fostering financial inclusion and comments on further research on the subject.

2. Current status of financial inclusion

2.1. Brief profile of the financial sector3

In 2014, the Philippine banking system remained strong and stable despite external challenges—especially in international capital markets—with continuing growth in resources, deposit liabilities, and loans. The total resources of the whole banking system increased by 11.8 percent to more than P11 trillion from P10.3 trillion in the preceding year, 2013. This can be attributed to growth in loans, financial assets, and equity investments (Table 1). Total deposits of banks rose to P8.52 trillion in the same period, a 12 percent year-on-year increase from end-December 2013. The number of banking institution head offices decreased to 648 as of end-December 2014 from the previous year’s 673 head offices, signifying a consolidation of banks and the closure of weaker banks in the sector (Table 2). The number and types of banking offices are also shown in Table 2. The notable information here is the growth at 11 percent of micro-banking offices of mostly microfinance-oriented banks. Together with microfinance- oriented branches, they cater to the lower-income groups and are the access points that are accessible to the excluded segments of the population.

Asset quality indicators also improved with the decline of the banking system’s gross non-performing loan ratio from 2.8 percent as of end-December 2013 to 2.3 percent as of end-December 2014. Likewise, net non-performing loans were reduced. Capital adequacy ratios remained above the international standards imposed under the Basel III framework, which became effective on January 1, 2014. As of end-September 2014, the capital adequacy ratios of universal and commercial banks stood at 17 percent, while overall the capital adequacy ratio for all types of banks stood at 16.7 percent at end-September 2014 (Table 1).

3 This paragraph was drawn from Llanto [2015].


TABLE 1. Resources, deposits, and loans outstanding, all banks, December 2014

All Universal Commercial Banks

Thrift Banks

Rural Banks

Number of Banks

Total number of Banks 10,361 6,330 1,920 2,608

Head offices 648 51 69 543

Other Offices 9,713 6,279 1,851 2,065

Resources (P billion) 11,128 10,398 916 208

Deposits Liabilities (P billion) billion) 8,522 7,680 696 144

Loans Outstanding (P billion) 5,532 4,822 571 138

GNPL to Total Loans (%) 2.3 2.3 4.4 11.9

NNPL to Total Loans (%) 0.6 0.6 1.95 5.9

Capital Adequacy Ratio (%) 16.7* 17.0**

* As of end-June 2014; ** As of end-September 2014 Source: Bangko Sentral ng Pilipinas.

TABLE 2. Number of banking offices, by type, 2013 and 2014

2013 2014 Growth Rate (%)

TOTAL 9,935 10,361 4.3

Head Offices 673 648 -3.7

Branches/Other Offices 9,262 9,713 4.9

Regular Branch 8,077 8,442 4.5

Micro-finance Oriented Branch 98 99 1.0

Regular Other Banking Office (ROBO) 420 448 6.7

Microbanking Office (MBO) 465 517 11.2

Extension Office (EO) 166 176 6.0

Representative Office 15 13 -13.3

Remittance Desk Office 16 14 -12.5

Marketing Office 2 2 0.0

Sub-Branch 2 1 -50.0

Limited Purpose Branch 1 1 0.0

Sources: Bangko Sentral ng Pilipinas

The regional distribution of banking offices in Table 3 gives a rough idea of the spatial distribution of access to banking facilities. A finer distribution by municipalities and cities is available at the bsp web site. The population residing in richer regions, which have more banking facilities, have easier access to financial services. Those residing in poorer regions with fewer banking facilities do not have this advantage. Cities/municipalities/provinces/regions with a higher bank density have more financially included individuals in their respective


176

populations than those with lower bank density. The leading regions from this perspective are the National Capital Region, Regions III and IV-A in Luzon, and Regions VI and VII in the Visayas. The regions in Mindanao have fewer banking offices and presumably have a bigger share of the financially excluded segments of the population.

TABLE 3. Regional distribution of banking offices, 2010-2015p

2010 2011 2012 2013 2014 2015*Growth 2012-2013

Growth 2013-2014

Percent Share (%)

2014

Philippines 8,843 9,015 9,375 9,884 10,315 10,410 5.4% 4.4% 100.0

NCR 2,876 2,892 2,993 3,141 3,275 3,299 4.9% 4.3% 31.7

CAR 138 146 148 150 155 156 1.4% 3.3% 1.5

Region I 403 401 413 436 456 466 5.6% 4.6% 4.4

Region II 257 272 286 310 329 337 8.4% 6.1% 3.2

Region III 914 940 975 998 1,033 1,050 2.4% 3.5% 10.0

Region IV-A 1,314 1,350 1,406 1,509 1,566 1,575 7.3% 3.8% 15.2

Region IV-B 184 189 206 220 234 236 6.8% 6.4% 2.3

Region V 271 281 315 353 380 383 12.1% 7.6% 3.7

Region VI 515 531 543 572 600 604 5.3% 4.9% 5.8

Region VII 580 584 627 653 683 688 4.1% 4.6% 6.6

Region VIII 165 172 174 183 186 187 5.2% 1.6% 1.8

Region IX 171 181 190 199 201 202 4.7% 1.0% 1.9

Region X 320 328 328 347 368 372 5.8% 6.1% 3.6

Region XI 333 338 355 388 396 400 9.3% 2.1% 3.8

Region XII 190 192 195 197 214 216 1.0% 8.6% 2.1

Caraga 193 199 201 207 218 218 3.0% 5.3% 2.1

ARMM 19 19 20 21 21 21 5.0% 0.0% 0.2

Note: * - as of March 2015Source: BSP Statistics

2.2. Financial inclusion

The latest data on financial inclusion are those reported by the bsp from data gathered through the National Baseline Survey on Financial Inclusion [bsp 2015]. During the launch of the National Strategy for Financial Inclusion on July 1, 2015, the bsp reported the following statistics: 25 percent of Filipino adults have never saved; 32 percent used to save; and only 43 percent presently have savings. Of those with savings, only 32 percent save in banks, while 68 percent keep their savings at home. Around 65 percent of unbanked adults cited lack of money as the main reason for not having a bank account. About 47 percent of adults have outstanding loans. The main source of borrowing is informal: 62 percent borrow from family, relatives, or friends; while 10 percent borrow from informal lenders. About 44 percent of adults sent or received money, while 42


percent made payments. Only 3.2 percent of adults have a microinsurance coverage [bsp 2015].

There are several salient findings of the National Baseline Survey on Financial Inclusion. In terms of access, the bsp [2015] reports that among the available access points, Filipino adults are most aware of banks (98.3 percent), pawnshops (95.7 percent), and automated teller machines (93.5 percent). Seemingly, there is a relatively low awareness of other access point, e.g. microfinance nongovernment organizations (30.5 percent), e-money agents (25.6 percent), and non-stock savings and loan associations (13.6 percent). As of end-December 2014, 36 percent of municipalities do not have a banking office. While the physical network of banks and atms continues to experience sustained growth, there are disparities in the regional distribution of access points.

In terms of usage, most of those who are aware of the access points had also conducted transactions using the previously mentioned top access points. Nonetheless, only 5 out of 10 Filipino adults have conducted transactions with banks. Also, there are certain access points which are more frequently used than others depending on the geographic location (i.e. island group) of the user. For instance, adults in Mindanao tend to transact more often with cooperatives and microfinance nongovernment organizations; in Visayas, the majority of adults have transacted with non-stock savings and loan associations and pawnshops. Significant usage disparity was also evident between users in urban rural areas [bsp 2015].

Meanwhile, the proportion of Filipino adults who save remains small at 43.2 percent; while 32.3 percent of the respondents used to save in the past, and the remaining (24.5 percent) have never experienced saving money. It is worth noting that the majority (7 out of 10 adults or 68.3 percent) prefer to save their money at home; 32.7 percent save through banks; and the remaining through other financial institutions and informal savings group. It seems that such behavior among most of the Filipino adults stems from the main reasons for saving, i.e. to use in case of emergencies (63.8 percent); for future expenses on food (55.6 percent); and education (47.4 percent). This may imply that a significant percentage of Filipinos would rather forgo the interest income from savings deposits in banks in exchange for easier access to savings, that is, keeping cash at home. Some of the other reasons cited for not saving in the banks were lack of money (65 percent), limited knowledge and capability to manage an account (16.8 percent), cost (11.2 percent), proximity of the banks (7.6 percent), and failure to meet documentary requirements (4.6 percent), among others. In terms of loans, 47.1 percent of adults borrow money of whom 61.9 percent borrow from family, relatives, or friends, and 10.1 percent borrow from informal lenders. Among the main considerations for borrowing are interest rate and loan amount. On insurance, most are aware of health and life insurance. Results showed that the most common reasons for not enrolling in life, health, or accident insurance are lack of money and perception of high cost.


178

More than half who have accessed banks and automated teller machines are only somewhat satisfied with their transactions. This is most common in automated teller machines, cooperatives, and microfinance nongovernment organizations. Finally, on welfare, the results of the survey indicated that 86 percent of the Filipino adults perceive access to financial products and services is important, while 88 percent believed that it is beneficial to them. Also, majority of the adults believe that saving, borrowing, and insurance are important to them as well. Nonetheless only half of the (potential) borrowers would want to borrow from financial institutions. The findings of the first National Baseline Survey on Financial Inclusion indicate that much work has to be done to achieve financial inclusion in the country. This includes extensive information dissemination, financial education especially for poor households, and development of financial products and services that are responsive to the financial needs of the excluded segment of the population.

It is good to benchmark financial inclusion levels in the Philippines relative to other countries in order to provide a good perspective and understanding of the local situation. Data are sourced from the 2014 Global Findex. Account penetration in the Philippines was lower than all of the countries within the peer group (asean), except in Vietnam and Cambodia (Figure 1). Account penetration in the Philippines was lower than all of the countries within the peer group (asean) except in Vietnam and Cambodia (Figure 1). The percentage of adults who have savings in a financial institution was lower than all countries within the peer group except Cambodia. The Philippines and Viet Nam have more or less the same percentage of adults with savings in a financial institution. (Figure 2). Surprisingly, the percentage of adults with loans obtained from a financial institution was lowest in the Philippines (Figure 3).

3. Data and methodology

The methodology for estimation is as follows. Using the Heckman selection model estimation, the likelihood of availing of loans regardless of the source of loans—that is, loans from formal institutions or informal lenders—was first tested, followed by the likelihood of getting a formal loan (access to formal credit), as a series of Probit models for households. Both procedures used Full-Maximum Likelihood estimation. Here the observations were limited to those households with access to formal loans. The Heckman selection model estimation addressed the sample selection problem that could arise from the use of samples that include those that did not avail of loans. The other reason is that it will be interesting to find out what factors matter for household access to loans from formal and informal sources, and to loans from a formal financial institution, e.g., a bank.


World

East Asia and the Pacific

Singapore

Malaysia

Thailand

Indonesia

Philippines

Vietnam

Cambodia

0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0

World


Singapore

Malaysia

Thailand

Indonesia

Philippines

Vietnam

Cambodia

0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0

Source: Global Findex, 2014

FIGURE 1. Percentage of adults with a formal account


FIGURE 2. Percentage of adults with savings in a formal financial institution


180

World


Singapore

Malaysia

Thailand

Indonesia

Philippines

Vietnam

Cambodia

0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0

The second step is to test whether or not financial inclusion helps improve household income using a two-stage instrumental variable approach to take care of the possible endogeneity problem between financial inclusion (proxied by access to formal credit) and household income. Using the number of formal lending institutions (i.e., universal, thrift, and rural banks) as instrument, the Two-Stage Least Squares estimation was employed upon satisfaction of the weak identification and heteroskedasticity tests.

3.1. Heckman selection model

The first step is to test which factors significantly influence a household’s decision to access financial services (proxied by access to formal credit). To take into account the endogenous borrowing decisions of households, the Heckman selection model, which is specified as follows, is employed:

q = xβ+ u (1)

Dloan

= I(zδ + v > 0) (2)

Equation 1 estimates the probability of a household using formal financial services; that probability is determined by a set of exogenous variables included in vector x. β is a vector of parameters while u is a normally distributed error term


FIGURE 3. Percentage of adults with loans from a formal financial institution


with mean 0 and variance 1 and is assumed to be independent from x. Equation 2 is the first-stage equation—or the so-called “selection” equation—that estimates the probability of a household availing of a loan (regardless of type), conditional on a set of exogenous variables included in vector z. D

loan is a dummy variable

taking the value of 1 if the household availed of a loan and 0 otherwise. I(zδ + v > 0) is an indicator function that takes the value of 1 if the inequality inside the function holds and 0 otherwise. z contains the same set of variables as in x plus an instrumental variable, which is an exclusion restriction. δ is a vector of parameters and v is assumed to have a standard normal distribution and is independent from z.

Essentially, Equation 1 is estimated when q is observed or when Dloan

= 1. Thus, taking the expectation of Equation 1, conditional on z and D

loan = 1, with u

and v being jointly normal with mean 0, the conditional mean can be written as follows:

E(y | z, Dloan

= 1) = xβ + [ρ*λ(zδ)] (3)

where λ(c) = Φ(c) / Φ(c) is the inverse Mills ratio, which is the ratio of the standard normal probability density function to the standard normal cumulative density function, and ρ is the coefficient. In Equation 3, the inverse Mills ratio is evaluated at zδ.

3.2. Instrumental variable regression model

The second step is to test whether financial inclusion (proxied by access to credit) is a significant factor affecting household income. To address the potential endogeneity between access to credit and household income, the two-stage instrumental variable regression model is estimated, with the number of banks in a province as the instrument. The specification of the two equations are as follows:

ln(y) = wθ + xβ + ε (4)

w = zα + τ (5)

Equation 4 is the outcome equation that estimates the natural logarithm of household per capita income, ln(y), conditional on the credit variable, w, and a vector of x, which includes household head profile, household composition and location. θ and β are the coefficient of the credit variable and a vector of coefficients of the other explanatory variables, respectively, while ε is the error term that is assumed to have a standard normal distribution and is independent from the explanatory variables. Equation 5 is the credit equation that measures access to credit. Like in Equation 2, z contains the same set of variables as in x plus an instrumental variable, which is the number of banks in a province. The number of banks is assumed to have a significant effect on the credit variable but


182

has no direct effect on the outcome variable or per capita income of a household. α is a vector of parameters and τ is assumed to have a standard normal distribution and is independent from z.

The estimations of factors conditioning the likelihood of using financial services by households and the impact of financial inclusion on household income used micro-data from the 2013 apis. The survey provides rich data on the households’ socio-economic characteristics and other aspects of the household economy,e.g., access to formal and informal loans. A chief limitation is the absence of information on types of financial products or services used or accessed by households, e.g., use of savings accounts, remittances services, and the like. Hence, in this paper, access to formal credit was used as proxy for access to various types of financial services.

Data used for the estimation come from the apis, a nationwide survey that collects information on poverty-related indicators, such as those pertaining to the socioeconomic and living conditions of households and their members, their access to government programs, and the impact of economic crisis, among others. This particular sample survey is used for the estimation of the country’s poverty statistics during years when the Family Income and Expenditure Survey has not been conducted.4 Although the apis does not collect comprehensive information on household income and expenditure compared to what the Family Income and Expenditure Survey does, nevertheless it gathers information on a number of non-income indicators that are important in poverty monitoring and assessment. The apis is one of the nationally representative sample surveys in the country that gathers household-level information on both income and credit availment for the operation of economic activities of households (psa [2015]).

The 2013 apis has a sample size of 10,864 households, which is relatively lower than the sample size of the earlier rounds. In addition, this survey was the first round of the apis that included the income module of the Family Income and Expenditure Survey. The 2013 apis was conducted by the Philippine Statistics Authority in July 2013 and was funded by the Department of Budget and Management (Balamban et al. [2013]).

4. Estimation results

Several regression analyses were done, and the following provided the best empirical results. Annex A shows the variables used and their definition.

4 The Family Income and Expenditure Survey is conducted every three years.


4.1. Heckman selection model estimation5

The first-stage equation looked at the characteristics of households having access to a loan, regardless of the source of loans. The first-stage regression results show what matters to households in getting a loan, regardless of the source of loans (Table 4). The age of the household head matters in accessing loans. Bigger family size and a high dependency ratio lead household heads to borrow, while being employed is also a significant factor. The presence or availability of banks does not necessarily matter in household decisions to borrow at this stage. The source of loans could be informal lenders, which, as the literature shows, are mostly the source of loans for poor households.

The second-stage equation looked at the characteristics of households which had access to formal credit. In the second-stage regression, shifting to a formal loan source (a bank) household decision to use financial services is positively and significantly correlated with family size as well as sex, age, marital status, and educational attainment of the household head (Table 4). The dependency ratio, measured as the number of dependents below 15 years old, exerts a negative and significant influence on the decision to use financial services. The bigger family size means there are more members in the households, and they may not necessarily be dependents.

Primary and secondary education and more so with tertiary education help household heads in deciding to access loans from formal financial institutions. Data tabulations reveal that there are more households with heads having tertiary level education who availed of formal loans than the household heads with primary and secondary education. The results seem to indicate that level of education—more properly, tertiary education—is a significant factor in household decisions to access formal loans. Similar results by Honohan and King [2012] indicate that income and education are key demand side determinants of access to formal banking. On the other hand, level of education does not matter to households in accessing loans, regardless of source.

Poor households located in the National Capital Region do not necessarily approach a formal financial institution, such as a bank, to borrow money. In the first-stage regression, the location of households does not matter in accessing loans, regardless of the source of loans.

5 Using the Full-Maximum Likelihood (ml) estimation; at the second stage where access to formal credit is the dependent variable, the observations are limited to those with access to such credit.


184

TABLE 4. Estimated models on access to any type of loans and on access to formal loans

Notes: Figures in parentheses are analytical standard errors;* significant at 10% ** significant at 5% *** significant at 1%

4.2. Two-stage instrumental variable estimation

Turning now to the impact of financial inclusion on household income, Tables 5 and 6 show the results of a two-stage instrumental variable estimation. In the first-stage regression, a model on access to credit was estimated with the number of banks in a province as instrument. Among the borrowing households, the availability of banks and level of education (secondary and tertiary) are significant positive factors in accessing formal credit, while those with a large

First-stage equation:Dependent variable: Access to loansRegressor Estimate

Constant -1.1210 ***

(0.1681)

Natural logarithm of number of banks -0.0600 ***

(0.0118)

Profile of household head

Sex 0.0251

(0.0440)

Age 0.0142 **

(0.0064)

Square of age -0.0000 ***

(0.0001)

Married 0.0636

(0.0433)

At least elementary graduate 0.0257

(0.0381)

At least high school graduate -0.0140

(0.0395)

At least post-secondary/college graduate -0.0300

(0.0488)

Employed 0.2390 ***

(0.0437)

Household Composition

Household size 0.0552 ***

(0.0074)

Dependency ratio 0.2200 ***

(0.0737)

Location

Within National Capital Region -0.0410

(0.0498)

Second-stage equation:Dependent variable: Access to formal creditRegressor Estimate

Constant -0.0922

(0.1151)


Sex -0.0861 ***

(0.0314)

Age 0.0144 ***

(0.0041)

Square of age -0.0001 ***

(0.0000)

Married 0.0878 ***

(0.0299)

At least elementary graduate 0.0805 ***

(0.0232)

At least high school graduate 0.1904 ***

(0.0245)

At least post-secondary/college graduate 0.3784 ***

(0.0319)

Employed 0.0358

(0.0329)

Household composition

Household size 0.0147 ***

(0.0051)

Dependency ratio -0.1909 ***

(0.0501)

Location

Within National Capital Region -0.1532 ***

(0.0281)


family size and high dependency ratio do not access formal credit. Intuitively, the demand for formal credit tends to increase with the number of formal lending institutions in the area. While this is true, an increase in the number of banks does not necessarily stimulate household borrowing. The presence of banks is a distinct advantage but not a sufficient condition for access to credit. Poorer and less educated households may decide not to transact with a bank for a variety of reasons, e.g., lack of information or familiarity with banking procedures, high transaction cost.

The finding on the education variable is consistent with the findings of Honohan [2008] and Park and Mercado [2015], who noted that primary education completion and literacy rates do not have a significant effect on the level of financial inclusion in developing Asia. It seems that it takes more than just primary education for households to be able to use financial services.

A high dependency ratio acts as a barrier to financial inclusion. This echoes the finding of Park and Mercado [2015] that higher age dependency ratio significantly reduces financial inclusion. This is because a larger segment of the population is either too young or above the retirement age, which impedes their access to financial services as they do not earn income.

Although not significant, the signs of the coefficients indicate that households with very young and very old heads either have higher probability of not availing of loans, regardless of source, or of accessing formal credit. The seemingly counterintuitive results on the employment variable is explained as follows. An inspection of the data showed that as households move from a situation where all the loan sources are informal lenders (loan2=0, Annex A) to where households do not borrow at all (loan2=2, Annex A), the proportion of households with employed heads increases slightly from 88.6 percent to 89.9 percent and then drops significantly to 79.4 percent. On the other hand, the proportion of households with heads who were not employed decreases slightly from 11.4 percent to 10.1 percent and then increases substantially to 20.6 percent.

In the second-stage regression, Table 6 shows the results on the impact of financial inclusion on household income. Financial inclusion (here proxied by access to credit) has a positive and significant impact on household income. Higher-income households may not necessarily be availing themselves of formal loans because they most likely have demand for other financial services, such as savings with a formal institution, payment services with a bank, insurance, and others. The financial inclusion survey of the bsp revealed that the proportion of adults who keep their money in banks is significantly higher in classes A, B, and C (around 71 percent) than in class D (32.7 percent) and class E (17.2 percent). Among those who borrow, higher-income households tend to have higher access to formal credit. Education at any level—whether primary, secondary, or tertiary—similarly has a positive and significant influence, and the magnitude of the impact increases with educational level. Apparently, the households with more educated heads tend to have higher-paying jobs relative to those with less educated heads.


186

TABLE 5. Estimated model on access to credit (first-stage regression)

First-stage equation:Dependent variable: Access to credit

Regressor Estimate

Constant 1.6328 ***

(0.0882)

Natural logarithm of number of banks 0.0284 ***

(0.0068)


Sex -0.0323

(0.0222)

Age -0.0023

(0.0032)

Square of age 0.0000

(0.0000)

Married -0.0135

(0.0224)

At least elementary graduate 0.0125

(0.0223)

At least high school graduate 0.0668 ***

(0.0228)

At least post-secondary/college graduate 0.1197 ***

(0.0255)

Employed -0.0981 ***

(0.0208)

Household Composition

Household size -0.0240 ***

(0.0043)

Dependency ratio -0.1714 ***

(0.0415)

Location

Within National Capital Region -0.0234

(0.0269)

Notes: Figures in parentheses are robust standard errors;* significant at 10% ** significant at 5% *** significant at 1%

The location variable, the National Capital Region, which can be broadly interpreted as an urban location variable, is also a significant and positive determinant of household income. Jobs seem to be more available in dense urban settings than in the rural areas.


The age of the household head seems to matter also on the level of household income. The sign of the coefficient of the squared age variable implies that household income increases with the age of the household head up to a certain point and then decreases thereafter. Very young household heads are at the start of their career, and it is reasonable to assume that they are earning relatively less compared to the older ones. On the other hand, very old household heads are no longer working and earning. Both of these cases result in a lower per capita income of a household.

Meanwhile, a higher dependency ratio has a significant negative correlation with household income for obvious reasons. Households with married and/or employed heads have greater chances of improving household incomes.

TABLE 6. Estimated model on the impact on household income (second-stage regression)

Second-stage equation:Dependent variable: Natural logarithm of per capita income

Regressor EstimateConstant 4.4746 ***

(1.3101)Access to credit 3.0266 ***

(0.7345)Profile of household headSex -0.0413

(0.0724)Age 0.0179 *

(0.0100)Square of age -0.0002 **

(0.0001)Married 0.1209 *

(0.0691)At least elementary graduate 0.2004 ***

(0.0701)At least high school graduate 0.4113 ***

(0.0912)At least post-secondary/college graduate 0.9286 ***

(0.1246)Employed 0.1919 *

(0.0994)Household compositionHousehold size -0.0089

(0.0223)Dependency ratio -0.3362 *

(0.1856)LocationWithin National Capital Region 0.3119 *** (0.0729)

Notes: Figures in parentheses are robust standard errors;* significant at 10% ** significant at 5% *** significant at 1%


188


Current policy discussions hold that financial inclusion is important for inclusive growth and poverty reduction. Theoretical and empirical studies have looked at the beneficial impacts of financial inclusion at the macroeconomic, household, and firm level. The present study paid special attention to households, especially poor households, to get a better understanding of what factors drive financial inclusion at this level.

The empirical findings showed what factors significantly influence household decision to access financial services. The findings clearly showed robust and significant correlation between household decision to use financial services (proxied by access to formal credit, due to data limitations) on the one hand, and the age of the household head, marital status, family size, and educational attainment of the household head, on the other. Poor households with more dependents (those below 15 years of age) tend not to use financial services. This probably has to do with weaker incomes or the financial capacity of households with too may mouths to feed to repay loans. There is a need to underscore the importance of a higher level of education (secondary and tertiary) as a positive and significant factor in household decisions to access and use formal financial services. On the other hand, households may be able to access informal loans regardless of the level of education of the household head.

The empirical findings from a two-stage instrumental variable estimation supports the hypothesis that financial inclusion improves household income. Financial inclusion (here proxied by access to formal credit) has a positive and significant impact on household income. The empirical findings of the paper lead to certain policy implications.

First, expanding access to and use of financial services by low-income households/individuals may have a positive effect on household/individual welfare. Access and use of financial services are an important tool of households for consumption smoothing, making productive investments, and coping with catastrophic risks. Based on evidence of a strong correlation between financial inclusion and higher household income, there is a need for policies and interventions that reduce barriers to financial inclusion.

Second, a key measure to address the financial exclusion of poor households is financial education. Education at any level—primary, secondary, or tertiary—builds human capacity and has a positive and significant influence on household income. This is an important finding in view of the fondness of some politicians to use credit subsidies to address the problem of lack of access and utilization of financial services by poor households, small farmers, and similar economic agents, e.g., microenterprises. It is good to pay attention as well to non-financial


factors, such as education, to equip such small economic agents with the capacity to access and use financial services.

Philippine Institute for Development Studies (PIDS)

The author acknowledges the invaluable research assistance provided by Christian Mina. The

views and opinions in this paper are solely the author’s and do not in any way reflect those of

the Philippine Institute for Development Studies.

References

Alliance for Financial Inclusion (afi) [2010] The afi survey on financial inclusion policy in developing countries: preliminary findings. Kuala Lumpur.

Balamban, B.B., M.B. Addawe, and M.G.G. Darunday [2013] “Official poverty statistics and poverty reduction programs of the Philippines”, http://adbi.adb.org/files/2013.05.08.cpp.sess4.3.poverty.reduction.philippines.pdf. Accessed 21 October 2015.

Bangko Sentral ng Pilipinas (bsp) [2014] State of financial inclusion in the Philippines, 2013 data updates. Manila: Bangko Sentral ng Pilipinas.

Bangko Sentral ng Pilipinas (bsp) [2015] National strategy for financial inclusion. Manila: Bangko Sentral ng Pilipinas.

Beck, T., A. Demirguc-Kunt, and R. Levine [2007] “Finance, inequality, and the poor”, Journal of Economic Growth 12: 27-49.

Cámara, N. and D. Tuesta [2015] “Factors that matter for financial inclusion: evidence from Peru”, Aestimatio the ied International Journal of Finance, 10: 8-29.

Clarke, G.R.G., L.C. Xu, and H. Zou [2006] “Finance and inequality: what do the data tell us?”, Southern Economic Journal 72(3): 578-596.

Consultative Group to Assist the Poor (cgap) [2011] “Global standard-setting bodies and financial inclusion for the poor: toward proportionate standards and guidance”, a white paper on behalf of the Global Partnership for Financial Inclusion.

Economic and Social Commission for Asia and the Pacific (escap) [2014] “Sustainable development financing: perspectives from Asia and the Pacific”, paper prepared for the Regional Outreach of the Intergovernmental Committee of Experts on Sustainable Development Financing for the Asia- Pacific Region, 10-11 June 2014.

Honohan, P. [2008] “Cross-country variation in household access to financial services”, Journal of Banking and Finance 32: 2493–2500.


190

Honohan, P. and M. King [2012] “Cause and effect of financial access: cross-country evidence from the Finscope Surveys”, iiis Discussion Paper No. 399, Trinity College, Dublin, May.

Kumar, K., C. McKay, and S. Rotman [2010] “Microfinance and mobile banking: the story so far”, Focus Note 62. Washington, d.c.: cgap.

Kuri, P.K. and A. Laha [2011] “Financial inclusion and human development in India: an inter-state analysis”, Indian Journal of Human Development 5(1): 61-77.

Llanto, G.M. [2015] “Financial inclusion, education, and regulation in the Philippines”, adbi Working Paper 541. Tokyo: Asian Development Bank Institute.

Park, C.Y. and R. Mercado, Jr. [2015] “Financial inclusion, poverty and income inequality in developing Asia”, adb Economics Working Paper Series No. 426 January.

Philippine Statistics Authority (psa) [2015] “Technical notes on the Annual Poverty Indicators Survey”, https://psa.gov.ph/content/technical-notes-annual-poverty-indicator-survey-apis. Accessed 21 October 2015.

Pickens, M. [2009] Window on the unbanked: mobile money in the Philippines. World Bank: Washington d.c.

Verbeek, M. [2008] A guide to modern econometrics. 3rd edition. West Sussex, England: John Wiley and Sons Ltd.

Wooldridge, J.M. [2002] Econometric analysis of cross section and panel data. Cambridge, Massachusetts: MIT Press.

World Bank (WB) [2014] Financial inclusion: global financial development report. Washington, d.c.


ANNEX 1. Definition of variables for Heckman selection odel

ln_pcinc = natural logarithm of per capita income of the household avail_loan = 1 if the household availed of a loan during the past 6 months,

regardless of source; 0, otherwiseformal_loan = 1 if the household availed of a loan from during the past 6 months

and at least one of the sources is a formal lending institution; 0, otherwisesex = 1 if household head is male; 0 if femaleage = age of household headagesq = squared age of household headeduc1 = 1 if head attained at most elementary undergraduate (base category)educ2 = 1 if head attained at least elementary graduate and at most high school

undergraduateeduc3 = 1 if head attained at least post-secondary or college graduatemarried = 1 if household head is married; 0, otherwiseemployed = 1 if household head is employed; 0, otherwisefsize = number of members in the householddep_ratio = proportion of household members aged below 15ncr = 1 if the household is located within the National Capital Region; 0 if located

outside ncr. In the absence of an urbanity variable, this variable was used to represent the location variable.

imr = inverse Mills ratio, or ratio of the standard normal probability density function to the standard normal cumulative distribution function of the predicted value of avail_loan; addition of this in the model as a regressor addresses sample selection bias

ln_banks = natural logarithm of the number of banks within a province (instrument)

_cons = constant termThe following variables were used in the two stage instrumental variable

estimation, in addition to those listed above.ln_pcinc = natural logarithm of per capita income of the household loan2 = 2 if the household did not avail of a loan during the past 6 months; 1

if the household availed of a loan and at least one of the sources is a formal lending institution; 0 if the household availed of a loan and all of the sources are informal lenders6

6 The definition of the loan variable was based on the following observations from the data set: the mean per capita income of households that did not avail of a loan was P32,696.79; the mean per capita income of households that availed of a loan and at least one of the loans was

sourced from a formal lending institution was P30,009.64; and the mean per capita income of households that availed of a loan and all loans were sourced from

informal lender(s) was P18,318.37.


PRE

The ASEAN business cycle and China’s slowdown

Bhanupong Nidhiprabha

The economies in the Association of Southeast Asian Nations (asean) have been integrated through increasing trade relations intensified by tariff reductions and increasing openness to foreign direct investment. Rising volume of networks trade has deepened interconnectedness between asean and China, a recent growth locomotive in the world. asean business cycle is shaped by volatile China’s trade volume. As China’s expansion slows down, adverse consequences on asean economies have become more pronounced. The extent of the damage depends on each member of asean’s trade exposure and China dependency. This paper identifies the most vulnerable asean economies to China’s business cycle. The slowing down of the Chinese economy would undoubtedly result in a decline in long-term growth of some asean economies, unless appropriate policy responses can be implemented.

JEL classification: E32, F43Keywords: ASEAN Growth, Business Cycles, China’s Slowdown

1. Introduction

Hill [2014] raises an important question on whether there is a Southeast Asia development model. The 10 countries of the Association of Southeast Asian Nations (asean) have become more integrated over time. Hill argues that large socio-economic and institutional disparities among asean members imply that the asean development model does not exist yet. Nevertheless, convergence in growth rates and social development are taking place. In time, through spillover and demonstration effects from asean’s leaders to its followers, we might find such coherent regional development strategies. According to Hill and Gochoco-Bautista [2013], there is policy consensus on the desirability of conservative fiscal policy, except during the crisis, and on the independence of central banks. However, rethinking of an export-led growth strategy should be for those countries with highly open and trade-dependent countries.


China, the world’s second-largest economy, has experienced a drop in economic growth rate from 10 percent in 2010-2011 to 7 percent in the first two quarters of 2015. The economic slowdown and the depreciation of the yuan would have significant negative consequences on some asean economies. Through vertical integration, Foreign Direct Investment (fdi) has made asean vulnerable to fluctuations in world trade volume. The asean business cycles are synchronized through regional trade linkages. On the financial flows, synchronization of asean-5 stock market has been well documented. Teng et al. [2013] report that stock markets in traditional asean countries are well aligned with economic activities in developed economies rather than those in emerging economies. Also, they do not react to external shocks originating in other asean markets. Each stock market in asean-5 responds differently in terms of direction and degree towards changing world economic conditions. In August 2015, the stock market crash in China led to fears of a market meltdown.1 The “China factor” is significant in real and financial sectors to the rest of the world, from commodity prices, exchange rates, and stock prices.

Because asean economies have some degree of heterogeneity, some asean countries would suffer from the slowing growth of China more than others. Differences in the level of trade integration, export commodity concentration, and export market dependency would lead to dissimilar impact of China’s new normal growth path. Furthermore, the quality of asean institutions differs greatly from authoritarian to democratic regimes. Hill and Jongwanich [2014] note that the determinants of outward fdi from China and other emerging East Asian economies are consistent with standard economic theory. In addition, the desire for natural resource security and exceptionally high domestic savings rates are particularly important factors for fdi in the region. fdi from China is, therefore, another important transmission channel of shocks from China’s economic fluctuations to asean economies, which have abundant natural resources.

The quality of institutions can affect the attractiveness of fdi in asean [Masron and Nor 2013]. The heterogeneity among asean economies implies differences in the effectiveness and efficiency of macroeconomic policy in response to China’s economic slowdown. China rising has caused a shift in global trade patterns, with China dominating Western markets at the expense of asean countries [Napoli 2014]. Nevertheless, China’s dominance does not appear to have had a significant negative effect on growth rates for asean gdp, exports, or fdi stocks. While China is crowding asean out of developed markets, increased Chinese demand for asean imports has more than offset this effect. Now the impact of China’s slowdown has become more pronounced in export sectors in the United States, Europe, and Japan. Since asean’s export sectors have market exposures to those

1 China’s Shanghai Composite index rose by 150 percent (year-on-year) in June 2015. The sharp fall in August 2015 represented a 40 percent fall from the June peak.

194 Bhanupong: The ASEAN business cycle and China’s slowdown

markets, although they do not directly rely on China’s imports, eventually they will be adversely affected by the slowdown in China.

This paper examines how the slowdown in China’s economic growth affects asean economies. If China’s rebalancing policy means increasing consumption and reducing investment expenditure, will the link between China and the asean business cycle become weaker? Which asean economies will be most affected by China’s new normal growth?

The rest of this paper is organized as follows. Section 2 discusses key aspects of the China factor that exerts impact on the asean business cycle. Section 3 examines the implication of exchange rate realignment. Section 4 analyzes the impact of China’s slowdown on asean economies. Section 5 presents simulation results from a vector autoregressive model, which contains China’s imports growth, asean business cycle movements, and asean members’ economic activity. The last section provides concluding remarks.

2. The China factor and the ASEAN business cycle

asean economies experienced similar shocks from the Asian financial crisis in the period 1997-1998 to the global financial crisis in the period 2008-2009. The spectacular falls in output were followed by V-shaped recoveries. The reason behind this co-movement of output can be traced back to similar patterns of exports and investment cycles in asean economies. Athukorala and Hill [2010] point out that dependence on fragmentation-based international specialization is proportionately higher in asean than in North America and Europe. The rapid integration of China into regional production networks does not crowd out asean members’ opportunities for international specialization.

Not only are their output growth paths related, inflation, interest rates, and stock market performance tend to move together in tandem. Since the Asian financial crisis, the saving-investment gaps in asean economies have widened as the current account deficit in the pre-Asian financial crisis turned around into surplus.

The loss of China’s competitiveness is shown by the appreciation of the real effective exchange rate, which weighs down on China’s exports. The export-led growth policy of China has been characterized by an artificially weak exchange rate and subsidized credit by public-own banking system. China has welcomed fdi inflows in order to increase the strength of its industrial development, which has contributed to enormous capacity. When the world could no longer support the growth of China’s exports, China’s exports collapsed spectacularly after the V-shaped recovery in 2010 (Figure 1). Consequently, imports also declined as a result of network trade. The global financial crisis caused the sharp export decline in 2009, but the sharp recovery in 2010 was not sustained. Since then both exports and imports started falling rapidly. Overcapacity among exporting firms would eventually lead to non-performing loans of banks that provided subsidized credit to the Chinese export sector.


Source: Federal Reserve Bank economic data

FIGURE 1. Growth rate of China’ s merchandise trade

In July 2015, China devalued the yuan after its exports declined by almost 9 percent from the same time in 2014. The sharp fall was due to a strengthening US dollar and rising wages. China’s devaluation was an attempt to compensate for the strength of the yuan as it has been pegged to the dollar. Total trade declined by 8 percent for the first ten months of 2015. This dramatic economic slowdown has led to a substantial export shortfall in asean countries.

China’s major sources of imports are Japan and the United States. Both Malaysia and Thailand are also important sources of China’s imports, ranking among the top 10 countries. To gauge the impact of China’s imports on asean economies (Figure 2), we compare the movement of business cycle in asean economies (right axis) with China’s imports (left axis). The asean business cycle is obtained from the first component of principal components of asean’s gdp growth rates between 1990 and 2014. There is a striking similarity between the upturns and downturns of the two magnitudes. A long-run relationship between the two series is confirmed by the Johansen cointegration test. The Granger causality test indicates that the direction runs from asean business cycle to China’s import growth. Booming economic activity in asean can predict the rising demand for imports from China. On the contrary, there is no evidence of cointegration between asean business cycle and the United States’ imports. Indeed, asean business cycle is synchronized with China’s economic activity, i.e., trade and output growth. After the establishment of asean Economic Community in 2016, with diversity in population and economic sizes, income levels, and natural resources, it is expected that synchronization of economic activity among asean members would be more intensified through increasing trade intensity within the asean Economic Community and with China.


Sources: Asian Development Bank data and the author’s calculation

FIGURE 2. The ASEAN business cycle and China’s imports

Another explanation of the business cycle synchronization between asean and China is related to the performance of the Japanese economy. The quantitative easing policy of the Bank of Japan, through its record us$665 billion annual asset-buying scheme, has had only limited impact on output on growth and inflation. Japan has suffered as exports to key trading partner China slumped. Japanese consumption spending is also sluggish as a result of an increase in sales tax in 2014. The deterioration in the global economic outlook, including the developments in China, would delay Japanese companies in expanding business investment and raising wages. If the Japanese economy is still in recession and deflationary mode, the demand for Japanese exports from China will also decline, implying a reduction in imported parts and components from asean countries.

The synchronized business cycle is related to credit cycles and exchange rates. Mohan and Nandwa [2009] find evidence of cointegration among the asean interest rates, and the direction of these interest rates is affected by China’s. This evidence points to the linkage between asean-5 and China. The synchronization can also be explained by the exchange rate channel. asean countries tend to avoid currency appreciation when the dollar trended downward. China also tries to peg the yuan to the dollar. Thus trade between asean and China can be conducted within an environment of stable exchange rates. Tang [2014] shows that, because of the region’s production networks, increasing intraregional exchange rate volatility among Asian economies leads to a decline in intraregional trade, in particular among asean economies.2 Ong and Habibullah [2012] find evidence of a continuous macroeconomic interdependence between asean-5 and China, which would lead to a successful asean-China economic cooperation.

2 Intraregional trade within asean amounted to only 24 percent, compared to 69 percent among the European Community.

ASEAN output cycle (right axis)China’s imports (50%)


The macroeconomic interdependence between asean and China can be demonstrated by co-movements of price levels in the two regions. Employing the principal component of asean inflation rates in 10 member countries, we can obtain the proxy of inflation cycles, as shown in Figure 3. When compared with China’s, it is obvious that price stability between China and asean is related. The long-run relationship between China and asean exist, with the causal relationship running from China’s inflation to asean’s.3 The global food crisis in 2008 led China to ban exports of rice and sent waves of panic to Asia. The inflation rate temporary skyrocketed because of food inflation, followed by export collapse caused by the global financial crisis.

FIGURE 3. Inflation synchronization: China and ASEAN

Source: Asian Development Bank data and the author’s calculation

Tham and Kam [2014] employ a gravity model that captures the effects of trade in parts and components as well as final manufactured goods from asean to China. Devadason [2009] also finds evidence that China’s integration in the asean region increases the size of the asean export market, rather than reducing asean’s export expansion. His result is in line with the fact that, although China has become an important export destination and a source of imports for individual asean 5 countries, there is no reduction in intra-asean 5 trade. From the literature, we can conclude that trade seems to be the most important channel explaining the synchronization between China and asean business cycle.

3 The long-run relationship is confirmed by Johansen cointegration test. The hypothesis that China’s inflation rate does not Granger- cause the inflation cycle in asean is rejected at 0.1 percent statistically significant level.

China inflation

Principal component of ASEAN inflation


3. Implications on exchange rates

The quantitative easing monetary policy of the Fed led to capital inflows to emerging markets, as investors were looking for higher rates of return from portfolio investment. After 2009, the Chinese yuan, the Thai baht, the Philippine peso, and the Malaysian ringgit considerably appreciated against the us dollar. In terms of international competitiveness, the real effective exchange rates of these currencies had appreciated substantially since 2010. The yuan strengthened by 30 percent, the peso by 20 percent, and the baht by 10 percent at the end of 2014 (Figure 4). The loss of international competitiveness has compounded economic problems of export-led growth asean economies.

Source: Federal Reserve Bank economic data

FIGURE 4. Real effective exchange rates

Hooy et al. [2015] find that the yuan real exchange rate has a significant positive impact on asean’s total exports to China. Parts and components exports are sensitive to the yuan depreciation, because of the recent production relocations of multinational corporations from the asean region to China and Vietnam. The huge capital inflows after the quantitative easing operations prompted the asean central banks to buy dollars and to accumulate international reserves so as to prevent currency appreciations. However, the trend of appreciation of asean currencies turned to a rapid depreciation in 2015 (Figure 5). By 2015, the US dollar gained strength as the us economy rebounded and the fear of the Fed’s interest rate hike led to fear of capital outflows from the asean region.


Source: The Economist

FIGURE 5. Asian currency depreciation against the US dollar, October 2015 (% year-on-year)

By November 2015, the Indonesian central bank’s reserves fell by us$1 billion to us$100.7 billion. The decline was due to rising costs to service the government’s foreign debt and the use of reserves to stabilize the Indonesian rupiah, which depreciated by 10.7 percent (year-on-year) in October 2015 (Figure 5). The large depreciation corresponds to a high inflation rate of about 6.4 percent during the same corresponding period. The gdp growth rates in Indonesia and Malaysia slowed down to around 4.8 percent in 2015; while Indonesia experienced current account deficit, Malaysia still maintained current account surplus. The steep depreciation of the ringgit must have been related to capital outflows and political risks.

Vietnam is another country that faces inflation at a higher rate than other asean members. The Vietnamese dong has depreciated gradually in line with the inflation level. However, Vietnam experienced strong economic growth in 2015, helped by robust export growth. The dong depreciated by 6.5 percent in 2015. The gain in Vietnam’s exports to the United States was at the expense of Thailand, which is losing world market share in electrical and electronic equipment, apparel, and garments. This is partly due to rising wages and the appreciation of the baht from 2010 to 2014. Thailand also experienced currency appreciation, although the Bank of Thailand does not intervene in the foreign exchange markets as much as it used to. As a result, the baht depreciated by 9.2 percent (year-on-year) in October 2015.

Vietnam imports input materials mostly from China to serve its local production. Therefore, as the Chinese yuan has depreciated, Vietnam is able to buy materials at lower prices, thus enabling it to cut production costs and raise Vietnam’s competitiveness. Furthermore, the price decreases in the world market,


caused by the weaker demand from China, will also help Vietnamese businesses cut production costs.

Vietnam still has cheap labor, which can be a main factor in attracting fdi away from China, which is experiencing rising wages and economic slowdown. Another factor that attracts fdi to Vietnam is the benefit Vietnam would obtain from joining the Trans Pacific Partnership Agreement (TPP). Vietnam is the country that would benefit the most from the Trans Pacific Partnership agreement as it has the lowest wages among the 12 members.

Since China surprised world markets by devaluating the yuan by 2 percent in August 2015, net capital outflows have reached us$200 billion. The People’s Bank of China is reported to have spent us$229 billion in foreign exchange intervention to prop up the yuan in the third quarter of 2015. There has been speculation that the Fed interest rate hike is imminent as the us economy has been approaching the full employment level in the last quarter of 2015. While the economic recovery in the Eurozone is still protracted and fragile, due to a decline from Chinese imports, the European Central Bank would still maintain easy monetary policy. As a result, the gap between the interest rates in the United States and in the Eurozone will widen in 2016. The dollar may continue to appreciate. Because of the yuan peg with the dollar, there would be speculation of another round of yuan devaluation. In the future, we would observe the similar pattern of Asian currency depreciation as shown in Figure 5.

4. Trade exposure and dependency of China

Trade exposure to external shocks can be measured by the value of total trade volume relative to gdp. In general, the ratio can be thought of as a measure of the degree of openness, reflecting trade liberalization over time. The increasing trend of trade openness also implies increasing risk exposure to external shocks, such as terms of trade and volume of trade shocks. Small countries tend to have higher trade volume relative to gdp, while large countries with large domestic markets would tend to have lower trade exposure to trade shocks from the rest of the world. The high trade exposure ratio bodes well for degree of trade integration, which can serve as an economic growth driver.

Hill and Menon [2011] point out the high degree of vulnerability of the Cambodian economy, which arises from a narrow economic base, a pre-crisis asset price boom, a fragile financial system, and limited defensive economic policy instruments. After extremely rapid economic growth during 2000-2007, it experienced a sharp growth collapse in 2008-2009. The economy has begun to rebound since early 2010, and the crisis episode has provided the government with an opportunity to conduct policy reform by diversifying the economy and creating the preconditions for de-dollarization. From Figure 6, between 2010 and 2014, the average ratio of the volume of Cambodia’s trade to gdp almost reached 100 percent. It is evident that, despite having the highest dollarization level in


Mya

nmar

Indon

esia

China

Philippine

s

Lao

PDR

Brune

i Dar

ussa

lam

Cambod

ia

Thail

and

Mala

ysia

Vietna

m

Singap

ore

the region, dollarization has contributed to a stable price level and favorable environment to trade.

Source: Asian Development Bank

FIGURE 6. Average degree of trade exposure, 2010-2014

Since China is a major trading partner of Myanmar, and bilateral trade with China accounts for 50 percent of total trade, China’s economic growth and imports slowdown would adversely affect Myanmar’s exports. Despite low trade exposure (21.4 percent), the falling price of rice and nationwide floods would compound Myanmar’s economic problems. The gdp growth of Myanmar is highly correlated with China’s total import growth rate (Figure 7), but not as high as Cambodia’s.

According to Bird and Hill [2010], the past two decades of reform in Lao People’s Democratic Republic (pdr) have been largely successful, with accelerating growth despite unfavorable initial conditions such as the country’s being landlocked and having weak institutions and ill-defined property rights. Neighborhood effects have obviously been supportive in Lao pdr’s case, but their importance should not be overstated. Lao pdr’s gdp growth rate is independent of China’s economic activity (Figure 7). Since the major trading partner of Lao pdr is Thailand, its output growth is related more to Thailand’s gdp growth through export channel.

The Philippine growth rate in 2015 also slowed down compared to the high rate in 2014. Similar to Indonesia and Lao pdr, the Philippine economy would not severely affected by China’s slowdown as much as countries which have high overall trade exposure and high dependency on China. The slump in export demand in the Philippines cannot be compensated by currency depreciation of 4.5 percent in 2015. The depreciation was caused by expectations of capital outflows in anticipation of the Fed’s rate hike.


Mya

nmar

Indon

esia

Philippine

sLa

o

Brune

i

Cambod

ia

Thail

and

Mala

ysia

Vietna

m

Singap

ore

Source: Author’s calculations based on Asian Development Bank data

FIGURE 7. GDP growth correlation with China import growth

TABLE 1. China dependency and ASEAN trade exposure

China dependency High trade exposure (% Trade/GDP)

Low trade exposure (% Trade/GDP)

High (Growth correlation: above 0.35)

Brunei (0.42) Cambodia (0.51) Malaysia (0.39) Singapore (0.47) Thailand (0.39)

Myanmar (0.44)

Low (Growth correlation: below 0.35)

Vietnam (0.3) Indonesia (0.19) Lao PDR (0.01) Philippines (0.2)

Sources: Figures 6 and 7

Malaysia is a country with highly open economy and with export-growth driven strategy. Hill et al. [2012] attribute Malaysia’s significant slowdown in investment and growth since the late 1990s to some specific institutional and political problems, rather than the middle-income trap. Thus major policy reforms are required to meet challenges to growth acceleration. In 2014, Malaysia’s output growth rate is highly related to China’s economic growth. The country has high trade exposure and high dependency on China. Malaysia has suffered from losing market shares in declining sectors: electrical and electronic equipment, vegetable oil, and rubber. Indonesia lost its market shares in declining sectors such as rubber, mineral fuels, and palm oil. In the same development, Thailand’s market shares in the growing sector of rice and seafood have deteriorated. Furthermore, to make matters worse, Thailand also lost the market share in a declining sector: rubber.


The only export sector that seems to be gaining market share is the automobile and parts sector. Due to rising wages, Vietnam has been replacing Thailand’s position as a major exporter of footwear, garment, and seafood. Viewed in this light, asean countries have certain degree of complementarity and competition in the area of labor-intensive industry.

The asean economies that depend on commodity exports suffer from the declines in export demand and commodity export prices. The extent of the damage is related to the reliance and concentration of commodity exports. Palm oil, rubber, and rice are primary commodity exports of many asean countries. They have experienced a declining trend since 2010 (Figure 8). The unfavorable terms of trade would continue as long as the crude oil price still remains below us$80 per barrel.4

As Asia’s only major net oil exporter and the world’s second-biggest producer of palm oil, Malaysia’s economic growth has been negatively affected by the weak oil price. The Malaysian government encountered a decline in fiscal revenues. The government derives about 22 percent from energy related business. The political scandal over funds transferred also led to the weakening of the ringgit. A collapse in commodity prices, caused by China’s slowdown in export demand for raw materials, has affected the asean community.

Source: Mundi Index

FIGURE 8. Declining commodity prices

4 Crude oil prices are positively related to the price of natural rubber, which competes directly with synthetic rubber, a byproduct of the oil refinery industry.

Palm oil

Rubber

Rice


In summary, China’s slowdown has reduced the fiscal space of asean countries, thereby impairing the governments’ ability to use fiscal policy to counteract the downturn of economic cycles. When the Fed starts normalizing monetary policy, an increase in interest rates would raise household debts in emerging economies. Rapid increases in household debts to gdp have made it difficult for asean monetary authorities to implement effective monetary policies to stimulate the economy after export collapses. Poverty and indebtedness have been a major problem for farmers since the collapse of primary commodity prices. As long as commodity prices remain weak, the inclusive development goals of poverty eradication and improved income distribution will be a challenge. China’s slowdown has a far-reaching impact on inclusive growth.

5. Impacts of China’s slowdown

We have established that there is a strong causal relationship between China and asean economic activity. In this section, we examined the impacts of the fluctuations on business cycles in asean and China. An unrestricted vector autoregressive model is constructed for this purpose by utilizing the annual data on growth rates in asean economies. The data from 1990 to 2014 are obtained from the Asian Development Bank. As discussed earlier, the first principal component of asean growth rate is employed as a proxy for the economic activity of asean economies as a whole (Figure 2). By utilizing two-period lags, impulse response functions from shocks in asean business cycle and China’s import growth are obtained. The accumulated responses of asean output growth rates are reported. In the first group, we select high-trade exposure countries (Table 1), namely Malaysia, Thailand, Cambodia, and Vietnam. The results reported in Figure 8 confirm the earlier conjecture that countries with high trade openness are likely to benefit more from China’s rapid expansion than countries with lower degrees of trade exposure. Thailand’s growth rates, followed by Cambodia, respond most strongly among the four countries in the first simulation group. On the other hand, during China’s slowdown, Thailand and Cambodia would bear the most burdens of adjustments in terms of growth slowdown. Except for Malaysia, the impact of China’s economic activity is more prominent than economic activity within asean community (as captured by impulse responses to the asean business cycle). For Malaysia, the impact of asean expansion is more pronounced than the China factor. This is because the trade intensity between Malaysia and other asean countries, namely Singapore, is higher than Malaysia’s trade intensity with China. In the case of Thailand, Cambodia, and Vietnam, the accumulated responses of growth of these economies from asean economic activity die off within five years, while China’s impacts on growth rates in high-trade exposure countries are sustained more than a decade. These findings underline the importance of China factor in contributing to long-term output growth in Thailand, Cambodia, and Vietnam since China’s accession to the


World Trade Organization. On the other hand, China’s new normal growth pattern would have a detrimental impact on the long-term growth of countries with high international trade exposure and high dependency on Chinese economy. Hill and Menon [2013] observe that despite Cambodia’s weak institutions and a legacy of history and small size that limit government’s policy space, Cambodia has achieved high economic growth. Through large public and private capital inflows, economic openness, and reasonably prudent macroeconomic management, the Cambodian economy grows rapidly in a dynamic, integrating neighborhood of the asean region. The empirical finding in this paper underlines the importance of the China factor in generating high growth for low-income asean countries since they started engaging in trade liberalization. With the new growth target set for 7 percent in the China’s new development plan for the next five years, heavy reliance on Chinese expansion can be detrimental to their long-term growth, as they cannot depend on China’s strength.

For countries with low trade exposure to external shocks, the impact of China slowdown will be minimal. Figure 10 illustrates that, except for Myanmar, the China factor is less important than the asean factor; economic activity in the asean region produces stronger growth impact. Indonesia, the Philippines, and Lao pdr benefit less from China rising than the expansion of asean economies. Lao pdr’s growth rate is hardly related to China’s growth. Distance matters in trading activity. Lao pdr trades more with Thailand than China. Although Myanmar has low international trade exposure, its gdp growth is highly associated with China’s. Figure 10 demonstrates that Myanmar’s growth increases substantially with rising economic activity in China, whereas it does not benefit much from increasing asean economic activity. This conclusion may change after the investment and trade impacts of the asean Economic Community become more pronounced in the next decade.

Indonesia, which has a relatively low trade exposure, due to its large size, has less influence from the China factor. Basri and Hill [2011] analyze Indonesian growth dynamics, in particular the V-shaped recoveries after the Asian financial crisis and global financial crisis. The years 1997-1998 were regarded as a watershed in the country’s economic history and political economy. The growth of Indonesian economy has never matched that of high-growth East Asian Economies. In Figure 10, Indonesia and the Philippines have not benefited much from China rising, although the two countries’ growth rates respond vigorously in line with expansion of asean economic activity.5

5 See Hill [2015] for a discussion on political economy of policy reforms in Indonesia and the Philippines.



China’s new normal growth implies that the expansion of asean economies would slow down. China’s slowdown has reverberated around the world. Japan was also in recession again in the third quarter of 2015. This is Japan’s fifth recession in seven years. Some asean economies would not be affected, unless public investment can be implemented effectively to induce growth via enhancing effect of public infrastructure, which can lead to crowding in effect on private and consumption expenditures. asean countries are seemingly homogenous, yet each country has heterogeneity in trade exposure, export markets, and export products concentration. When China’s growth declines, it would affect asean economies differently. Countries which have large trade exposure and are highly dependent on China would suffer the most. On the other hand, countries which have low trade exposure and rely less on China’s economic activity would suffer the least from China’s growth slowdown.

The adverse impact of China’s slowdown would be felt less in Cambodia, Lao pdr, Myanmar, and Vietnam. In particular, developing rapid transportation network in the region can compensate for the declining volume of trade with China through increasing trade among asean countries, which would be enhanced by reduced transportation costs. In addition, fdi outflows from asean-5 into Cambodia, Lao pdr, Myanmar, and Vietnam can offset declining export demand from China. Indonesia and the Philippines would be the least to suffer from China’s slowdown. Market size and age structure of population can ensure dynamism in investment and consumption expenditures. Thailand would be the laggard in the region unless a democratic government can be restored. Authoritarian rules reflect extractive institutions that create risks, and uncertainty can drive out investment and lower long-term growth.

In the scenario of global growth slowdown, currency depreciation, and rising interest rates, the main challenge in asean will be household and business sector debts that would eventually impair the quality of bank loans and prolong economic recovery. The road to economic recovery will be hard unless the asean community can increase their intra-regional trade through improving regional connectedness through massive infrastructure investment to compensate for the declining China’s demand for exports.

Thammasat University


FIGURE 9. Impacts of high trade exposure on ASEAN economies

FIGURE 10. Impacts of low trade exposure on ASEAN economies


References

Athukorala, P. and H. Hill [2010] “Asian trade: long-term patterns and key policy issues”, Asian-Pacific Economic Literature 24(2): 52-82.

Basri, M.C. and H. Hill [2011] “Indonesian growth dynamics”, Asian Economic Policy Review 6(1): 90-107.

Bird, K. and H. Hill [2010] “Tiny, poor, land-locked, indebted, but growing: lessons for late reforming transition economies from Laos”, Oxford Development Studies 38 (2): 117-143.

Devadason, E.S. [2011] “Reorganization of intra-asean 5 trade flows: the ‘China factor’”, Asian Economic Journal 25 (2): 129-149.

Hill, H. [2015] “The political economy of policy reform: insights from Southeast Asia”, in I. Coxhead, ed., Routledge handbook of Southeast Asian economics. Oxford and New York: Routledge. 327-344.

Hill, H. [2014] “Is there a Southeast Asian development model?”, Malaysian Journal of Economic Studies 51: 89-111.

Hill, H. and J. Jongwanich [2014] “Emerging East Asian economies as foreign investors: an analytical survey”, Singapore Economic Review 59(3): 1-26.

Hill, H. and M.S. Gochoco-Bautista [2013] “Perspectives and issues”, in H. Hill and M.S. Gochoco-Bautista, eds., Asia rising: growth and resilience in an uncertain global economy. Cheltenham, U.K. and Northampton, Mass.: Elgar; Manila: Asian Development Bank. 3-45.

Hill, H. and J. Menon [2013] “Cambodia: rapid growth with weak institutions”, Asian Economic Policy Review 8(1): 46-65.

Hill, H., T.S. Yean, and R.H.M. Zin [2012] “Malaysia: a success story stuck in the middle?” World Economy 35(12): 1687-1711.

Hill, H. and J. Menon [2011] “Reducing vulnerability in transition economies: crises and adjustment in Cambodia”, asean Economic Bulletin 28(2): 134-159.

Hooy, C.W., S-H. Law, and C. Tze-Haw [2015] “The impact of the renminbi real exchange rate on asean disaggregated exports to China”, Economic Modelling 47: 253-259.

Ishaq, M. and M.A.U. Rehman [2013] “Surmounting the individual: establishing a common currency in Asia--a case study of East Asian economies”, Global Economy Journal 13(1): 63-88.

Masron, T.A. and E. Nor [2013] “fdi in asean-8: does institutional quality matter?” Applied Economics Letters 20(1-3): 186-189.

Mohan, R. and B. Nandwa [2009] “Examining interest rate linkages among asean-5, China and India”, asean Economic Bulletin 26(2): 174-179.

Napoli, C. [2014] “China’s economic rise: implications for asean trade flows”, Journal of Southeast Asian Economies 31 (3): 345-360.

Ong, H.-B. and M.S. Habibullah [2012] “Is China compatible with asean-5? a gradual cointegration analysis”, Journal of Economic Studies 39(3-4): 356-367.


Tang, H.C. [2014] “Exchange rate volatility and intra-Asia trade: evidence by type of goods”, World Economy 37(2): 335-352.

Teng, K.T., S.H. Yen, and S.Y. Chua [2013] “The synchronization of asean-5 stock markets with the growth rate cycles of selected emerging and developed Economies”, The Journal of Applied Economic Research 7(1): 1-28.

Tham, S.Y. and A.J.Y. Kam [2014] “Re-examining the impact of ACFTA on asean’s exports of manufactured goods to China”, Asian Economic Papers 13(3): 63-82.


PRE

Goods trade liberalization under the ASEAN Economic Community: effects on the Philippine economy

Ramon L. Clarete* and Philip Arnold P. Tuaño**

The economy-wide effects of the asean Free Trade Area (afta) on the Philippine economy are computed using the Global Trade Analysis Project (gtap) model of the world economy. Of the 40 industries representing the Philippine economy that were simulated to assess the impacts of afta on sectoral output, 24 industries declined. However, the order of magnitudes of the percentage declines is low, except for rice, whose output decreased by about 4.5 percent.

Notwithstanding the contraction of production in the majority of industries, the country comes out a net gainer in aggregate output by around 1.4 percent in total gross domestic product. This implies that, overall, the Philippines is slightly better off with the preferential trade liberalization, with an equivalent variation gain of us$237.4 million.

Considerable movement of workers across industries is observed. Reductions in skilled worker employment resulted in 31 industries, and 35 industries do the same in the case of unskilled labor. There is no change in unemployment due to the fact that the empirical model assumes full employment of productive factors. Given the fact that there is a change in employment patterns during trade liberalization episodes, it is important to assess the empirical relationship between trade liberalization and unemployment which remains difficult to pin down.

The empirical literature in developing countries shows that the employment response of trade reforms is dependent on infrastructure, trade facilitation measures, and other policies. Therefore, policies related to providing information and new market opportunities, organizing value chains, facilitating coordinated investments along the chain, clustering of related small and medium enterprises to boost external economies, and providing information to employers and workers of job opportunities, including training, are important.

JEL classification: F150 Keywords: ASEAN Economic Community, ASEAN Free Trade Area, trade

liberalization


1. Introduction

This paper examines the economy-wide effects of goods trade liberalization in the Association for Southeast Asian Nations (asean) region on the Philippine economy using the Global Trade Analysis Project (gtap) model of the world economy. Goods trade liberalization is a key reform of asean leaders under the asean Economic Community (aec). By transforming the region into a single market and production base, its leaders seek to make the region more competitive, attract more direct investments, generate more jobs, and increase productivity, trade, and per capita income growth.

However, as in every trade liberalization episode, many view the removal of tariff barriers as a process resulting in cheaper and ostensibly better quality imported products, resulting in the loss of domestic jobs. On the other hand, trade reforms are also seen as creating new markets for domestic products and creating more employment opportunities.

The two effects on employment are likely to happen in the process. Jobs are lost in some industries of the economy, and they are created in other industries. This study also provides a brief assessment of the suitability of general equilibrium economic models in estimating the net impact on jobs of goods trade liberalization under aec.

Hill [2003:232] and Clarete [2006] noted that one of the development puzzles in the Philippine economy is that while trade reforms resulted in marginally better growth levels in the 1990s (at least compared to the previous decade), it did not trigger improvements in manufacturing employment. Several explanations that were explored include supply side constraints in the economy and the comparative lack of demand.

This article is divided into six parts. In the following section, the key policy reforms, including preferential trade liberalization, in aec are discussed. The third section reviews the methodology used in computing the effects of tariff reforms. The results of the ex-ante analysis are in the fourth section of the paper. Considering that the ex-ante analysis abstracts from structural rigidities of the economy, the fifth section takes up the adjustment process to freer trade; the empirical and analytical assessments of the relationship between trade liberalization and employment effects are examined. Key observations and recommendations are provided in the concluding section.

2. Goods trade reforms in AEC

This section provides the context of goods trade liberalization in asean. Even before the 2003 aec declaration at Bali, asean leaders had already declared the region to be a free trade area. aec prohibited non-tariff barriers, introduced rules of origin and other non-tariff measures, and provided trade facilitation measures.

212 Clarete and Tuaño: Goods trade liberalization under the ASEAN Economic Community: effects on the Philippine economy

2.1. The ASEAN Free Trade Area

The agreement that created the asean Free Trade Area (afta), which came into force in 1992, was signed during the period when the world “exploded” with free trade agreements [Soloaga and Winters 2001]. According to the World Trade Organization [2011], there hardly were any preferential trade agreements created between the 1950s and 1980s. It was in the 1990s that the number surged to about 70 agreements, and then it rose to 300 agreements in 2010.

Under the agreement that launched afta, member countries were required to reduce the trade taxes imposed on goods imported from their fellow member countries. Brunei, Indonesia, Malaysia, the Philippines, Singapore, and Thailand, collectively known as the asean-6, initially comprised the asean membership. Cambodia, Lao People’s Democratic Republic (or Lao pdr), Myanmar, and Vietnam, known as the clmv countries, subsequently joined the regional bloc. This Common Effective Preferential Tariff (cept) Scheme, which afta called the system of tariffs that would be imposed on the products imported from the other asean member countries, ensured that the tariff rates would be brought down at zero to five percent rates, in a time frame of eight years from the signing of the agreement.

The rules of origin were also agreed upon in the trade agreement. This meant that the goods, excepting for some commodities, would be subject to preferential trade and should have a local content of at least 40 percent of the freight on board. It was also agreed that quantitative restrictions would be removed on goods whose tariffs would be reduced and all other non-tariff measures related to these products would be reduced in a five-year period.

With the entry of the clmv countries, the free trade agreement was reaffirmed in November 1999 when asean agreed to implement a reduction of tariffs to zero by 2010 for the asean-6 and by 2015 for clmv countries. Products in the priority sectors, especially in the manufactured goods sectors, would be eliminated by 2007 for the asean-6 and by 2012 for the clmv, while all tariffs would be eliminated by 2015 for the asean-6 and by 2018 for the clmv countries.

2.2. Leap to AEC

The transformation of afta into an “economic community” took place in 2003 with the Bali declaration. asean leaders adopted the vision of creating a “single market and production base” in order to make the regional economy “stable, prosperous, and highly competitive” [asean 1997] as one of the three pillars of regional cooperation, including security, sociocultural integration, and economic integration, as agreed upon in the asean Concord (Bali Concord II).

During the asean Summit in Cebu in January 2007, which developed the aec Blueprint, the deadline for integration was brought forward from 2020 to 2015. Because of this, the asean Trade in Goods Agreement (atiga) was


signed in 2009, which formalized the tariff agreements that have been made in the past and improved the transparency and predictability of changes in tariffs. atiga also emphasized trade facilitation measures with a work plan to be put in place from 2009 to the start of aec in 2015.

2.3. Goods trade reforms under AEC

atiga also emphasized trade facilitation measures with a work plan to be put in place from 2009 to the start of aec in 2015. Besides tariff reduction, the following are the significant provisions of atiga and the ancillary agreements: elimination of non-tariff barriers [De Dios 2007]; continuous reform of the rules of origin rules; implementation of harmonized trade facilitation processes; establishment of an asean single window to expedite clearance processes for trade flows; and harmonization of standards and technical barriers to trade.

2.3.1. Tariff reforms

The afta agreement came into force in 1992, which aimed, among other purposes, to reduce intra-regional import tariffs to no more than five percent under the cept Scheme. However, the scheme allowed members to draw up their respective Inclusion List, Sensitive List, Highly Sensitive List, General Exclusion1 List, and tariff reduction schedules. The new member states—Cambodia, Lao pdr, Myanmar, and Vietnam (or clmv)—were given flexibility in terms of a longer implementation period than the asean-6.

In 1995, asean leaders decided to accelerate the cept process by moving the completion date from 2008 to 2003. The tariff rates on sensitive imported products from the region were eventually phased into the cept process. Member states were legally bound to reduce the tariff rates on these products to no more than 5 percent. For asean-6, the target year of completion of the cept process for Sensitive List imports was 2010 and 2015 for clmv states.

The 2007 aec Blueprint affirmed the agreements of member states regarding the parameters of the preferential tariff reforms, i.e. the target rates, deadlines, and schedules of tariff reduction. Nonetheless, the blueprint emphasized the urgency of these reforms and desirability of minimizing the cases of departure from the agreed tariff reduction schedules. The blueprint extended the cept Scheme to 2015 for asean-6 and up to 2018 for clmv states to cover for the integration of Sensitive List imports into the cept Scheme, as well as to set the final tariff rates on Highly Sensitive List products.

1 General Exclusion List products are permanently exempted from the tariff reduction process to uphold national security, public morals, and public health, as well as protect the environment and articles of artistic, historic, or archaeological value (asean Secretariat, 1999).


As part of its commitments to asean, in December 2009, the Philippines passed Executive Order No. 8502 which removed tariffs on imports from asean, except for products in the Sensitive List and Highly Sensitive List. Approximately 94 percent of tariff lines in the country’s Tariff Reduction Schedule is already set to 0 percent.

Agricultural tariffs remain high in the Philippines. The average tariff of dutiable agricultural and fisheries products in 2013 is 13.01 percent, which is set to fall to 10.23 percent by 2015. This is due to the feature of the country’s cept that reduced the tariffs of its Sensitive List imports down to only 5 percent, which is an acceptable ending rate of the cept. The Philippine Sensitive List includes swine, poultry, cassava, sweet potatoes, corn, grain sorghum, and sugar for the detailed list. Except for sugar, the tariff rates of products in the Sensitive List are already down to 5 percent since 2010. Rice is the only item in the Philippine Highly Sensitive List. Based on the Tariff Reduction Schedule, rice and sugar tariffs go down to 35 percent and 5 percent respectively in 2015.

The reduction and elimination of tariffs in asean has been considered successful. As of 2010, intra-asean tariff rates were virtually zero in the asean-6, and 2.6 percent was the mean preferential tariff rate under atiga in the newer clmv member states. By 2015, rates on 98 to 100 percent of all tariff lines are expected to be in the 0-5 percent range. Table 1 shows the percentage of tariff lines with 0 percent tariff as of 2013 for each of the asean countries. According to asean Secretariat, 99.85 percent of regional tariffs of asean-6 member states are zero, while the corresponding number for newer asean members was 69 percent. About 30 percent of the goods in the clmv currently have rates greater than zero, while slightly over a third of a percent is observed for the asean-6. The residual share of tariff lines is for those not offered for preferential reduction, and these tariffs apply to products in the General Exclusion List. The General Exclusion List tariff lines for the Philippines is nearly a third of slightly over a quarter of 1 percent and 0.45 percent or the asean-6 states.

2.3.2. Elimination of non-tariff trade barriers

Quantitative restrictions to imports, whether explicit or not, are non-tariff trade barriers (ntbs) and are thus prohibited under atiga. Under the aec Blueprint, all member states are to commit not to expand the number of ntbs they maintain and to roll back any that they currently implement. In the interest of promoting transparency, the asean member states commit to notify the asean Secretariat

2 “Modifying the Rates of Duty on Certain Imported Articles as Provided Under the Tariff and Customs Code of 1978, As Amended in Order to Implement the Commitment to Eliminate the Tariff Rates on the Remaining Products in the Inclusion List in Year 2010 Under the Common Effective Preferential Tariff (CEPT) Scheme for the ASEAN Free Trade Area (AFTA)/ASEAN Trade in Goods Agreement (ATIGA)”


of their respective use of ntbs. The secretariat has to maintain an effective surveillance mechanism to monitor the implementation of ntbs and non-tariff measures. Member states are legally bound to eliminate all ntbs by 2010 for asean 5, and by 2012 for the Philippines, although this had already changed because of the June 2014 waiver.3 The clmv states have until 2018 to remove the ntbs that they maintain.

TABLE 1. Percentage of tariff lines at 0 percent in the ATIGA tariff schedule of 2013

Percentage of tariff lines (%)

Country 0% Greater than 0% Other1

Brunei 99.27 - 0.73

Indonesia 98.87 0.17 0.96

Malaysia 98.74 0.59 0.66

Philippines2 98.62 1.11 0.27

Singapore 100.00 - -

Thailand 99.85 0.15 -

ASEAN-6 99.20 0.35 0.45

Cambodia 40.77 59.23 -

Lao People’s Democratic Republic 78.73 20.36 0.91

Myanmar 79.66 19.69 0.65

Vietnam 72.24 25.77 1.99

CLMV 68.88 30.20 0.92

Source: ASEAN Secretariat [undated]

2.3.3. Trade facilitation

With tariff rates and ntbs eliminated in the region, the agenda for closer integration in goods trade is increasingly focused on the harmonization and efficient administration of the various regulations and para-taxes that affect international trade.

asean [2011] determined the major ntbs in the region. The first category comprises customs surcharges. In a survey conducted, asean reported that about 70 percent of the ntbs implemented are customs surcharges. Technical measures, another category, make up 14 percent, followed by product characteristics requirement, 10 percent. The remaining 6 percent of observed ntbs included charges other than those collected by customs, state trading and single channel rules, marketing requirements, and technical regulations.

3 The Philippines presently has a waiver from the World Trade Organization to continue to maintain the country’s quantitative restrictions on rice imports in the form of the monopoly of rice imports under the National Food Authority. The 2012 deadline for the Philippines to eliminate its non-tariff barriers reflected the country’s original plan not to extend its quantitative restrictions for rice after it would have expired in 2012. The government changed plans and decided to apply for a waiver from the World Trade Organization.


Non-tariff measures regulate the flow of trade to act on observed problems that international trade may inadvertently cause. An important category is the set of sanitary and phyto-sanitary standards and regulations on traded plant and animal-based products. These ensure that imported foods, feeds, planting materials, and breeding stocks are safe for consumption or for use by the public, or that their entry into the country does not introduce diseases and pests to native plants and animals.

In non-agricultural products, governments require that imports meet their respective standards and conform to existing regulations. These standards and regulations mitigate the risk of substandard imports that may jeopardize public interest.

Because of the preferential nature of trade liberalization under asean, rules of origin have become part of the free trade area rules to determine eligibility and to mitigate risk of trade deflection. atiga states that the administration of rules of origin be kept simple and continuously responsive to the requirements of private businesses in the region, particularly with respect to promoting regional and global value chains.

The aec Blueprint has also called for the modernization of customs procedures and regional integration of national customs administration, which are set in the asean cargo clearance and customs declaration document [asean 2007].

Another important contribution to facilitating trade is the harmonization of standards, technical regulations, and conformity assessment procedures. The aec Blueprint calls on member states to work together to develop and implement such harmonized systems of standards and technical regulations based on international best practices. Trade facilitation programs, required by atiga, are intended to avoid or mitigate any adverse effects of administering non-tariff measures, whether it is before the goods arrive or are exported, during cargo clearance at customs, or after cargoes are cleared.

Another area that aec is focusing on is improving transparency of regulations. aec calls for the establishment of a region-wide asean Trade Repository of existing trade regulations, where information on regulations of member states is accessible to stakeholders.

The aec Blueprint provides for measures that reduce trade costs. It calls on member states for collective action through the asean Secretariat to undertake studies on trade and to implement a comprehensive work program in order to simplify, harmonize, and standardize trade licensing and cargo clearance processes and procedures.

3. Structure of the model

A computable general equilibrium model of the global economy is used in analyzing the impact of aec tariff reforms. The model allows the tracking of the economy-wide effects of changes in asean tariff protection in a way that takes


into consideration their effects on global markets of products and in turn their feedback to its component national economies.

3.1. Production structure

The model comprises 40 production sectors and 23 countries/regions of the world. Table 2 lists and classifies the production sectors of the model by broader sector categories that they belong to, namely agriculture, natural resources, manufacturing, and services. There are nine primary agricultural production activities. Forestry, mining and oil, and gas are three natural resource extraction activities. There are 19 manufacturing activities and nine services industries.

Each of the 40 sectors produces only one unique product. The production activity for each sector is modeled by a nested Leontief-constant elasticity of substitution production function, i.e., there are several input-to-output transformation processes at various levels of the production activity (see Figure 1).

At the top level, the activity’s output is a nested function of two composites, one for intermediate inputs and the other for primary inputs or value added. The former in turn depends upon another set of composites, called Armington [1969] goods, with each good or service made up of a locally produced input and its imported equivalent, which are substitutable subject to a constant elasticity of substitution parameter.

Value added is a function of primary factors of which there are four in the model: labor; skilled labor; land; and capital. These inputs are constant elasticity of substitution-substitutable among themselves, i.e., they substitute with each subject to a constant elasticity of substitution parameter. All primary factors are free to be moved from one industry to another. However, land is an input in only agricultural production activities, and it is freely movable among the latter.

The outputs are used for final consumption, exports, and intermediate inputs into Armington composites. Both final and intermediate uses of the product are sales to the domestic markets. Output is split between domestic sales and exports through a constant elasticity of transformation parameter.

3.2. Demand structure

Twenty-three countries or regions represent the global economy (see Table 4). Fifteen of these regions are in Asia, with Southeast Asia having the most number of individual countries represented. Of the 10 asean member states, Cambodia and Lao People’s Democratic Republic are grouped as one region, while Myanmar, Brunei Darussalam, and East Timor make up the rest of Southeast Asia.


Local sales Export

Output

Intermediate input composite Value added

Low skilled labor

High skilled labor

Land CapitalAG1 AG2 AG40

LP1 MP1 LP2 MP2LP40 MP40

TABLE 2. The production sectors of the model

1. Paddy rice A 21. Wood products M2. Cereals A 22. Paper and publishing M3. Oil seeds A 23. Leather M4. Sugarcane and beets A 24. Chemicals, rubber, and plastic M5. Vegetables and fruits A 25. Petroleum and coal M6. Other crops A 26. Non-metal mineral products M7. Other animal products A 27. Metal products M8. Cattle A 28. Machinery products M9. Fishery A 29. Electrical products M10. Forestry NR 30. Transport equipment M11. Mining NR 31. Other manufacturing M12. Oil and gas NR 32. Construction S13. Meat preparations M 33. Fuel, electricity, and water S14. Dairy M 34. Transport services S15. Vegetable oils M 35. Trade S16. Processed rice M 36. Communications S17. Milled sugar M 37. Financial intermediary S18. Other food products M 38. Public administration, education, and health S19. Beverages and tobacco M 39. Real estate and commercial services S20. Textile and garments M 40 Rest of services S

AbbreviationsA: agriculture NR: natural resources M: manufacturing S: services

Notes LP: local product MP: imported product AG: Armington good

FIGURE 1. Structure of production activities of the model


TABLE 3. The geographical regions of the model

1. Indonesia SEA 13. India SA2. Cambodia and Lao People’s

Democratic Republic SEA 14. Rest of South Asia SA

3. Malaysia SEA 15. Australia and New Zealand ANZ4. Philippines SEA 16. Canada A5. Singapore SEA 17. United States of America A6. Thailand SEA 18. Brazil A7. Vietnam SEA 19. Rest of North and South Americas A8. Rest of Southeast Asia SEA 20. European Union 27 EU9. China EA 21. Middle East and North Africa MENA10. Hong Kong and Taiwan EA 22. Rest of Africa Af11. Japan EA 23. Rest of the World ROW12. South Korea EA

Abbreviations SEA: Southeast Asia EA: East Asia SA: South Asia A: Americas

Each region has a representative private household and a government. Both are treated in the model as maximizing a utility function subject a budget constraint, with the latter having a Cobb-Douglas utility function.

The representative regional consumer draws its income from its ownership of the four primary factors of the model, and income transfers from the government and the rest of the world. After deducting the taxes it pays to the government, the consumer then apportions the disposable income to savings and consumption. The government’s income comes from taxes collected and spends it on local and imported products. Savings in the model are pooled by a global financial intermediary, and are allocated to the various regions based on their respective investment financing requirements.

3.3. Equilibrium conditions

The computable general equilibrium modeling structure of gtap entails the latter observing accounting relationships that correspond to the various conditions that define when the global and regional economies are in a state of equilibrium or balance, utilizing a regional household and a global bank.

The regional household receives and allocates income and expenditure flows. It receives the incomes of the primary factors in consideration of the use of the respective services of these factors in producing goods and services. It also receives taxes, paid by the private household and the government when they buy goods and services, of which there are two types. Internal income and indirect tax revenues are those from taxes on local purchases, and trade taxes are those from imports and exports of goods and services. The border taxes are the revenues from import tariffs, export taxes, and other related tax measures collected at the border.


The regional household spends all its income on two income transfers and savings. One of the two transfers goes to the private household, and the second is given to the government. These transfers then figure out as the respective budgetary constraints of the two when they maximize their respective utility functions as discussed above. The savings that the regional household generates goes to a global bank. The bank pools the savings from all regional households of the model and allocates the total across various competing uses of the savings by the regions. The manner how this is allocated is that all regions with competing uses of savings face the same terms of use.

3.4. Global Trade Analysis Project data and model

The data used in the study comes from version 8. 1 of the gtap data [in Narayanan, Bardi, and McDougall 2012]. The gtap 8.1 data set comprises a total of 137 countries and 57 sectors. The Center for Global Trade Analysis at Purdue University maintains and regularly updates the data set since 1993 [Hertel 1999]. The baseline year is 2007.

This study used the gtap model in computing the impacts of aec tariff reforms. It is a multi-regional and multi-sectoral model, and it is used in counterfactual analysis such as on the possible implications of the policy changes (Hertel [1999]; Brockmeier [2001]). The gtap model is set up as one that can be solved using gempack [Harrison and Pearson 2002]. gempack solves for percentage changes of the economic variables following a change of policies.4

4. Economic effects of AFTA

4.1. Production effects

Table 5 portrays the changes of production outputs, measured in billions of pesos, for each of the 40 production sectors of the gtap model. These changes reflect how the resources of the economy get to be reallocated in the economy as a result of lower preferential tariff protection in asean.

With lower protection, Philippine industries lose their domestic markets to imported substitutes and thus reduce their outputs, as shown in the Table 4. Altogether, 24 industries, or more than half of the total, are observed to contract production.

The order of magnitudes is in the range from less than a percent to up to 4.5 percent. These are rather low changes, except for rice, whose output decreased by about 4.5 percent. This is despite the fact that rice continues to have the highest asean tariff rate of 35 percent.

4 An alternative solution technique is to solve for the counterfactual values of the economic variables of the model. For this, modelers use the GAMS software, and a version of the model in GAMS was developed by Rutherford [2005].


TABLE 4. Effects of AFTA on production, by industry, Philippines* (in billions of pesos)

Sectors Base value

Change % Sectors Base

value Change

%

Paddy rice 142.36 -4.58 Wood products 86.28 -1.63

Cereal 56.64 -0.40 Paper and publishing 96.28 0.07

Oil seeds 35.84 -1.02 Leather 30.28 -0.94

Sugar cane and beets 25.48 -1.83 Chemicals, rubber, and plastic 327.6 -0.86

Vegetables and fruits 140.72 -0.18 Petroleum and coal 260.56 0.30

Other crops 43.96 15.13 Non-metal mineral products 87.28 0.28

Other animal products 227.96 0.17 Metal products 341.76 0.90

Cattle 28.12 0.35 Machinery products 438.24 -1.36

Fishery 180.08 -0.03 Electrical products 1912.04 -1.56

Forestry 14.52 -1.15 Transport equipment 246.32 16.54

Mining 92.52 -0.92 Other manufacturing 117.32 -0.62

Oil and gas 13 -0.91 Construction 388.4 1.29

Meat preparations 278.56 0.06 Fuel, electricity, and water 388.4 0.00

Dairy 52.56 0.45 Transport services 585.64 0.08

Vegetable oils 83.52 -1.82 Trade 1232.68 0.14

Processed rice 141.28 -4.78 Communications 213.92 -0.12

Milled sugar 43.8 -1.95 Financial intermediary 396.48 0.11

Other food preparations 430.64 -0.17 Public administration, education, and health 812.72 -0.05

Beverages and tobacco 130 4.62 Real estate and commercial services 724.72 -0.32

Textile and garments 278.56 -0.97 Rest of services 217.48 -0.16

All sectors 11,344.60 1.41

Note: Shaded entries are those for industries with increases of their respective outputs. Source: Authors’ computations

The economic resources they lost went to the 15 industries.5 Transport equipment increased the greatest with an increase of 16.54 percent. Although its component production activities have very low base values, the other crops industry increased its output by 15.13 percent. Beverages and tobacco comes second with growth of 4.62 percent. The overall production of the economy went up by 1.41 percent. In the non-agricultural industries and services, transport equipment and construction show significant expansion of their outputs.

The Philippines, despite the output contraction of 24 or more than half of its industries, comes out a net gainer in aggregate output. With a base value of aggregate output at P11.3 trillion, the country comes out of this tariff liberalization process as gaining by 1.41 percent. The gains more than outweigh the losses in aggregate production in the economy.

5 No change in output is observed for the fuel, electricity, and water industry.


4.2. Employment effects

Given the changes in output, considerable movement of workers across industries is observed from the results (see Table 5).6 Thirty-one (31) out of forty industries contract employment of skilled workers, and 35 industries do the same in the case of unskilled labor. In agriculture and natural resources, ten out of 12 industries lay off skilled workers. Rice paddy, sugar cane, and oil seeds cut back on skilled labor employment by at least 1 percent. The other crops industry appears to absorb those displaced with an expansion of hiring at the rate of 16.22 percent.

Thirteen (13) out of 20 manufacturing industries give up skilled workers to other industries. In the case of low-skilled labor, 15 industries are observed to do the same. The industries with the largest layoff rates are agriculture-based, namely milled rice, milled sugar, and vegetable oils. Six and seven industries in the case of skilled and low-skilled workers, respectively, follow, whose layoff rates are between a percent and two, except for the wood products industry that cut back on low-skilled workers employment by 2.13 percent. The rest of manufacturing cut back on employment by less than a percent.

The manufacturing industries that increase employment of skilled workers are the following: dairy; beverages and tobacco; petroleum and coal; non-metal mineral products; metal products; transport equipment; and construction. The same industries increase as well their employment of low-skilled workers, except for petroleum and coal, non-metallic mineral products, and dairy industries, which shed off jobs albeit at less than a percent.

All eight services industries lay off workers, but the rates of change are all less than a percent. This is due to the fact that there are services industries that contracted output; these include communication, public administration, real estate and commercial services, and the rest of services. These output cutbacks may explain the reduction in employment reported in Table 5. However, the remaining three services industries expanded production, and fuel, electricity and water utility neither contracted nor expanded their output. But all four services

industries gave up workers despite the lack of output contraction.

6 In Table 6, entries colored green show the industries that increase employment of workers. Under the allocation columns, shares that are less than a percent are not shaded. Those industries having shares between 1 and less than 4 are shaded light red, while light green is used to shade entries with shares of at least 4 percent.


TABLE 5. Employment changes and allocation of labor by industry, baseline and simulated levels due to AFTA tariff reforms (in %)

IndustryEmployment changes

Share of sector in employmentSkilled workers Low-skilled workers

Skilled workers

Low-skilled workers Baseline AFTA Baseline AFTA

Paddy rice -5.29 -5.46 0.004 0.004 0.251 0.237Cereal -0.77 -0.95 0.029 0.029 0.566 0.561Oil seeds -1.44 -1.63 0.011 0.011 0.239 0.235Sugar cane and beets -2.32 -2.51 0.003 0.003 0.076 0.074Vegetables and fruits -0.53 -0.71 0.052 0.052 1.687 1.675Other crops 16.22 16.01 0.035 0.041 0.625 0.725Other animal products -0.15 -0.34 0.021 0.021 0.623 0.621Cattle 0.04 -0.15 0.034 0.034 0.693 0.692Fishery -0.35 -0.54 0.006 0.006 0.233 0.232Forestry -1.46 -1.64 0.013 0.013 0.336 0.330Mining -1.24 -1.41 0.234 0.231 0.665 0.655Oil and gas -1.45 -1.64 0.310 0.305 0.660 0.649Meat preparations -0.16 -0.44 0.167 0.167 0.516 0.514Dairy 0.23 -0.06 0.114 0.115 0.322 0.322Vegetable oils -2.04 -2.32 0.035 0.035 0.103 0.101Processed rice -4.99 -5.26 0.020 0.019 0.065 0.061Milled sugar -2.17 -2.45 0.037 0.036 0.093 0.091Other food preparations -0.39 -0.67 0.659 0.656 1.371 1.362Beverages and tobacco 4.38 4.07 0.279 0.291 0.568 0.591

Textile and garments -1.18 -1.47 0.479 0.473 1.505 1.483

Wood products -1.83 -2.13 0.364 0.357 0.974 0.954Paper and publishing -0.15 -0.45 1.214 1.212 1.721 1.713Leather -1.15 -1.44 0.089 0.088 0.291 0.287Chemicals, rubber, and plastic -1.08 -1.37 2.173 2.150 2.670 2.634Petroleum and coal 0.09 -0.21 0.110 0.110 0.170 0.169Non-metal mineral products 0.06 -0.24 0.479 0.479 1.096 1.094Metal products 0.69 0.39 1.613 1.624 3.356 3.369Machinery products -1.57 -1.87 3.957 3.895 4.257 4.177Electrical products -1.77 -2.07 0.908 0.892 1.170 1.146Transport equipment 16.28 15.91 1.768 2.056 2.575 2.985Other manufacturing -0.83 -1.11 0.377 0.374 0.891 0.881Construction 1.10 0.78 4.957 5.011 10.189 10.269Fuel, electricity, and water -0.23 -0.53 1.434 1.430 1.468 1.460Transport services -0.10 -0.45 2.337 2.335 4.807 4.786Trade -0.04 -0.39 7.507 7.504 14.894 14.836Communications -0.35 -0.64 2.289 2.281 1.389 1.380Financial intermediary -0.12 -0.42 9.481 9.470 4.861 4.841Public administration, education, and health -0.30 -0.59 36.753 36.643 20.660 20.537

Real estate and commercial services -0.55 -0.84 15.348 15.264 8.930 8.855

Rest of services -0.39 -0.68 4.300 4.283 2.432 2.416Total 100.000 100.000 100.000 100.000

Note: Shaded entries are those for industries with increases of their respective outputs. Source: Authors’ computations


Note: Agriculture, fishery and natural resources (ANFR)Source: Table 5

FIGURE 2. Allocation of skilled and low-skilled labor, base and simulated levels due to AFTA tariff rates (%)

4.3. Effects on trade

Table 6 shows the changes to the country’s exports and imports that afta may bring about. In agriculture, fishery and natural resources, the top exporting industries are vegetables and fruits, mining, other crops, and fisheries industries.

Altogether, the agriculture and fisheries industries come out with a positive net gain in exports at us$183.64 million. The country’s exports in vegetables and fruits decline by nearly a percent. Other crops industry registered to have the highest expansion rate at 280.8 percent. The base value is relatively significant, but this industry is a collection of several crops not elsewhere specified, each of which may have relatively low levels of exports.

The three natural resources industries suffer cuts of their respective export levels, except for oil and gas. This sector’s gain is inadequate to offset export losses in mining and forestry. Mining is a far significant exporter than oil and gas. The loss of natural resources exports amounts to us$10.7 million, mostly in mining.

It’s in the manufacturing sectors that the country gains relatively the most in exports, altogether nearly us$405.4 million. Nine out of 20 industries are observed to have reduced their exports. The biggest exporter, electrical products, with base export value of more than us$41.8 billion, experience a 1.59 percent decline of its exports.

The gainers more than offset the losses of the three large export performers. Automotive spare parts, under transport equipment, have a significant gain, 47.87 percent, and their base export value is about us$2 billion. Although it expands only by a percent, metallic products industry has a large base export value of us$3 billion.

120

100

80

60

40

20

0

19.80

79.45

0.7520.04

79.21

0.75 33.91

59.44

6.65

34.20

59.11

6.69

Base level, skilled labor

AFTA level, skilled labor

Base level, low skilled labor

AFTA level, low skilled labor

Industry AFNR Servces


TABLE 6. Effects of AFTA on Philippine exports and imports, by industry* (in million US$)

Industry Base export values

Change (%)

Base import value

Change (%)

Paddy rice 0.0 9.09 0.0 -6.77Cereal 1.8 0.4 1.9 0.32Oil seeds 1.5 -0.59 1.5 0.47Sugar cane and beets 0.0 0.5 0.0 -0.55Vegetables and fruits 792.0 -0.81 943.0 1.26Other crops 68.0 280.79 73.2 15.29Other animal products 11.4 -1.01 11.9 0.58Cattle 0.0 -3.52 0.0 1.92Fishery 121.0 -0.63 136.0 3.3Forestry 3.8 -2.39 4.2 0.15Mining 1,295.0 -0.82 1,682.0 1Oil and gas 0.4 2.9 0.4 0.34Meat preparations 56.9 9.82 58.0 2.16Dairy 146.0 7.47 159.0 1.19Vegetable oils 768.0 -1.22 808.0 2.02Processed rice 21.9 3.96 21.9 20.62Milled sugar 90.4 1.58 98.2 36.97Other food preparations 1,043.0 0.86 1,138.0 2.8Beverages and tobacco 165.0 76.79 172.0 1.35Textile and garments 2,648.0 -1.11 2,792.0 0.97Wood products 1,112.0 -1.38 1,237.0 3.8Paper and publishing 228.0 3.1 252.0 1.42Leather 151.0 -0.27 163.0 2.13Chemicals, rubber, and plastic 1,687.0 -0.48 1,833.0 0.99Petroleum and coal 812.0 4.23 860.0 0.88Non-metal mineral products 298.0 0.5 350.0 2.95Metal products 3,028.0 1.0 3,119.0 2.14Machinery products 5,415.0 -1.55 5,613.0 1.74Electrical products 41,858.0 -1.59 42,331.0 -0.96Transport equipment 2,094.0 47.87 2,190.0 6.76Other manufacturing 527.0 -2.07 556.0 2.28Construction 113.0 -1.46 113.0 2.05Fuel, electricity, and water 141.0 -2.53 141.0 1.26Transport services 3,160.0 -0.43 3,160.0 0.46Trade 651.0 -1.83 651.0 1.08Communications 579.0 -1.98 579.0 1.06Financial intermediary 349.0 -2.26 349.0 1.28Public administration, education, and health 367.0 -2.16 367.0 1.15

Real estate and commercial services 2,653.0 -2.29 2,653.0 0.96

Rest of services 470.0 -1.89 470.0 0.08All Sectors 72,927.1 0.62 75,088.2 0.32

Note: Shaded entries are for industries that expanded their levels of trade due to AFTA tariff reforms. Source: Authors’ calculations


Textile and garmentsBeverages and...

Other food...Milled sugar

Processed riceVegetable oils

DairyMeat preparations

Oil and gasMining

ForestryFishery

CattleOther animal products

Other cropsVegetables and fruitsSugar cane and beet

Oil seedsCereal

Paddy rice

4.74

-0.3 0.2 0.7 1.2

Rest of servicesReal estate and ...

Public...Financial intermediary

CommunicationsTrade

Transport servicesFuel, electricity, and...

ConstructionOther manufacturingTransport equipment

Electrical productsMachinery products

Metal productsNon-metal mineral

Petroleum and coalChemicals, rubber

LeatherPaper and publishing

Wood products

-0.3 0.1 0.5 0.70.3-0.1

All eight services industries have reduced exports, and together lose about us$126 million. Transport services and real estate and commercial services industries are among the country’s largest services exporters. The former declines by less than half of a percent, while the latter’s exports falls by 2.29 percent. Five out of the eight services sectors increase their export value by at least 2 percent. Summing up, the country gains us$452.36 million. The manufacturing industries take the lead in the country’s exports, followed by agriculture and fisheries.

Source: Authors’ calculations

FIGURE 3. Effects of AFTA on product prices (in %)

Source: Authors’ calculations

FIGURE 4. Effect on returns to primary factors (in %)

0.60

0.40

0.20

0.00

-0.40

-0.20

Land Skilled Low skilled

Capital


In contrast, the country’s import bill rises by us$241.26 million. The majority of the import activities expanded their activities, except for paddy rice, sugar cane, and electrical products. Electrical products, metallic products, machinery, textiles and garments, and transport equipment are the country’s largest importers among the manufacturing industries. The reduction in imports in electrical products is the largest, and this result appears consistent with lower exports coming out of this industry. The overall exports increased by about .62 percent, while imports expanded by .32 percent. The country has a base trade deficit of us$2,161 million. In the simulated trade levels, the deficit is observed to go down by us$211 million.

4.4. Effects on prices

Figures 3 and 4 show the changes of the prices of goods and returns to the primary resources used in production. In Figure 4, all goods, except 5, gain increases in prices. That is, simulating the preferential tariff reforms results in pushing up world prices, albeit at different rates. The highest rate of increase, 4.74 percent, is observed for other crops. The order of magnitudes of price increases is up to about half of a percent.

Figure 4 shows the changes in returns to owners of resources used as primary inputs in the various production activities of the economy. Wages of both skilled and low-skilled workers rise by nearly half of a percent. Land rents, however, go down by about a fourth of a percent. Land is mostly used in agriculture and particularly in rice.

4.5. Overall economic effects

The changes in the country’s gross domestic product and economic welfare are shown in Table 8, which provides information on the gross domestic product at constant prices which is the total value added generated by an economy. This is presented in million US dollars as the changes in other countries in the gtap model are likewise shown for ease of comparison with that of the Philippines. As expected, asean member states are better off with afta. The percentage changes are small, and this reflects the fact that the base year of gtap is 2007 which is already just about the end of the cept process.

Is the Philippines better off with the preferential trade liberalization in goods under aec? The equivalent variation, which measures the amount of money a country is willing to pay in order to be as well-off as having afta under aec, in the last column of Table 7 is used to indicate the economic well-being of the country. Singapore tops the gainers, with us$2.4 billion, followed by Malaysia and the Philippines. The Philippines has an equivalent variation of us$237.4 million. With this figure, the reforms marginally benefit the country.


TABLE 7. Changes in real GDP and economic well-being, by country (in US$ million)

Country Real GDP Economic welfare (in equivalent variation of income)

Country

Real GDP Economic welfare (in equivalent variation of income)

Base % Base %

Indonesia 432,103.19 0.019 64.87 India 1,232,816.38 -0.008 -328.49Cambodia and Lao People’s Democratic Republic

12,644.03 0.194 182.27 Rest of South Asia 266,235.88 -0.002 -32.80

Malaysia 186,642.11 0.145 430.37 Australia and New Zealand 995,227.63 -0.002 -101.61

Philippines 144,070.47 0.052 237.41 Canada 1,424,062.63 0.000 -15.45

Singapore 176,759.66 0.037 2,389.27 United States of America 14,061,782.00 0.000 -601.56

Thailand 247,109.83 0.135 210.79 Brazil 1,365,983.25 -0.001 -47.33

Vietnam 68,435.25 0.022 24.31Rest of North and South Americas

2,470,987.75 0.001 36.62

Rest of Southeast Asia 28,601.50 0.172 76.41 European

Union 27 17,003,710.00 -0.001 -767.54

China 3,494,058.00 -0.002 -421.73Middle East and North Africa

2,530,414.00 -0.001 183.00

Hong Kong and Chinese Taipei 600,833.13 -0.003 -191.83 Rest of Africa 879,130.94 -0.001 -16.45

Japan 4,377,944.00 -0.004 -827.83 Rest of the World 2,782,553.00 0.002 111.00

Source: Authors’ computations * Shaded entries refer to industries with positive gains in real GDP and equivalent variation of income.

5. Adjustment process

This section identifies several issues in the country’s adjustment process to determine the complimentary policies and programs of the Philippine government to fully realize the economic benefits of goods trade integration. It basically asks the question: What effects may the Philippines expect if all the goods trade liberalization reforms under aec are fully implemented without regard as to when those effects are going to take place?

Does asean integration help the Philippines create jobs? As with every trade liberalization process, local workers, farmers, and business people view the economic integration undertaken through aec to be a process whereby local products will be substituted out in the local markets by cheaper and ostensibly better quality imported products. This implies loss of domestic jobs, travel to look for jobs overseas, and adjustment costs.

However, local exporters see the economic integration in asean as providing them new markets for their products and creating more and new jobs for the country. With the talk about aec supporting the formation of regional


value chains, local exporters view this process as bringing them into a more cooperative—rather than competitive—process.

The two effects on employment are likely to happen in the process. Jobs are lost in some industries of the economy and created in others. As conventionally structured, general equilibrium models are full employment models. As such, no net jobs are created or lost by trade liberalization. What can be observed is a change in the allocation of the workforce across the various industries as shown in Table 6. Thus the model is not useful in discovering if trade liberalization creates more than destroys jobs in the economy. Therefore, an examination of other factors is important; the rest of the discussion looks at the literature on this topic.

5.1. Trade liberalization and employment: empirical assessments

The empirical relationship between trade liberalization and employment has been difficult to pin down. In the early 2000s, a series of case studies, done under the auspices of the International Labor Organization, on the impact of trade liberalization on manufacturing employment in selected emerging economies failed to yield a clear picture [Lee 2005]. In the three Asian emerging economies covered by the case studies (i.e. China, India, and Malaysia), the expansion of trade led to increases in employment in manufacturing, employing greater number of low- compared to high-skilled workers. The driver was the increased growth of export-oriented manufacturing industries, which happened to be relatively labor intensive compared to import-competing industries. However, jobs were not altogether lost in the latter despite the competition with imports.

However, these positive impacts of growth on employment failed to transpire in the Latin American countries in the sample, such as Brazil and Mexico. Manufacturing employment has either stagnated or declined. Low-skilled workers tended to lose jobs. The results were attributed “to unfavourable initial conditions (e.g., extremely unequal distribution of assets), problems of macroeconomic management and overdependence on external resources, but more work is required to develop adequate insights” [Lee 2005:8].

5.2. Recent assessments

More recent assessments of the relationship between trade liberalization and employment take off from the trade model that features heterogeneous firms, differentiated products, trade costs to export [Melitz 2003], and labor matching with equilibrium unemployment [Mortensen and Pissarides 1999].

Helpman and Itskhoki [2010] developed such a trade model for two countries. They showed analytically that the country with lower labor market search costs gains proportionately more than the other. Reducing trade barriers between the two countries may raise unemployment, defined as the excess of workers seeking work and the available job vacancies that are filled up.


The state of the country’s labor market institutions has the important role in determining whether trade liberalization raises or reduces unemployment. When labor market frictions are high, trade reforms raise the rate of unemployment, but they lower the unemployment rate under a more flexible labor market.

5.3. Market transaction costs

A related gap in analyzing the effects of trade on employment incorporates product market transaction costs. Allen [1991] suggested classifying transaction costs into those related to exchange of assets and those associated with defining and enforcing property rights. The former arise “because parties to exchanges must find one another, communicate and exchange information” and must need “to inspect and measure goods to be transferred, draw up contracts, consult with lawyers or other experts and transfer title” [Stavins 1995:134]. Lack of logistics infrastructure, communications and banking facilities, and other common services constrains the capability of producers to take advantage of known Vmarket opportunities.

The latter are costs related to establishing and maintaining property rights, which are needed in organizing and keeping cooperative business relationships. Property rights transaction costs have the potential of dampening the investments needed to facilitate adjustments to reforms.7 Exports require investments, which in turn depend upon the investment climate of the country.

The omission of transaction costs in market models may help explain the deviation between the simulated effects of economic policy reforms and the observed secondary data. An ex-post assessment of the effects of unilateral trade reforms in the Philippines yielded results that did not meet the findings from an analysis of the reforms using a computable general equilibrium model without transaction costs8 [Clarete 2006].

5.4. Baseline unemployment

An interesting scenario is if the trade liberalization occurs when there is starting labor unemployment in the economy. Suppose first that labor or export product market transaction costs are weak. In this scenario, trade liberalization will generate new jobs. Without transaction costs, the export industries can absorb the unemployed by exporting to the rest of the world as much products as its

7 Allen [2001] regards Coase [1937] as the one who raised the importance of the transaction costs in establishing and enforcing property rights. Interestingly, in about the same period, Hicks [1935] wrote about “the cost of transferring assets from one form to another,” referring to the “neoclassical” meaning of transaction costs.8 The simulation of the ex-ante effects of trade liberalization was conducted by Habito and Cororaton [2000] using a 50-sector computable general equilibrium of the Philippine economy.


resources permit. In this scenario, regional trade liberalization is a remedy to labor unemployment.

However, it is very likely that the realistic scenario is one that combines baseline unemployment of labor in an economy and significant export product and labor market transaction costs. It is not surprising to expect that jobs may be lost more than created because of short-run lack of capacity of the export industries to take advantage of new market opportunities. Trade liberalization in this case could worsen the baseline unemployment instead of alleviating it. But such is not the only scenario. If transaction costs are weaker compared to the willingness of workers to accept a lower wage just so that they can get employment, then trade liberalization can generate new jobs.

6. Concluding observations and policy recommendations

The economy-wide effects of the goods trade liberalization in asean on the Philippine economy are computed using the gtap model of the world economy. Goods trade liberalization is a key reform made by asean leaders under the asean Economic Community, comprising the free trade area tariff reforms, prohibition of non-tariff trade barriers, and trade facilitation. Altogether, 24 of the 40 industries representing the Philippine economy reduce production. The economic resources they lost went to the 15 industries that expanded their respective production. Notwithstanding the contraction of production in 24 industries, the country comes out a net gainer in aggregate output. With a base value of aggregate output at P11.3 trillion, the country comes out of this tariff liberalization process as gaining by 1.41 percent. Exports and imports both increase in value, but there is a positive trade balance as goods and services outflow outweigh inflows.

Considerable movement of workers across industries is observed. Reductions in skilled worker employment resulted in 31 industries, and 35 industries do the same in the case of unskilled labor. The industries with the largest layoff rates are agriculture-based, namely, milled rice, milled sugar, and vegetable oils. Manufacturing industries come next. Thirteen to fifteen out of 20 manufacturing industries let go of their workers to other industries. While all services industries lay off skilled or low-skilled workers, the layoff rates are all less than a percent. Despite the loss of jobs, the other industries that expanded production absorbed the laid-off workers.

The study looked at also the important topic of how the economy adjusts towards the expected changes as a result of goods trade liberalization in asean. Does asean integration help the Philippines create jobs? As conventionally structured, general equilibrium models, like the one used in this study, are full employment models. As such, no net jobs are created or lost by trade liberalization. One needs to adjust the model to take into account the possible unemployment in the economy, a task yet to be done.


The analysis of labor market adjustment in this paper to freer trade indicates that workers are likely to get unemployed for some time, as the export-oriented activity tends to adjust at a slower rate than imports. The adjustment period can be shortened with appropriate investment in adjustment facilitation programs. These may aim to provide information to the private business sector on new market opportunities, organizing value chains, facilitating coordinated investments along the chain, clustering of related small and medium enterprises to boost external economies, and providing information to employers and workers of job opportunities, including training.

*University of the Philippines School of Economics **Ateneo de Manila University

References

Allen, D. W. [1991] “What are transaction costs?”, Research in Law and Economics 14: 1-18.

Armington, P.S. [1969] “A theory of demand for products distinguished by place of production”, IMF Staff Papers 16(2): 179-201.

asean Secretariat [undated] “Annex 2: tariff schedules”, http://www.asean.org/news/item/annex-2-tariff-schedules. Accessed 1 June 2015.

asean Secretariat [2007] asean Cargo Clearance and customs clearance document. Jakarta: asean Secretariat.

asean Secretariat [1999] asean Economic Community Blueprint. Jakarta: asean Secretariat.

asean Secretariat [2011] asean on non-trade barriers. Jakarta: asean Secretariat.

Brockmeier, M. [2001] “A graphical exposition of the gtap model”, gtap Technical Paper No. 08. Purdue University.

Clarete, R. [2006] “Ex-post effects of trade liberalization in the Philippines”, in S. Laird and F. de Cordoba, eds., Coping with trade reforms: a developing-country perspective on the wto industrial tariff negotiations. New York: Palgrave Macmillan.

Coase, R. [1937] “The nature of the firm”, Economica 4 (November): 386-405.De Dios, L. [2007] “Non-tariff barriers to trade in the asean priority goods

sectors”, in D. Hew Wei-Yen, ed., Brick by brick: the building of an asean Economic Community. Singapore: Institute of Southeast Asian Studies.

Habito, C. and C. Cororaton [2000] “wto and the Philippine economy: an empirical and analytical assessment of post-wto trade reforms in the Philippines”, a report prepared for the usaid/Philippines agile Project.

Harrison, W.J. and K.R. Pearson [2002] gempack User Documentation: Release


8.0. Melbourne, Australia: Center of Policy Studies and impact Project.Helpman, E. and O. Itskhoki [2007]. “Labor market rigidities, trade and

unemployment”, nber Working Paper 13365. Cambridge, Mass. Hertel, T. [2001] “Global applied General Equilibrium analysis using the gtap

framework”, gtap Working Paper No. 66. Center for Global Trade Analysis, Purdue University.

Hertel, T. [1999] Global trade analysis: modeling and applications. New York: Cambridge University Press.

Hicks, J. [1935] “A suggestion for simplifying the theory of money”, Economica 2(1): 1-19.

Hill, H. [2003] “Industry”, in A. Balisacan and H. Hill, eds., The Philippine economy: development, policies and challenges. Quezon City: Ateneo de Manila University Press.

International Labor Organization (ilo) and Asian Development Bank (adb) [2014] asean Community 2015: Managing integration for better jobs and shared prosperity. Bangkok, Thailand.

Laird, S. and F. de Cordoba [2006] Coping with trade reforms: a developing-country perspective on the wto industrial tariff negotiations. New York: Palgrave Macmillan.

Lee, E. [2005] “Trade liberalization and employment”, DESA Working Paper No. 5 st/esa/2005/dwp/5. http://www.un.org/esa/desa/papers.

Levinsohn, J. [1999] “Employment responses to international liberalization in Chile”, Journal of International Economies 47: 321-344.

Melitz, M.J. [2003] “The impact of trade on intra-industry reallocations and aggregate industry productivity”, Econometrica 71(6): 1695-1725.

Mesquita, M. and S. Najberg [2000] “Trade liberalization in Brazil: creating or exporting jobs?”, Journal of Development Studies 30(3) February: 78-100.

Mortensen, D. T. and C.A. Pissarides [1998] “Technological progress, job creation and job destruction”, Review of Economic Dynamics 1(4): 733-753.

Narayanan, B.G., A. Aguiar, and R. McDougall [2012] Global trade, assistance and production: the gtap 8 database. Purdue: Center for Global Trade Analysis Project, Purdue University.

Rutherford, T. [2005] “gtap6 in gams: the dataset and static model,” prepared for workshop on applied General Equilibrium modeling for trade policy analysis in Russia and the CIS, Moscow, December 1-9.

Soloaga, I. and L. A. Winters [2001] “Regionalism in the nineties: what effect on trade?”, The North American Journal of Economics and Finance 12(1): 1-29.

Stavins, R.N. [1995] “Transaction costs and tradable permits”, Journal of Environmental Economics and Management 29(2): 133-148.

World Trade Organization [2011] World trade report 2011. Geneva: World Trade Organization.


PRE

Just how good is unemployment as a measure of welfare? A note

Emmanuel S. de Dios* and Katrina Dinglasan*

Governments are rightly concerned with employment generation to make growth inclusive. The use of the open unemployment rate to measure success, however, may be misplaced. In a developing country like the Philippines, with a large informal sector and in the absence of unemployment insurance, open unemployment is primarily a middle-class phenomenon: the unemployed are not predominantly poor, and the poor are not predominantly unemployed. Measures of productivity and shifts of labor across sectors may contain more information and be more welfare-relevant.

JEL classification: J21, I32, O15Keywords: unemployment, underemployment, labor force, welfare, poverty,

development economics

1. Introduction

The Philippine government, like many others, has placed employment generation at the center of its objective of “inclusive growth” [Philippine Development Plan 2010]. Employment generation is also regarded as the principal tool for halving poverty incidence—as committed under the earlier Millennium Development Goals. This has naturally focused the attention of policy-makers and the public on the unemployment rate—the headcount of the unemployed as a proportion of the labor force—as a measure of success or failure of the government’s performance in achieving inclusive growth.

This note, however, cautions against an uncritical use of the unemployment rate as a measure of welfare or of inclusion. The reason is that in the specific conditions of a developing country—particularly one with a large informal sector and a poorly developed social insurance system—unemployment correlates only very poorly with poverty. The proposition may be put most starkly as follows:


most of the unemployed are not poor; and most of the are not unemployed.None of this is new: what is merely curious is how hard-won lessons from the

earlier period of development economics have been so casually forgotten.1 Writing almost forty years ago, Gunnar Myrdal [1968:961-1027], among others, took great pains to question the uncritical appropriation of unemployment concepts derived from developed-country experience—or their use as a significant guide to policy in less developed economies. In industrial-country contexts, he argued, open unemployment represented a readily available labor reserve that could be readily mobilized through the management of aggregate demand. In developing, and especially agricultural countries, by contrast,

...[T]he readily available labor supply [i.e., that which is measured by open unemployment—esd] represents only a very small proportion of the real waste of labor. A massive waste of labor—whether because labor is not utilized at all, or is utilized for only parts of the year, month, week, and day, or is utilized in an almost useless way, that is at a low level of productivity—is one of the obvious facts of economic life in the region. In the present context, the important point is that little of this slack in the labor force can be taken up by turning on the tap of aggregate demand. Underutilization of labor vastly exceeds the supply that could be mobilized by expansion in monetary demand [Myrdal 1968:999].

Myrdal drew the conclusion that open unemployment rates understated the true extent of the problem in many economies where agriculture predominates, and moreover that real labor underutilization was unlikely to be addressable through short-term Keynesian aggregate measures: “...‘[F]ull employment’ is a distant goal and not one which can be reached. The limited scope of organized markets, among other things, makes aggregative measurement of the underutilization of labor far less possible” [Myrdal 1968:1001]. Open unemployment rates, in short, are poor welfare measures for many less-developed countries.

2. Most of the unemployed are not poor

The continuing relevance of critiques such as Myrdal’s can be demonstrated empirically. We proceed by using the merged files of the (Philippine Statistics

1 Part of this may be due to the decline in interest in development issues per se among (especially North American) academic departments, as Krugman later [1994] noted. Hal Hill, with other fellow scholars of the region, has been among the faithful remnants who stayed the course until the ultimate vindication of development economics.

236 de Dios and Dinglasan: Just how good is unemployment as a measure of welfare? A note

Authority, psa2) Labor Force Survey and the Family Income and Expenditure Survey of 2009 to examine the poverty status of the various sections of the labor force L, namely, the unemployed U and the employed N, with the latter consisting of the fully employed F and the underemployed D. We have L = U + N = U + (F + D). The most salient results are shown in Table 1.

TABLE 1. Poverty in the labor force, 2009

PovertyIncidence (%)

Number of poor Share in poorpopulation %

Unemployed 17.01 485,009 3.72

Employed 22.80 8,202,347 62.92

Of whom:

Fully employed 19.37 5,511,609 42.28

Underemployed 35.76 2,690,738 20.64

Labor force 22.38 8,687,356 66.64

Not in the labor force 20.35 4,348,001 33.36

Total 21.66 13,035,357 100.00

Source: Computed from PSA 2009 Labor Force Survey and Family Income and Expenditure Survey

It will be immediately evident from Table 1 that poverty incidence is actually lowest among the unemployed. Of some 2.85 million unemployed persons in 2009, only 17 percent—less than half a million—were classified as being poor. This should be compared with the much higher poverty incidence of 36 percent among the underemployed—who, it should be remembered, are among those regarded as already employed. Indeed, compared to the unemployed, poverty was even slightly higher among those who were fully employed (19 percent). Poverty among the unemployed was also significantly less than the national average in that year, namely, 22 percent.

Therefore, in terms of the simplest welfare measure—poverty incidence—the unemployed are paradoxically the best-off group in the population; somewhat worse is the situation of the fully employed, followed closely by people not in the labor force. By far, the worst-off are the underemployed.

The other half of the statement is also true, namely, the majority of poor people in the country are not among the unemployed but rather among the employed. This is also seen in Table 1, which shows that of the 13 million persons officially classified as poor in 2009, less than four percent were unemployed. Most of the poor are in fact employed—indeed 42 percent of them are even fully employed, while 21 percent are underemployed.

2 Formerly the National Statistics Office


3. Unemployment is mainly a middle-class phenomenon

Table 2 shows rates of unemployment among various income groups of the population, ranging from the poorest twenty percent (Quintile 1) to the richest (Quintile 5). Open unemployment is lowest among the poorest fifth of the population, where it is only 5.1 percent. It then rises steeply to between 7 and 9 percent among the middle classes (Quintiles 2-4) before dropping slightly among the richest. As a result, more than two-thirds of all the unemployed are from the second to fourth quintiles, while only 15 percent of the unemployed are from the poorest 20 percent of the population.

TABLE 2. Unemployment across income quintiles, 2009

Q1 Q2 Q3 Q4 Q5

Unemployment rate (%) 5.1 7.2 8.5 9.1 6.9

Share of unemployed (%) 14.9 20.1 23.2 24.0 17.7

Source: Computed from NSO data(Q1 = poorest income quintile; Q5 = richest income quintile)

The same conclusion is drawn when one looks at educational attainment among the unemployed (Table 3). Almost half of the employed have not completed a secondary education.

TABLE 3. Educational attainment among the unemployed and the employed, 2009

Unemployed Employed Share difference

No education 0.64 1.82 1.17

Incomplete primary 7.26 15.74 8.48

Complete primary 7.60 15.55 7.95

Incomplete secondary 13.81 13.74 (0.06)

Complete secondary 33.09 25.76 (7.33)

Incomplete college 19.09 13.00 (6.09)

Complete college 18.49 14.21 (4.28)

Complete postgrad 0.02 0.18 0.16

Source: Computed from PSA 2009 Labor Force Survey and Family Income and Expenditure Survey

The bottom line is that unemployment in the Philippine case is primarily a problem of the middle class. It is a phenomenon that is bound to assume increasing social significance as the country progresses. For the present, however, it merely implies that fighting poverty and battling open unemployment are two different things.

The weak correlation between poverty and unemployment will surprise some, since it flies against mental pictures formed in the context of industrial economies.


For the u.s., for example, a table similar to Table 1 can be computed. Table 4 shows the large difference in poverty incidence as between the unemployed (28 percent) and among the employed (7 percent). The same table also shows the large gap in unemployment rates as between the poor and the nonpoor in the labor force (i.e., 26 percent and 6 percent, respectively).

TABLE 4. Poverty incidence and employment status in the u.s., 2012 (in thousands)

Poor Nonpoor Total Poverty incidence (%)

All persons 46,496 264,152 310,648 15.0

Unemployed 3,367 8,802 12,169 27.7

Employed 9,587 133,006 142,593 6.7

Labor force 12,954 141,808 154,762 8.4

Memorandum (%) Unemployment rate 26.0 6.2 7.9

Source: Computed from the (U.S.) Current Population Survey 2012.Notes: “Unemployed” includes those who have just been laid off and those who are looking for work; “employed” includes those at work and those who have a job but are not at work; the labor force excludes those in the military.

The close association is further seen when one relates the period of unemployment with poverty. Poverty incidence was only 2.9 percent among full-time workers, but it was 16.6 percent among those who worked less than a full-time year [Nichols 2013]. Household evidence in the u.s. also shows poverty incidence rising with longer spells of unemployment. u.s. data for 2010 show that poverty incidence was 13 percent among people who experienced no unemployment, but it was 19 percent among those unemployed for 1-28 weeks and as high as 30 percent for those unemployed for 27 weeks or more [Nichols and Callan 2013]. Econometrically, Hoynes, Page, and Stevens [2006] find the unemployment rate to be one of the labor-market opportunity variables that affect the incidence of poverty at the aggregate level. By contrast, no such relationship has, to our knowledge, been established in the Philippines. What appears to have been established instead is a relationship between measures of unemployment and subjective measures of household satisfaction with government performance [Mapa et al. 2013]. This actually jibes with our interpretation of unemployment as a middle-class phenomenon.3 It is, after all, the middle class that performs a vital role of forming and influencing national political opinion (e.g., through

3 As an unemployment variable, Mapa et al. [2013] use self-reported “joblessness” as found in the public opinion polls of the Social Weather Stations. This differs from the official definition in some respects, notably the reference being to current idleness rather than to a reference week. Like the official statistic, however, there is “no outstanding correlation” between those who self-report as poor and the self-reported jobless (Personal communication with Mahar Mangahas).


media). One should not be surprised, therefore, if a phenomenon affecting them (unemployment) should figure in the more general opinions of government, even if it does not necessarily affect the greater majority of the poor.

The reason for the discrepancy in the welfare significance of unemployment as between poor and rich countries is as follows: Under standard statistical definitions, being unemployed requires one not to have worked even a single hour during the past week, to have actively sought work, and to be available for work. In richer societies, unemployment insurance, welfare benefits, and other transfers typically kick in when one is out of work. Such a system allows unemployed people to devote time to job search and still sustain themselves. Since unemployment and welfare benefits typically pay less than the average wage, people who are unemployed are counted close to or even below the poverty line. Finally, the fact that part-time jobs may pay less than welfare and unemployment benefits discourages unemployed people from accepting such jobs (possibly forfeiting or reducing their benefits) and tend to keep them fully unemployed. This explains the closer relationship between unemployment and poverty in those cases.

In the Philippines, as in many other poor countries, however, two things stand out: (a) there is no system of unemployment or welfare benefits; and (b) an informal sector exists which is easy to enter and exit owing to low skill demands and low productivity. The first removes the feasibility for the poor to devote themselves to full-time job search, since there is no means to support themselves in the process. At the same time, a large informal sector beckons that is easy to enter and to exit. Easy entry into low-productivity, low-wage jobs will suffice to remove one from the ranks of the unemployed, but will hardly ameliorate poverty. As the old development adage goes, “The poor cannot afford to be unemployed.” Indeed, the fact of their employment is a sign not of improvement in their welfare, but of their lack of choice.

By contrast, it is people who are better able to support themselves through a spell of job search who will be found among the openly unemployed. These will be those who can rely on personal savings, or who come from families with sufficient means, who have better access to social networks, or people with some education and who, therefore, have better job prospects—or all of these—in short the middle class. For this reason, an unemployed person is more than 80 percent likely to be non-poor.

4. The poverty impact of falling unemployment

How is a change in the unemployment rate related to a change in poverty incidence? An answer in purely accounting terms can be provided as follows. Let P be the number of poor persons in the labor force L and p = P/L the (headcount) poverty incidence in that category. Then, using the fact that L = U + N = U + (F + D), where U, N, F, and D are defined as before, and letting P

k , k = U, F, D,

be the poverty headcounts among the unemployed, the fully employed, and the


underemployed, respectively, we obtain:

p = (PU + P

F + P

D) / L

= (αUU + α

FF + α

DD) / L

= αU(U/L) + α

F(F/L) + α

D(D/L) (1)

where the αk, are rates of poverty incidence among k = U, F, D. We note that the

unemployment rate u = U/L = (1 – N/L) so that N/L = (1 – u); F/L = (N/L) – (D/L); and the underemployment rate d = D/N, so that D/L = (D/N)(N/L) = d(1 – u). Substituting these into the last identity of (1) above yields:

p = αUu + α

F(1 – u) – α

F (1 – u) d + α

D(1 – u) d

= αF + u (α

U – α

F) + (1 – u) d (α

D – α

F ) (2)

This last expression relates overall poverty incidence in the labor force with poverty incidence in its various categories. The association between poverty incidence and a change in the unemployment rate can then be approximated as

.

∂p/∂u = (αU – α

F) – d(α

D – α

F ) (3)

If using Table 1 we substitute into (3) the values αU = 0.17, α

F = 0.19, α

D =

0.34, and d = 0.19 we obtain a value of: –0.00326. This is remarkable not only for its small magnitude but more importantly its sign. It suggests not only that an increase in the unemployment rate has little effect on poverty, but that a higher unemployment rate might indeed even improve it!

It is plainly wrong, of course, to interpret this to mean that poverty incidence could actually be reduced by increasing the rate of unemployment. It merely reflects the accounting identity that (with a fixed labor force) the ranks of the unemployed can fall only by drawing away from the employed. Given the existing rates of poverty, however, the random unemployed person is even less likely to be poor than her employed counterpart; so a move from unemployment to employment can be an ambiguous matter.

More constructively, one might look for conditions under which the expression in (3) is positive—i.e., where an increase (decrease) in unemployment is likely to increase (reduce) poverty. The sufficient condition is given by

αU > (1 – d) α

F + dα

D (4)

As is readily evident, this says that poverty among the unemployed must be worse than average poverty among the employed (the weight being represented by d): a reduction in unemployment is more likely to reduce poverty if poverty among


the fully employed is far less among the unemployed and the underemployed, and the lower is the rate of underemployment.

The pathology of the Philippine case (which may also be true of other developing countries), however, is that α

F > a

U, α

D > α

F so that condition (4)

is impossible to fulfill for any d in the interval [0, 1]. This explains the perverse result.

At any rate, this simple exercise does focus attention on the key problem, which is the poverty incidence among those who are employed, particularly those who are fully employed. Somewhat paradoxically, in order for unemployment reduction to imply poverty reduction, poverty must be reduced among those who are already employed.

5. Poverty among the employed

For more detail, we can disaggregate the employed by sector as well as by their poverty status (Table 5). Most of the employed poor can be found in sectors where informal employment relations predominate and which are notorious for low-productivity jobs. The most prominent is agriculture, which alone already accounts for almost two-thirds of the employed poor. Other sectors that serve as major collecting pools for the employed poor are wholesale and retail trade (think vendors and hawkers); private household services (e.g., domestic help); informal sector manufacturing (e.g., sweatshops and small household businesses); and transport (e.g., jeepney drivers, tricycles, kuliglig, and pedicabs).

Agriculture is also the sector with the highest incidence of poverty (44 percent) among those it employs. Poverty among people engaged in the mining sector is also extremely high (42 percent), although the poor in that industry are only a small percentage of the total poor. The high incidence of poverty in mining doubtless also reflects the desperate conditions of the informal mining sector, as exemplified by the small-scale mining operations in Compostela Valley. This example also illustrates the duality of conditions existing in many important economic sectors. There will in many cases be a wide gulf in scale, skills, productivity, and pay as between informal and formal employment even in the same sector: e.g., high- v. low- productivity manufacturing; high- v. low-productivity services; high and low productivity mining; and so on. As a result, simple classification of the employed according to industries will not be a reliable guide to their welfare status. Small exceptions to this are sectors such as finance, education, and utilities, where poverty is low in both incidence and extent.


TABLE 5. Poverty among the employed: distribution and incidence by sector, 2009

Distribution (%) Incidence (%)

Agriculture and fishing 63.7 44.3

Mining and quarrying 1.0 41.5

Manufacturing 5.1 13.8

Electricity, gas, water 0.1 2.7

Construction 4.6 19.1

Wholesale and retail services 10.1 11.7

Hotels 0.9 6.8

Transport 4.5 13.7

Financial 0.0 1.0

Real estate 0.4 3.1

Public administration 2.1 9.4

Education 0.2 1.6

Health and social services 0.2 4.1

Other community services 1.8 15.5

Private households 5.2 20.1

Source: Computed from NSO data

6. The takeaway for policy

The foregoing has merely sought to demonstrate how employment status can be a poor guide to policy. For government (and its critics) to use open unemployment—especially by itself—as a measure of failure or success is to completely miss the mark and underestimate the development task at hand. An undue focus on unemployment could induce policy-makers, for example, to mistakenly engage in large-scale emergency job-creation schemes financed by public spending. Such stopgap schemes are likely to have adverse budgetary consequences without making a real dent on poverty, since all they would do is transfer people who are already employed in low-productivity jobs to similar low-productivity jobs—except underwritten now by government.

It has been suggested that perhaps the extent of unemployment taken together with underemployment might provide a better measure for policy makers to track. What has been called a “job misery index” (see, e.g., Mapa et al. [2013]) takes the unemployed and underemployed together as a proportion of the labor force.4 Such a statistic is an improvement over the simple unemployment rate, especially considering how poverty is markedly higher among the underemployed. But it

4 A convenient expression involving only rates for computing (U + D)/L is u + (1 – u)d, where u and d are the unemployment and underemployment rates, respectively.


unfortunately still falls short of the mark. First, the unemployed are markedly different as a group from the underemployed, so that adding the two is something of a statistical pastiche, since the former are predominantly from the middle class, while more of the underemployed are from the poor. Second, even the job misery index neglects the fact that far more of the poor are to be found among the fully employed. If job “misery” and dissatisfaction plague even the fully employed, then real job misery would have to include virtually the entire labor force, which threatens to render the concept meaningless.

The mismeasure can impart a wrong sense of the scale of the problem of employment and its relation to poverty. A recent World Bank development report, for example, rightly focuses on the problem of providing “good jobs—meaning jobs that raise real wages and bring people out of poverty”. But its assessment of the scale of the task is hampered by an inability to sort out the most crucial welfare aspects of the problem. It defines the “jobs challenge” as one of providing jobs to “around 10 million Filipinos who were either unemployed (three million) or underemployed (seven million) in 2012, and to around 1.15 million potential entrants to the labor force every year…In addition, better jobs need to be provided to another 21 million Filipinos who are informally employed. All in all informal workers comprise about 75 percent of total employment” [World Bank 2013:5]. (Emphasis supplied).

As already demonstrated, to regard unemployment plus underemployment as the target is certainly too narrow. But to lump all the unemployed plus all the informally unemployed as the problem is also certainly too broad, since that would comprise some 77 percent of the entire labor force.5 This certainly exaggerates the welfare problem, since poverty in the entire labor force is no more than 22 percent (Table 1). The result is that no clear focus is achieved.

More importantly, the policies required to address unemployment are vastly different from those needed to solve low-productivity employment, so that lumping the two together makes little sense. The former requires mainly improving the workings of labor markets and the matching of expectations as between qualified job seekers and employers—so physical and virtual job fairs, information given to parents and students regarding career options, measures facilitating labor mobility, and perhaps temporary unemployment benefits for people between jobs are effective policies to lower the open unemployment rate.

But these measures are obviously unlikely to reduce poverty. Solving the poverty problem ultimately means raising the productivity and incomes of people who are already employed. Again it is worth noting how Myrdal’s early lucidity on

5 Here we accept the World Bank’s estimate that 75 percent of the employed are in the informal sector (which, it is important to note, does not necessarily make them poor). If unemployment and employment rates are approximately 0.93 and 0.07, respectively, then the informally employed plus the unemployed are 0.77 (= (0.75)(0.93) + 0.07) as a proportion of the labor force, as stated in the main text.


these issues has been forgotten. He noted that productivity in the labor force could be broken into three components: participation, duration, and efficiency. Letting L, N, H, and Q represent the labor force, employment, hours worked, and output, respectively, one has Q/L = (N/L)(H/N)(Q/H). “Broadly speaking,” he wrote, “the modern approach has been preoccupied with the first of these ratios, and then only in a partial and unrealistically biased way, whereas all three are essential to an understanding of labor utilization in South Asia” [Myrdal 1968:1016]. Myrdal’s complaint was that the concern for unemployment related only to N/L, i.e., its complement (1 – N/L). Since his time, statistical authorities have generated data, albeit partially, relating to H/N in the form of visible underemployment rates. It is still the case, however, that no labor statistic in developing countries captures the efficiency of hours worked, i.e., Q/H.

Yet ultimately the quality or efficiency of employment matters. Either people must attain higher productivity in their current employment, or they must transfer to higher-productivity sectors. This means, for example, increasing productivity in agriculture through higher private and public investments in that sector; the infusion of new entrepreneurship and the linking of small farm operators into higher value-added chains; extension, training, and education for small farmers and their families; and the gradual movement away from agriculture and fisheries into manufacturing and better service-sector jobs. (Always remembering of course that there are also low-productivity jobs in manufacturing and services.)

There is no direct reason the effect of such measures will be reflected in the unemployment rate; but they are more substantial and more relevant to welfare nonetheless.

*University of the Philippines School of Economics

This paper is offered to Hal C. Hill, development economist, steadfast friend, and patronus vini to the U.P. School of Economics. Support from the Philippine Center for Economic Development is gratefully acknowledged.

References

Hoynes, H., M. Page, and A. Stevens [2006] “Poverty in America: trends and explanations”, Journal of Economic Perspectives 20(1): 47-68.

Krugman, P. [1994] “The fall and rise of development economics”, Available from: http://web.mit.edu/krugman/www/dishpan.html. Accessed September 2014.

Mapa, D., S. Dineros, K. Flores, and C. Japlit [2013] “The link between job misery index and the net satisfaction rating of the President: evidence from household surveys” (Paper presented at the National Conference on Statistics)


Myrdal, G. [1968] Asian drama: an inquiry into the poverty of nations (3 vols.). Harmondsworth: Penguin Press.

Nichols, A. [2013] “Poverty in America”, Urban Institute Fact Sheet. Available from: http://www.urban.org/publications/412898.html. Accessed September 2013.

Nichols, A. and T. Callan [2013] “Unemployment and poverty”, Urban Institute Fact Sheet. Available from: http://www.urban.org/publications/412400.html. Accessed September 2013.

World Bank [2013] Philippine development report: creating more and better jobs. Philippine Office. Asia Pacific Region.

Volume LII No. 2 - PIDS Admin December... · 2015. 12. 30. · Volume LII No. 2 ISSN 1655-1516...

Documents

Transcript of Volume LII No. 2 - PIDS Admin December... · 2015. 12. 30. · Volume LII No. 2 ISSN 1655-1516...