GLOBAL COMPETITION FOR THE JAPANESE FRUIT...
Transcript of GLOBAL COMPETITION FOR THE JAPANESE FRUIT...
GLOBAL COMPETITION FOR THE JAPANESE FRUIT JUICE MARKET
By
SHIFERAW TESFAYE FELEKE
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
2006
Copyright 2006
By
Shiferaw Tesfaye Feleke
This document is dedicated to my mom
iv
ACKNOWLEDGMENTS
Writer William Arthur Ward once said, “Feeling gratitude and not expressing it is
like wrapping a present and not giving it.” I couldn't agree more. My first, and most
sincere, acknowledgment must go to the chairman of my supervisory committee, Dr.
Richard L Kilmer. I would like to express my deepest gratitude and sincere appreciation
to him for his meticulous review of the manuscript, guidance, encouragement and
patience to successfully complete my study. I gratefully acknowledge and thank him for
everything he did throughout my program. I was very fortunate to work closely with him.
Our frequent interactions were very invaluable learning experiences. I am also very
grateful to Dr. Jonq Lee for introducing me the differential demand systems and TSP
program and helping me understand the basics and analytics of differential demand
systems that provide the basis of this study. I sincerely thank him for his patience in
reviewing, providing me with invaluable comments and suggestions from the very
beginning of proposal preparation up until the completion of this dissertation. Many
thanks must also go to the other members of my supervisory committee, Drs. Ronald
Ward, James Sterns and Lawrence Kenny, for providing me with constructive comments
and suggestions. I would like to thank them all for their support and guidance. I am also
grateful to Dr. Mark Brown for his assistance with the data analysis.
I am grateful to the Food and Resource Economics Department of the University
of Florida for affording me the opportunity of research assistantship to pursue my studies
in the department for the last six years. Special thanks must go to the department chair,
v
Dr. Thomas Spreen, graduate coordinator Dr. Jeffery Burkhardt, and graduate program
assistant Jessica Herman. I am very appreciative of the support I received from Dr.
Spreen and Jessica Herman.
I am also thankful to my officemate Katherine Finn for every help she offered me
during the preparation of this dissertation and for being a nice officemate. I would also
like to thank my friends and classmates Marco, Angel, Lurleen, Joy, Mariana and Maria.
Special thanks go to Lurleen for being an important force of motivation. Our frequent
interactions have been the source of learning. I am indebted to my fellow friends Seleshi,
Worku, Abiy, Dr. Getachew, Dr. Ayalew, Dr. Tesfaye, Saba Haile Selasie, Saba Ataro
and Measho for their support, encouragement and friendship.
My final, and most heartfelt, acknowledgment must go to my father Tesfaye, my
sister Firehiwot, my wife Genet and my daughter Biruktawit. I dedicate this dissertation
to my mother Yeshi who passed away a couple of years ago.
vi
TABLE OF CONTENTS page
ACKNOWLEDGMENTS ................................................................................................. iv
LIST OF TABLES............................................................................................................. ix
LIST OF FIGURES .......................................................................................................... xii
ABSTRACT..................................................................................................................... xiii
1 INTRODUCTION ........................................................................................................1
Background...................................................................................................................1 Objectives .....................................................................................................................4 Hypotheses....................................................................................................................5 Outline ..........................................................................................................................7
2 GLOBAL PRODUCTION, TRADE AND CONSUMPTION OF FRUIT..................9
Global Fruit Production ................................................................................................9 The Production of Oranges, Lemons and Limes, and Grapefruits and Pomelos 12 The Production of Grapes, Apples, and Pineapples ............................................16
Global Fruit Trade ......................................................................................................19 Global Fruit Consumption ..........................................................................................21
3 THEORETICAL MODELS .......................................................................................24
Demand Approaches...................................................................................................24 Production Approach...........................................................................................25 Consumer Demand Approach .............................................................................28
Utility Maximization ..................................................................................................29 The Rotterdam Model.................................................................................................32
Block Independence ............................................................................................36 Block-wise Dependence ......................................................................................39 Uniform Substitute Hypothesis ...........................................................................42
Uniform substitute given block independence .............................................42 Uniform substitute given block-wise dependence........................................45
vii
4 EMPIRICAL MODELS AND ESTIMATION PROCEDURES ...............................48
Empirical Models........................................................................................................48 The Relative Price Version of the Rotterdam Model ..........................................48 The Absolute Price Version of the Rotterdam Model .........................................52 Block Independent Non-uniform Substitute-Rotterdam Model ..........................54 Block-wise Dependent Non-uniform Substitute-Rotterdam Model....................56 Block Independent Uniform Substitute-Rotterdam Model .................................60 Block-wise Dependent Uniform Substitute-Rotterdam Model ...........................63
Data Sources ...............................................................................................................66 Analytical Methods.....................................................................................................67
5 RESULTS AND DISCUSSION.................................................................................69
Descriptive Results .....................................................................................................69 Test for First-order Autocorrelation ...........................................................................70 Hypothesis Testing for Model Selection ....................................................................72
Block Independence and Uniform Substitute Hypothesis...................................72 Block-wise Dependence and Uniform Substitute Hypothesis.............................74
The relative Price Version of the Rotterdam Model...................................................76 Parameter Estimates ............................................................................................77 Expenditure Elasticities .......................................................................................82 Own-price Elasticities .........................................................................................88 Cross-price Elasticities ........................................................................................90
6 MARKET STRUCTURES AND STRATEGY OPTIONS .......................................97
Market Structures........................................................................................................97 Block Independence (Direct) with Non-uniform Substitution ............................97 Block Independence (Direct) with Uniform Substitution ...................................98 Block-wise Dependence with Non-uniform Substitution....................................98 Block-wise dependence with Uniform Substitution..........................................100 Parameter and Elasticity Estimates in Five Market Structures .........................101
Parameter estimates....................................................................................102 Expenditure elasticities ..............................................................................104 Price elasticities..........................................................................................106
Market Strategy Options...........................................................................................109
7 SUMMARY, CONCLUSIONS AND IMPLICATIONS.........................................110
Summary and Conclusions .......................................................................................110 Implications ..............................................................................................................114
APPENDIX
A PRICE COEFFICIENTS OF FRUIT JUICES IN JAPAN.......................................118
B PRICE ELASTICITES OF FRUIT JUICES IN JAPAN..........................................124
viii
C PARAMETER ESTIMATES OF ROTTERDAM MODEL UNDER DIFFERENT SEPARABILITY ASSUMPTIONS .........................................................................136
D PRICE ELASTICITIES OF FRUIT JUICES IN JAPAN IN DIFFERENT MARKET STRUCTURES.......................................................................................142
E TWO-STAGE ROTTERDAM MODEL..................................................................166
F PARAMETER ESTIAMTES OF FRUIT JUCIES IN A TWO-STAGE ROTTERDAM MODEL ..........................................................................................180
LIST OF REFERENCES.................................................................................................187
BIOGRAPHICAL SKETCH ...........................................................................................192
ix
LIST OF TABLES
Table page 2.1 Global citrus production, area harvested and yield per hectare, 2005 .....................10
2.2 Global production of oranges, grapefruit and pommels, and lemons and limes in 2005..........................................................................................................................12
2.3 Global production of apples, grapes, and pineapples, 2005.....................................16
2.4 Fruit juice imports to Japan by country of origin.....................................................22
2.5 Per capita consumption of fruits in industrialized and developing countries ..........23
4.1 Codes for countries exporting fruit juice to Japan ...................................................50
5.1 Fruit juice quantity and price log-changes, and expenditure shares, Japan, December 1995 to May 2005 ...................................................................................70
5.2 Test for first-order autocorrelation...........................................................................71
5.3 Hypothesis testing for model selection ....................................................................74
5.4 Marginal expenditure shares of imported fruit juices in Japan ................................77
5.5 Parameter estimates of cross prices of fruit juices in Japan.....................................80
5.6 Parameter estimates of own prices of fruit juices in Japan ......................................82
5.7 Expenditure elasticity estimates of fruit juices in Japan ..........................................84
5.8 Own price elasticities of fruit juices in Japan...........................................................89
5.9 Cross-price elasticity estimates of substitutes ..........................................................94
5.10 Cross-price elasticity estimates of complements .....................................................96
6.1 Importance of country of origin in five market structures .....................................101
6.2 Relative price coefficients of fruit juices in five market structures .......................104
6.3 Expenditure elasticity estimates of fruit juices in Japan in five market structures 105
x
6.4 Uncompensated own price elasticity estimates of fruit juices in Japan .................107
6.5 Compensated own price elasticity estimates of fruit juices in Japan .....................108
6.6 Market strategies by market structures...................................................................109
A-1 Relative price coefficients of fruit juices in Japan .................................................118
A-2 Slutsky price coefficients of fruit juices in Japan ..................................................121
B-1 Uncompensated price elasticities of fruit juices in Japan.......................................124
B-2 Compensated price elasticities of fruit juices in Japan ..........................................130
C.1 Marginal value shares of fruit juices in a block independent Rotterdam model....136
C.2 Relative price coefficients of fruit juices in a block independent Rotterdam model ......................................................................................................................136
C.3 Marginal value shares of fruit juices in a block independent uniform-substitute Rotterdam model ....................................................................................................137
C.4 Marginal value shares of fruit juices in a block-wise dependent Rotterdam model ......................................................................................................................138
C.5 Constant of proportionality of fruit juice groups in a in block-wise dependent Rotterdam model ....................................................................................................138
C.6 Within-group relative price coefficients of fruit juices in a block-wise dependent Rotterdam...............................................................................................................138
C.7 Marginal value shares of fruit juices in a block-wise dependent uniform-substitute Rotterdam model....................................................................................139
C.8 Constant of proportionality of fruit juice groups in a block-wise dependent uniform-substitute-Rotterdam model .....................................................................140
C.9 Within-group relative price coefficients of block-wise dependent uniform substitute Rotterdam model....................................................................................140
F.1 Marginal value shares of fruit juices in a two-stage block independent Rotterdam model ....................................................................................................180
F.2 Relative price coefficients of fruit juices in a two-stage block independent Rotterdam model ....................................................................................................181
F.3 Marginal value shares of fruit juices in a two-stage block independent uniform-substitute-Rotterdam model e.................................................................................182
xi
F.4 Marginal value shares of fruit juices in a two-stage block-wise dependent Rotterdam model ....................................................................................................183
F.5 Relative price coefficients of fruit juices in a two-stage block-wise dependent Rotterdam model ....................................................................................................184
F.6 Marginal value shares of fruit juices in a two-stage block-wise dependent uniform-substitute-Rotterdam model .....................................................................185
F.7 Relative price coefficients of fruit juices in a two-stage block-wise dependent uniform-substitute-Rotterdam model .....................................................................186
xii
LIST OF FIGURES
Figure page 2.1 Citrus productions (MT) of major producers, 1961-2005........................................11
2.2 Orange productions (MT) of major producers, 1961-2005......................................13
2.3 Lemon and lime production (MT) of the top four producers, 1961-2005................14
2.4 Grapefruit and pomelos production (MT) in the U.S. and China, 1961-2005 .........15
2.5 Grape productions (MT) of the top three countries, 1961-2005 ..............................17
2.6 Apple productions (MT) in the U.S. and China .......................................................18
2.7 Pineapple productions (MT) of major producers, 1961-2005..................................19
3.1 A two stage profit maximization..............................................................................26
3.2 A two-stage utility maximization.............................................................................27
xiii
Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
GLOBAL COMPETITION FOR THE JAPANESE FRUIT JUICE MAREKT
By
Shiferaw Tesfaye Feleke
August 2006
Chair: Richard L. Kilmer Major Department: Food and Resource Economics
This study identifies the market structure of fruit juices imported into Japan
within the context of a consumer demand theory using three different versions of the
Rotterdam model (the block independent uniform substitute-Rotterdam model, the block-
wise dependent uniform substitute-Rotterdam model, and the relative price version of the
Rotterdam model). The models were formulated under the hypotheses of block
independence/block-wise dependence among products that belong to different product
groups and uniform substitute among products that belong to the same product group.
They were estimated for six different kinds of fruit juices (orange, grapefruit, other citrus,
apple, pineapple and grape juices imported from 18 countries) on monthly per capita data
over the period December, 1995, to May, 2005, using the non-linear least squares (LSQ)
in the Time Series Processor (TSP) program. Statistical tests select the relative price
version of the Rotterdam demand model as explaining the allocation decisions better
compared with the other versions and identify a market structure which involves both
direct and indirect competition based on the country of origin.
xiv
The results have important implications for countries exporting fruit juices to Japan
for identifying marketing strategies such as price reduction, product promotion, market
integration, as well as export supply decisions in light of the expansion and contraction of
the Japanese market for imported fruit juices because of the change in income.
1
CHAPTER 1 INTRODUCTION
Japan, with the second largest economy in the world and a population of about
127 million, imports agricultural products worth over $30 billion each year (USDA). The
U.S. is the leading agricultural supplier accounting for nearly one-third of Japan’s total
agricultural imports, though this share has declined slightly since the mid-1990s. China
and the EU-15 are the next-largest suppliers, each with over 12% of Japan’s agricultural
imports (USDA).
This study focuses on a portion of Japan’s imports which include orange,
grapefruit, other citrus, apple, pineapple and grape juices. Orange, grapefruit, apple and
grape juices account for 86% of fruit juice imports on a value basis (JETRO). The
leading exporters of orange and grapefruit juices to Japan are Brazil and the U.S.,
respectively. The U.S. is also a leading exporter of grape and apple juices while Thailand
and Israel are the leading exporters of pineapple and other citrus juices, respectively.
Background
Following the deregulation of imports of apple, grapefruit, and pineapple juices as
of April 1990 and that of orange juice as of April 1992, the import penetration ratio (the
fraction of income spent on imports or the increase in the extent of consumption of
imports) of processed fruits into Japan has increased (JETRO). Furthermore, Japan is
undergoing a profound change as a result of its aging population. Japan's statistical
agency has measured a decline in population growth that is about to become an absolute
decline, and population shrank for the first time in 2006 and will gradually fall for a
2
number of years thereafter. The impact of this demographic change on the demand for
fruit in Japan is an empirical question, since either the aging affluent consumers may
increase consumption of fruits to stay healthy or demand may decrease with the absolute
decrease in population size. In either case, the increase of import penetration in the face
of an aging population and declining population growth will lead to an increased
competition among exporters.
The purpose of this study is to assess the competitiveness of the world’s largest
exporters of fruit juice into Japan through the analysis of market structure. The analysis
of market structure in marketing is concerned with identifying closely competing brands
of the same product (Clements and Selvanathan, 1988). Consumption theory is amenable
to the analysis of market structure in international markets through demand analysis. The
approach involves the analysis of the change in marginal utilities of a certain product due
to a change in consumption of a closely related product.
The decrease in marginal utility of one product with an increased consumption of
another product implies that the products are substitutes and are thus in a competitive
market structure. Otherwise, they are not substitutes (i.e., complements or independent)
and are thus in a non-competitive market structure. Substitute products can be uniform1
(close) or non-uniform. If two products are uniform substitutes, price-oriented marketing
strategies and/or generic product promotion are recommended because consumers are not
influenced by the country of origin of such products. If two products are non-uniform
substitutes, consumers are influenced by the country of origin and thus exporters can
exercise a monopolistic power over their respective products. In this case, a non-price 1 The change in the marginal utility of a dollar spent on product i is the same as that of another dollar spent on product j .
3
marketing strategy (e.g., product promotion) and/or price reduction is recommended to
increase market share.
Be it uniform or non-uniform, the decision to use a particular marketing strategy
depends on the price elasticity of demand for the product in question. Under a situation
in which a product is a uniform substitute but price inelastic, the decision to reduce price
is not advised because total revenue is reduced when price is decreased. However, the
response of demand to changes in price may be higher under the uniform substitute
relationship than under the non-uniform substitute relationship. This implies that both
the nature of substitution (uniform/non-uniform) and the magnitude of substitution
(elasticities) are important in international trade since they have different implications to
exporters for marketing strategies such as market promotion, product differentiation as
well as a product supply plan (expansion or contraction of supply).
Most empirical studies have pursued the estimation of conditional demand
functions in isolation without testing for the nature of substitution within a product group,
and the nature and magnitude of substitution between product groups. However,
conditional demand parameters thus estimated are rarely of interest for policy analysts
because the appropriateness of marketing strategy depends on the relationship between
products within the same product group and across different product groups.
If, for example, the relationship between products within the same product group
is uniform, the appropriate marketing strategy is price reduction because consumers view
those products as homogenous. If, however, the products in the group are non-uniform,
product promotion is recommended because consumers can pay a different price since
they view them as differentiated products. Furthermore, since the optimal allocation of
4
expenditure to products in any one partition may depend on prices of products outside
that group in a uniform or non-uniform fashion, the failure to consider the nature and
magnitude of substitution between products in different products groups may misguide
marketing strategists. For example, the effect of a change in price of Chinese apple juice
on the demand for Brazilian may be the same as that on the demand for Florida orange
juice. The marketing strategy that is appropriate for this situation is different from the
situation in which the effect of a change in the price of Chinese apple juice on the
demand for Brazilian orange juice is different from that on the demand for Florida orange
juice. To be useful for policy applications in terms of designing an effective marketing
strategy, the demand for fruit juices in this study is estimated under different scenarios of
market structures consistent with consumer’s preference structure.
Objectives
The objectives of this study are the following.
(1) To characterize the trend and pattern of the world fruit production, trade and consumption.
(2) To identify the market structure of fruit juices imported into Japan by
estimating a differential consumer demand system. (3) To assess the competitiveness of the world’s largest exporters of fruit juice
into Japan. (4) To simulate the impact of changes in population growth on the growth rate of
demand for fruit juices by country of origin.
In order to identify the market structure of fruit juices in Japan, two hypotheses
are tested. These are block independence/uniform substitute and block-wise
dependence/uniform substitute hypotheses.
5
Hypotheses
Block Independence/Uniform Substitute Hypothesis
The hypothesis of block independence/uniform substitute states that there is no
change in marginal utility of a dollar spent on a product in one product group caused by
an extra dollar spent on another product in another product group. But, the change in the
marginal utility of a dollar spent on a product in one product group caused by an extra
dollar spent on another product in the same product group is the same for all pairs of
products in that group. This hypothesis represents the market structure of block
independent (direct competition) with uniform substitution such that a change in the price
of a product in one group (e.g. orange juice group) does not affect the demand for another
product in another group (e.g. apple juice group). But, the change in the price of a
product in one group (e.g. orange juice group) uniformly affects the demand for another
product in the same group.
The failure to reject the null hypothesis implies that exporters of one fruit juice
group don’t have to worry about the change in price of products that belong to other juice
groups because competition occurs only between products of the same product group or
the same products differentiated by country of origin. Furthermore, exporters of products
that belong to the same product group can only compete by reducing price (i.e. use a
price-oriented marketing strategy and/or generic product promotion) because under such
circumstances consumers are not influenced by the country of origin of the product, since
they perceive products from different countries as homogenous. Brand promotion is not
recommended because brand promotion for a uniform substitute product is technically a
generic promotion. For example, if Florida orange juice is a uniform substitute to
6
Brazilian orange juice, promoting Florida orange juice may rather help raise the sales of
Brazilian orange juice.
In summary, if two products are uniform, only a slight decrease in price makes a
big difference in sales, implying that the market of uniform substitute products is very
competitive. This may lead firms to merge so that they will be able to exercise a
monopolistic power.
Block-wise Dependence/Uniform Substitute Hypothesis
The hypothesis of block-wise dependence/uniform substitute hypothesis states
that the change in the marginal utility of a dollar spent on a product in one product group
caused by an extra dollar spent on another product which belongs to a different product
group is the same for all pairs of products that belong to the two product groups. Also,
the change in the marginal utility of a dollar spent on a product in one product group
caused by an extra dollar spent on another product in the same product group is the same
for all pairs of products in that group. This hypothesis represents the market structure of
block-wise dependent with uniform substitution such that a change in the price of a
product in one group (e.g. orange juice group) affects the demand for another product in
another group (e.g. apple juice group) in a similar fashion. Furthermore, the change in
the price of a product in one group (e.g. orange juice group) uniformly affects the
demand for another product in the same group.
The failure to reject the null hypothesis implies that exporters of one fruit juice
group need to watch the change in price of products in other juice groups because
competition occurs between products of different product groups. Since the competition
between products in different groups occurs in a similar fashion, a slight change in price
7
of one product in one group will significantly affect the demand for products in other
groups. Furthermore, exporters of products that belong to the same product group can
only compete by reducing price because under such circumstances consumers are not
influenced by the country of origin of the product, since they perceive products from
different countries and product groups as homogenous.
In summary, if two products are uniform within and across product groups, only a
slight decrease in price makes a big difference in sales, implying that the exporters of
products that belong to different product groups is very competitive. Hence, exporters of
products that belong to different product groups should pay close attention to the price
behavior of either product because only a slight change in price of one juice group
significantly affects the sales of another juice group.
Based on results of the test of the above hypothesis, the study will identify the
market structure of Japan’s fruit juice market. This will allow analyzing the
competitiveness of countries exporting fruit juices to Japan, and drawing implications in
terms of marketing strategies. Results will be useful for providing a structure for
marketing research on closely related products and identifying marketing strategies
involving price reduction, product differentiation and market promotion.
Outline
The dissertation is organized as follows. Chapter 2 presents the global fruit
production, trade and consumption. In this chapter, the trend, pattern and quantity of
production, trade and consumption of major players are investigated.
Chapter 3 presents the theoretical section in which the common approaches in
import demand analysis and the different demand models are reviewed. The chapter also
8
derives the different versions of the Rotterdam model used for empirical estimation and
tests the hypothesis presented in chapter 1.
Chapter 4 presents the empirical model and the estimation procedure. This
chapter includes (1) the systems of equations that are empirically applied to statistical
data (2) the procedures that need to be followed to estimate the models (3) the source of
data and analytical methods.
Chapter 5 presents the results and discussion. This chapter discusses (1) the
model that best describes the import data of fruit juices (2) the expenditure and price
elasticities estimated from the selected model (3) results of simulation about the effect of
the decline in population growth on the growth of demand for fruit juices.
Chapter 6 presents different market structure scenarios and compares the results
of these different market structures with the results of chapter 5.
Finally, chapter 7 summaries the results and draws conclusions. Based on the
conclusions, implications are drawn.
9
CHAPTER 2 GLOBAL PRODUCTION, TRADE AND CONSUMPTION OF FRUIT
This chapter presents a description of global fruit production, trade and
consumption. Both citrus and non-citrus fruits are included. The citrus fruits include
orange, grapefruit, and lemons and limes while the non-citrus fruits include apples,
grapes and pineapples. Data for this report come mainly from the website maintained by
the Food and Agricultural Organization (FAO).
Global Fruit Production
Citrus (Citrus L.) is one of the world’s most important fruit crops commercially
grown primarily between the latitudes 40°N to 40°S (University of Pretoria). According
to the University of Pretoria, Yunnan province in south-central China may be the center
of origin due to the diversity of species found, and the network of rivers in this area
which could have provided “on route dispersal” to the south. From there, they slowly
spread to northern Africa mainly through migration and trade. Citrus spread throughout
Europe during the Middle-Ages and were then brought to the Americas by Spanish
explorers. Worldwide trade of citrus fruits didn't appear until the 1800s and trade in
orange juice developed as late as 1940. Citrus production in Florida dates back to the
colonization of the state by the Spaniards in the 15th century (Spreen et al. 2006). Today,
the major types of edible citrus include citron, sour orange, sweet orange, lime, lemon,
shaddock (pomelos), grapefruit, mandarin, and kumquat.
The world’s largest producers of citrus fruits are Brazil, China, U.S. and Mexico
whose combined production accounted for half of the world’s total in 2005. During the
10
same year, Brazil’s production accounted for the highest proportion (19%) followed by
that of China (15%), U.S. (10%) and Mexico (6%) of the world’s total (Table 2.1). In
terms of area, China, Brazil, Nigeria and Mexico rank first, second, third, and fourth,
respectively, accounting for about 23%, 12% and 10% and 7% of the global citrus area
harvested in 2005, respectively. During the same year, the world’s highest yield per ha
was obtained in Turkey, Syria, S. Korea and U.S., each producing about 26 Mt per
hectare. The productivity of citrus in China as measured by yield per ha is one of the
lowest in the world (FAO, 2005).
Table 2.1 Global citrus production, area harvested and yield per hectare, 2005 Country Production(MT) % Yield (MT/Ha) Area (ha) % Brazil 20,142,100 19 Turkey 26.7 China 1,714,300 23 China 16,019,500 15 Syria 26.3 Brazil 930,379 12 U.S. 10,317,200 10 S. Korea 26.2 Nigeria 730,000 10 Mexico 6,475,411 6 U.S. 26.0 Mexico 523,505 7 Spain 4,867,300 5 Guatemala 24.7 U.S. 397,080 5 India 4,750,000 5 Palestine 24.5 India 264,500 3 Italy 3,836,793 4 Israel 23.7 Spain 240,759 3 Iran 3,825,000 4 Cyprus 23.2 Iran 232,500 3 Nigeria 3,250,000 3 Australia 22.8 Pakistan 185,400 2 Egypt 2,797,600 3 Italy 22.5 Italy 170,338 2 Total 78,801,620 74 Total 5,388,761 70 World 105,431,984 100 World 13.9 World 7,605,363 100
(Source: FAO, 2005)
During the last four decades, global citrus production showed a period of
sustained growth, primarily due to expansion of cultivation (Figure 2.1). Over the same
period, the world citrus production increased more than four fold from 24,999,430 Mt to
105,431,984 Mt, growing at an average annual rate of 1.5 % (Figure 2.1). The rate of
growth could have been higher, were it not for the occurrence of freezes in Florida in the
1980s. Both bearing tree numbers and production declined by 40% between 1975 and
1986 as freezes destroyed a large portion of the industry in Lake, Orange, and Pasco
11
counties of the state of Florida (Spreen, et al. 2006). However, the increase in prices
caused by the slowed production in Florida stimulated the development of new plantings
(Spreen et al.). Given the lag between price signals and output changes, an increase in
production occurred in the 1990s and 2000s (Figure 2.1).
Until the early 1980s during which freezes devastated the Florida citrus
production, the U.S. was the world’s largest producer of citrus. During the decade of the
1980s, Brazil became the largest citrus producer in the world and the first, and almost
exclusive, orange juice exporting country (UNCTAD). Brazil’s citrus production grew at
an average rate of 4.5% over the last four decades while that of the U.S. grew at 0.6%.
0
20,000,000
40,000,000
60,000,000
80,000,000
100,000,000
120,000,000
1961 1965 1969 1973 1977 1981 1985 1989 1993 1997 2001 2005
World
Brazil
China
Mexico
U.S.
Figure 2.1 Citrus productions (MT) of major producers, 1961-2005
Over the last few years, the Chinese citrus production experienced a fast growth
(over 3%) over the last few decades (particularly in the 1990s) mainly due to the
expansion of cultivation, thus emerging as the second largest producer of citrus fruits in
the early 2000s (Figure 2.1).
12
The Production of Oranges, Lemons and Limes, and Grapefruits and Pomelos
Oranges. The major citrus fruits are oranges, lemons and limes, and grapefruit
and pommels, whose combined production accounted for 55% of the world’s total citrus
in 2005 (FAO, 2005). Among citrus fruits, orange is the most important fruit, accounting
for about 43 percent of the world’s citrus production in 2005. The world’s largest
producers of oranges are Brazil and U.S, whose combined production in 2005 was 44%
of the world’s total orange production with Brazil alone accounting for 30% of the world
production. The U.S. produced 14 percent of the world production in 2005 (Table 2.2).
The top ten countries produced 76 percent of the world production in 2005.
Table 2.2 Global production of oranges, grapefruit and pommels, and lemons and limes in 2005
Oranges Grapefruit and Pommels Lemons & Limes Country Production % Country production % country production %
Metric tons Brazil 17,804,600 30 U.S. 914,440 25 Mexico 1,824,890 15 U.S. 8,266,270 14 China 443,000 12 India 1,420,000 11 Mexico 3,969,810 7 Mexico 257,711 7 Argentina 1,300,000 10 India 3,100,000 5 Israel 250,000 7 Iran 1,100,000 9 Italy 2,533,535 4 Cuba 226,000 6 Brazil 1,000,000 8 China 2,412,000 4 S. Africa 212,348 6 U.S. 745,500 6 Spain 2,149,900 4 Argentina 170,000 5 Spain 734,300 6 Iran 1,900,,000 3 Turkey 150,000 4 China 634,500 5 Egypt 1,789,000 3 India 142,000 4 Italy 609,435 5 Indonesia 1,311,703 2 Tunisia 72,000 2 Turkey 600,000 5 Total 45,236,818 76 Total 2,837,499 77 Total 9,968,625 79 World 59,858,474 100 World 3,667,862 100 World 12,554,879 100
(Source: FAO, 2005)
From 1961 to 2005, global orange production increased almost four fold from
15,946,492 Mt to 59,858,474 Mt, growing at an average annual rate of 1.4 % (FAO.
2005). Most of the growth was accounted for by developing countries, primarily in South
America but also in Asia and to a lesser extent in Africa. In South America, the volume
of production expanded considerably in Brazil and Mexico (Figure 2.2). In Asia,
production expanded significantly in China, India and Pakistan and Iran. Orange
13
production in China, Brazil and Mexico increased at an average annual rate of 4.3%,
2.7% and 1.4%, respectively over the same period (FAO, 2005). Spreen and Brown
(1995) noted that freezes in Florida in the 1980s provided a major impetus to the
expansion of orange production in Brazil. The average orange production of Brazil and
Mexico in the 1990s was 50 percent and 60 percent larger than the average production in
the 1980s, respectively (FAO, 2005).
Figure 2.2 Orange productions (MT) of major producers, 1961-2005
Lemons and limes. Lemons and limes are the second most important citrus crops
accounting for about 9 percent of the global citrus production in 2005. Like the case with
oranges, there has been a significant increase in production of lemons and limes through
expansion of cultivation. Over the last four decades, the global lemon and lime
production increased more than five fold from 2,625,865 MT in 1961 to 12,554,879MTt
in 2005, growing at the average rate of 1.6% per annum (FAO, 2005). Most of the
growth was accounted for by Mexico, India and Argentina (Figure 2.3).
0
5,000,000
10,000,000
15,000,000
20,000,000
25,000,000
1961 1965 1969 1973 1977 1981 1985 1989 1993 1997 2001 2005
Brazil China MexicoU.S.
14
Figure 2.3 Lemon and lime production (MT) of the top four producers, 1961-2005
The world’s largest producers of lemons and limes are Mexico, India and
Argentina whose production in 2005 was 15%, 11% and 10% of the world production,
respectively. Other major producers of lemons and limes include Spain, China, Italy and
Turkey, each accounting for about 5 percent of the world’s total in 2005 (Table 2.2). The
top ten countries produced about 80 percent of the world’s total in 2005.
Until the mid-1980s, the U.S. was the world’s largest producer of lemons and
limes (Figure 2.3). Between the mid-1980s and mid-1990s, the U.S. production slowed
while that of Mexico continued to rise particularly in the mid-1990s during which it
emerged to be the world’s largest producer of lemons and limes. Over the last four
decades, Mexico’s production grew at an average annual rate of 2.3% while that of U.S.
grew at 0.4%. In 2005, U.S. produced 6% of the world’s total, which is way below the
production of Mexico, India, Argentina, Iran and Brazil (Table 2.2). Over the same
period, India and Argentina also increased their production and emerged as the second
0
200,000
400,000
600,000
800,000
1,000,000
1,200,000
1,400,000
1,600,000
1,800,000
2,000,000
1961 1966 1971 1976 1981 1986 1991 1996 2001
Argentina India MexicoU.S.
15
and third largest producers of lemons and limes, respectively (Table 2.2). India and
Argentina increased their production at an average annual rate of 1.4 and 2.8%,
respectively.
Grapefruit and pommels. Grapefruit and pommels are the third most important
citrus crops, accounting for about 3.5% of the world citrus production. Over the last four
decades, the global grapefruit and pommels production increased by 73% from 2,120,896
MT in 1961 to 3,667,862 MT in 2005, growing at average rate of 0.8% per annum
(Figure 2.4).
Figure 2.4 Grapefruit and pomelos production (MT) in the U.S. and China, 1961-2005
The growth rate of grapefruit and pomelos production over the last four decades
was modest compared to the growth rate of other citrus fruits. This is due to the slow
growth of grapefruit production in the U.S. On average, grapefruit and pomelos
production in the U.S. grew at 0.2% per annum. Over the same period, China’s
production grew at a 3.8%.
0
500,000
1,000,000
1,500,000
2,000,000
2,500,000
3,000,000
1961 1966 1971 1976 1981 1986 1991 1996 2001
ChinaU.S.
16
U.S. is the world’s largest producer with 25% of the world’s total (Table 2.2).
China is the second largest producer with 12 percent of the world’s total. Mexico and
Israel are also important producers, each producing about 7% of the world’s total. The
top ten countries produced 77 percent of the world production in 2005 (Table 2.2).
The Production of Grapes, Apples, and Pineapples
Grapes. Among non-citrus fruits, grapes are the most important non-citrus fruit in
terms of production. The major producers of grapes are Italy, France and the U.S. whose
production in 2005 accounted for 14%, 10% and 10%, respectively (Table 2.3). China
and Spain are also important producers of grapes, each accounting for about 9 percent.
The top ten countries produced about 71 percent of the world’s total in 2005.
Table 2.3 Global production of apples, grapes, and pineapples, 2005 Apples Grapes Pineapples
Country production % Country production % Country production % Metric tons
China 25,006,500 39 Italy 9,256,814 14 Thailand 2,050,000 13 U.S. 4,254,290 7 France 6,787,000 10 Philippine 1,800,000 11 Turkey 2,550,000 4 U.S. 6,414,610 10 China 1,460,000 9 Iran 2,400,000 4 Spain 5,879,800 9 Brazil 1,418,420 9 Italy 2,194,875 3 China 5,698,000 9 India 1,300,000 8 France 2,123,000 3 Turkey 3,650,000 5 Nigeria 889,000 6 Poland 2,050,000 3 Iran 2,800,000 4 C. Rica 725,224 5 Russia 2,050,000 3 Argentina 2,365,000 4 Mexico 720,900 5 Germany 1,600,000 3 Chile 2,250,000 3 Indonesia 673,065 4 India 1,470,000 2 Australia 1,834,000 3 Kenya 600,000 4 Total 45,698,665 72 Total 46,935,224 71 Total 11,636,609 73 World 63,488,907 100 World 66,533,393 100 World 15,886,647 10
0 (Source: FAO, 2005)
The production of grapes in the U.S. has been growing steadily, while that in Italy
and France appears to be declining since the mid-1990s (Figure 2.5). Unlike the case
with citrus fruits, the increase in global grape production is modest. It increased by a
little more than 50% over the last four decades, growing at an average rate of 0.2% per
annum (FAO, 2005). This is due to the decline of production in the two major producing
17
countries (France and Italy) whose production declined at an average rate of 0.4% and
0.2%, respectively.
Figure 2.5 Grape productions (MT) of the top three countries, 1961-2005
Apples. Apples are the second most important non-citrus fruits. Over the last
four decades, the world apple production increased nearly four fold from 17,053,651 MT
in 1961 to 63,488,907 MT in 2005, growing at an average rate of 1.2% per annum. The
world’s largest producers of apples are China and the U.S. China produced 39% of the
world’s total in 2005. U.S. production accounts for 7% of the world’s total (Table 2.3).
Until the early 1990, the U.S. was the largest producer of apples (Figure 2.6).
Since then, China has become the world’s largest producer of apples. On average,
China’s apple production grew at the rate of 4.7% per annum while that of the U.S grew
at 0.7% per annum over the last four decades. The growth of apple production in China is
explained by an increase in area expansion.
0
2,000,000
4,000,000
6,000,000
8,000,000
10,000,000
12,000,000
14,000,000
1961
1964
1967
1970
1973
1976
1979
1982
1985
1988
1991
1994
1997
2000
2003
France ItalyU.S.
18
Figure 2.6 Apple productions (MT) in the U.S. and China
Pineapples. Pineapples are the third most important non-citrus fruit. Over the last
four decades, the global pineapple production increased almost four fold from 3,831,437
MT to 15,886,647 MT at an average rate of 1.4% per annum. Until the early 1980s, U.S.
was the world’s largest producer of pineapples (Figure 2.7). Since then, its production
has declined so that it is not in the list of the top 10 producing countries (Table 2.3).
Over the last four decades, the U.S. production declined at an annual rate of 1.3
percent per annum (FAO, 2005). Currently, the world’s largest producers of pineapple
are Thailand, and the Philippines, accounting for 13% and 11% of the world’s total,
respectively (Table 2.3). They increased their production over the last four decades at
2.8% and 2.5% per annum, respectively. China and Brazil have also emerged as the third
and fourth largest producers, each producing about 9% of the world’s total. The top ten
countries produced 73 percent of the world’s total in 2005.
0
5,000,000
10,000,000
15,000,000
20,000,000
25,000,000
30,000,000
1961 1966 1971 1976 1981 1986 1991 1996 2001
ChinaU.S.
19
0
500,000
1,000,000
1,500,000
2,000,000
2,500,000
3,000,000
3,500,000
4,000,000
1961 1965 1969 1973 1977 1981 1985 1989 1993 1997 2001 2005
Brazil China Philippines ThailandU.S.
Figure 2.7 Pineapple productions (MT) of major producers, 1961-2005
Global Fruit Trade
International trade in fruits and vegetables has expanded more rapidly than trade
in other agricultural commodities, especially since the 1980s (Huang, 2004). This is
attributed to rising incomes, falling transportation costs, improved technology, and
evolving international agreements. Citrus fruits rank first in international fruit trade in
terms of value (UNCTAD). As a result of trade liberalization and technological advances
in fruit transport and storage, the citrus fruit industry is becoming more global in scope.
The major players in the global trade of fruits and vegetables are the E.U, the North
American Free Trade Agreement (NAFTA) countries, China and Japan.
Exports of fresh citrus fruits represent roughly 10% of total citrus fruit production
(UNCTAD). The international trade on fruits and vegetables is dominated by processed
forms. According to UNCTAD, international trade in citrus juice only started to increase
in the 1940s, after World War II, when citrus processing technologies were invented and
developed. The advent of frozen concentrated orange juice (FCOJ) after World War II
20
provided a new impetus for the citrus industry (Spreen et al. 2006). Citrus fruit
processing accounts for approximately one third of total citrus fruit production. More
than 80% of it is orange processing, mostly for orange juice production. The major
feature of the world market for orange juice is the geographical concentration of
production. There are only two main players: the State of Florida in the U.S. and the
State of Sao Paulo in Brazil. Production of orange juice between these two players
account for over 80% of world orange juice production (Spreen et al. 2006).
The major difference between them is that Brazil exports 99 percent of its production
while 90 percent of Florida’s production is consumed domestically and only 10 percent is
exported (UNCTAD). The citrus industry in Florida currently faces two major
challenges (citrus canker and citrus greening) and increasing urbanization in the state,
which has resulted in increasing land values (Spreen et al. 2006). Nonetheless, the
Florida citrus industry will continue to be an important supplier of citrus products to both
the U.S. and world market.
International trade in orange juice takes place in the form of frozen concentrated
orange juice (FCOJ), in order to reduce the volume used, so that storage and
transportation costs are lower. Spreen et al. (2006) notes that FCOJ provided a means to
(1) store orange juice from the harvest season into other time periods, (2) provided a way
to produce a product with a consistent taste, and (3) offered new modes of transport and
new retail package alternatives to the consumer.
The E.U. is the largest importer of orange juice, accounting for over 80% of the
world orange juice imports (UNCTAD). The other major importers of orange juice are
Canada and Japan. Most of imports by the E.U. and Japan come from Brazil. Brazil’s
21
exports of orange juice to Japan account for over 70% of Japan’s total import of orange
juice (Table 2.4). In North America, the U.S. and Canada consume orange juice mainly
from Florida, while a small quantity of imports comes from Brazil. The U.S. is the
leading exporter of apple juice, grapefruit juice and grape juice to Japan. Thailand and
Israel are the leading exporters of pineapple juice and other citrus, respectively. The U.S.
share of grapefruit import is significant. However, the slow growth rate of grapefruit
production in U.S. implies that the U.S. is unlikely to continue as a dominant supplier of
grapefruit juice. The same is true with apple juice since the apple production growth rate
in U.S. is slower relative to other countries such as China. Currently, the U.S. is a
dominant supplier of apple juice to the Japanese market, followed by China and Austria.
With regard to grape juice, the U.S. is still the dominant supplier and is expected to
dominant the market since its production has been growing while that of France and Italy,
which are the world’s largest producers, has been declining.
Global Fruit Consumption
Higher income, urbanization, demographic shifts, improved transportation, and
consumer perceptions regarding quality and safety are changing global food consumption
patterns (Huang, 2004). Diet diversification and increasing demand for better quality
products have increased imports of high-value and processed food products in developed
countries. Fruits are mainly consumed in industrialized countries, not only because
consumers in these countries have high income levels but also because they have
increasing concerns about healthy eating. However, the growth of per capita
consumption of fruits in these countries seems to be stagnating. Over the period 1980 to
2003, the per capita consumption of citrus fruits (oranges, grapefruit and lemons and
limes) in these countries grew at an average rate of one percent per annum.
22
Table 2.4 Fruit juice imports to Japan by country of origin product Exporter %
Brazil 72.4 U.S. 23.7
Orange juice
Australia 1.4 U.S. 22.4 China 18.9
Apple juice
Austria 18.6 U.S. 87.1 Israel 9.6
Grapefruit juice
Australia 2.4 U.S. 46.9 Brazil 14.1
Grape juice
Argentina 11.7 Thailand 42.4 USA 28.6
Pineapple juice
The Philippines 27.6 Israel 40.5 Italy 21.8
Other citrus juice
Argentina 13.9 (JETRO)
Among 26 industrialized countries, the U.S. and Canada are the largest consumers
of orange and mandarins followed by the EU. In fact, some E.U. countries such as
Ireland, the Netherlands and Greece consume more oranges than do the U.S. and Canada
on a per capita basis. The average per capita consumption of oranges and Mandarins in
industrialized countries over the period 1990 to 2003 is 29 kilograms while that of
grapefruit and lemons and limes is 3.0 and 3.6 kilograms, respectively (Table 2.5).
Japan’s consumption of both citrus (except grapefruit) and non-citrus fruits is small
compared to other industrialized countries. The average annual per capita consumption
of oranges and apples in Japan over the period 1980 to 2003 is about 14 and 12 kilograms,
respectively, while those of grapes and grapefruit are 2.8 and 2.5 kilograms, respectively
(Table 2.5).
23
Table 2.5 Per capita consumption of fruits in industrialized and developing countries
Fruits Developing countries
Industrialized Countries E.U. Canada Japan U.S.
Orange and mandarins 8.00 29.23 27.52 46.28 13.80 39.87 Grapefruit 0.32 2.91 2.17 4.05 2.49 4.12 Lemons and limes 1.25 3.59 3.78 2.60 0.84 5.26 Apples 4.67 20.3 24.82 18.82 11.58 21.02 Grapes 2.20 7.60 8.67 10.19 2.79 8.18 Pineapples 2.01 3.61 1.97 2.61 1.43 7.01
(Source: FAO, 2005)
Japan’s domestic supply of pineapples is heavily dependent on imports. In 2003,
95% of the domestic supply of pineapples came from imports (FAO, 2005). Japan is also
heavily dependent on imports for its supply of lemons and limes. In terms of apples and
grapes, the significance of imports has been increasing since the last decade during which
the deregulation was in effect.
24
CHAPTER 3 THEORETICAL MODELS
Demand Approaches
Approaches common in the literature of import demand analysis involve use of
consumer demand theory and production theory. The consumer demand approach treats
imports as final products that directly enter a consumer’s utility function (Schmitz, A.
and Seale, J. 2002) while the production theory treats imports as inputs (Washington and
Kilmer, 2002). The first approach enables the derivation of the traditional consumer
demand and labor supply functions from utility maximization, while the second approach
enables the derivation of derived/input demand and output supply functions from profit
maximization or cost minimization.
The fact that output supply functions are derived in the production approach while
labor supply functions are derived in the consumer demand approach marks one major
difference between the two approaches. Another major difference between the two
approaches is that the parameter estimates of unconditional consumer demand and
unconditional input demand are different. However, similar parameter estimates can be
obtained for the conditional consumer demand and derived demand for inputs.
Furthermore, under the assumption of the constant percentage of retail price type of
marketing margin, the demand for any given quantity of product is equally elastic (or
inelastic) with respect to price at all market levels (Goodwin, 1994). This implies that
conflicts of interest between the producer level and subsequent market levels are reduced.
The constant percentage of retail price marketing margin is fairly typical for products for
25
which the marketing process involves fixed investments and substantial economies of
scale (Goodwin, 1994, pp. 292).
Production Approach
In the production approach, two allocation decisions, one involving outputs and
another involving inputs, are made. These two decisions can be made successively or
simultaneously through a two-step profit maximization or one-step or direct profit
maximization procedure yielding a system of output supply and input demand functions
(Washington, 2000). They are made successively in such a way that given output and
input prices, first the output manager decides on the quantity of output, and knowing the
quantity of output planned to be produced, the input manger decides on the quantity of
inputs required to produce the planned output. The simultaneous decisions are made by
one manager such that the input and output decisions are not independent of each other.
In this case, since the input demand and output supply functions are not independent of
each other and that their error terms are correlated (Laitinen, 1980), the input demand
function can not be estimated independently of the output supply function and vice versa.
Once the output supply and conditional input demand are estimated, the unconditional
demand parameters can be derived from the parameter estimates of the two functions
(Washington and Kilmer, 2002).
The input allocation decisions that involve the use of conditional input demand
functions can be implemented in stages/hierarchies (Theil, 1980b). That is, total
expenditure is first allocated over broader groups of inputs and then group expenditures
are allocated over individual inputs within each group. The two-stage input allocation
decision of the production approach is comparable to the two-stage utility maximization
of consumer demand approach (Figure 3.2). The consumer demand approach can yield a
26
system of group consumer demand and conditional demand functions from which the
parameter estimates of the unconditional demand function can be derived. As noted
earlier, the unconditional demand parameters thus estimated are not the same as those
derived from the system of output supply and conditional demand functions generated in
the production approach discussed earlier.
However, the parameter estimates of the input demand function (P1), group
demand function (P2) and conditional input demand (P3) in Figure 3.1 are the same as
that of the corresponding functions in Figure 3.2. That is, the parameter estimates of the
input demand function (P1) are the same as that of the unconditional consumer demand
(C1); the parameter estimates of the group input demand functions (P2) are the same as
that of the group consumer demand functions (C2); and that the parameter estimates of
the conditional input demand functions (P3) conditional consumer demand functions (C3)
in Figure 3.2.
Figure 3.1 A two stage profit maximization
2-Stage profit maximization
Output supply function Input demand function (P1)
Group input demand function (P2)
Conditional input demand function (P3)
27
Figure 3.2 A two-stage utility maximization
Although the two approaches provide the same empirical estimates with regard to
the conditional demand, and that the demand for any given quantity of product is equally
elastic (or inelastic) with respect to price at all market levels under the assumption of the
constant percentage of retail price type of marketing margin (Goodwin, 1994), the
production approach does not seem to lend itself to a theoretically consistent
investigation of demand relationships among narrowly defined import products because
of their independence. It may be realistic for broadly defined groups of imported
products. For example, Theil (1980b) applied the production approach to broad imported
products such as food, crude materials, semi-manufactures, finished- manufactures under
the assumption of input independence. However, when it comes to narrowly-defined
products such as fruit juices, it does not seem conceptually defensible and practical to
apply the production approach simply because the importing firm’s production function
of an imported fruit juice is independent of other imported juices.
Let the production function of a narrowly-defined import product such as orange
juice be given by
2-Stage utility maximization
Labor supply function Unconditional demand function
(C1)
Group consumer demand function (C2)
Conditional consumer demand function (C3)
28
(3.1) ( ) ( ) ( )( )11111 ,...,,..., mmgg xhxhxhhh ++++=
where gh is a production function of each import product or input; gx is the import
product or input. The groups run from 1 to m ; the number of inputs in each group is only
one to indicate that each import is a unique input that produces a unique output; the
number of inputs in group g is gn . The total number of products is mnn ++ ...1 .
Equation (3.1) implies that the elasticity of output with respect to each input is
independent of all other inputs; hence, all cross effects are zero. Let ( ).1h represents the
production function of, say, Florida orange juice. This function does not have the orange
juices of other countries as inputs because each individual input yields its own unique
output. Hence, the constrained cost minimization procedure will not yield a demand
function that consists of the prices of other orange juices. As a result, theoretically we
can’t investigate the relationship between Florida orange juice and other juices. The
presence of input independence in the production function precludes us from
investigating the substitution between imports of orange juice from different countries
and competition among exporting countries.
Consumer Demand Approach
The present study chooses the consumer approach over the production approach
since it allows investigating the nature of demand relationship among imported products
and competition among different exporters. Consumption theory is amenable to analyze
the market structure of commodities in fruit juice market. The theory involves the
analysis of the change in marginal utilities of a certain product due to a change in
consumption of a closely related product. The changes in marginal utilities are related to
the price substitution terms of demand functions.
29
Starting with a traditional utility function that is assumed to be well behaved
(twice differentiable, increasing in its arguments, strict concavity), we can derive the
Marshallian demand functions. They satisfy the properties of adding up, symmetry of the
cross price derivatives, homogeneous of degree zero in prices and expenditure, and
negative semi-definiteness in compensated price responses.
Utility Maximization
The maximization of a utility function ( )qu subject to a budget constraint
qpm '= is set up in a constrained optimization problem using the Lagrange method as
(3.2). ( ) ( ) ( )qpmquqL ', −+= λλ
where q is the vector of consumption products; λ is the Lagrange multiplier which can
be interpreted as the marginal utility of income; m is total expenditure; p is the vector
of prices.
The first order conditions are
(3.2.1) ( ) ( ) 0'.,=−≡
∂∂ pqu
qqL λλ
and
(3.2.2) ( ) 0',=−≡
∂∂ qpmqL
λλ .
The first order conditions imply that the marginal rate of substitution should equal
the price ratio at the optimum, which in turn implies that the internal rate of trade should
equal the external or market rate of trade. That is, a consumer will adjust purchases of
products until their willingness to trade one for the other just matches the rate at which
they can be traded in the marketplace, as given by the ratio of prices.
30
From the first order conditions, we derive the demand functions for all products i
and the marginal utility of income function as
(3.3) ( )pmfq ii ,=
and
(3.4) ( )pm,λλ = .
The choice of a functional form is at the interface of economic theory and the data.
In other words, the functional form should satisfy the economic proprieties and fit to a
statistical data satisfactorily. Two steps are followed in demand specification (Fousekis
and Revell, 2000). First, behavioral assumptions are imposed which lead to a cost or to
an indirect utility function. Second, a functional form is selected. Parsimony and
flexibility are desirable properties considered in the selection of functional forms.
The most common and parsimonious demand model, which dominated the import
demand literature in the past, was the Armington trade model. The application of the
Armington model to trade data dates back to the late 1970s and became popular in the
1980s and 1990s (Grennes et al. 1977, Sarris, 1981; Sarris, 1983; Abbot and Paarberg,
1986; Babula, 1987; Alston et al. 1990; Duffy et al. 1990; Haniotis, 1990). However,
the Armington trade model came to be increasingly criticized on both conceptual and
empirical grounds. The hypothesis of separability and homotheticity may not be
supported by import data (Alston, et al. 1990). Traditional methods of implementing the
Armington trade model result in theoretically and statistically inconsistent parameter
estimates (Davis and Kruse, 1993).
Consequently, system-wide demand models such as the Rotterdam model and the
Almost Ideal Demand Systems have come to be popular in the contemporary import
31
demand literature (Clements and Theil, 1978, Lee et al. 1990; Seale et al. 1992; Zhang et
al. 1994; Yang and Koo, 1994; Schmitz and Wahl, 1998; Fabiosa and Ukhova, 2000;
Soshnin et al. 1999; Schmitz and Seale, 2002; Washington and Kilmer, 2002).
The choice among different system-wide demand specifications (e.g., the
Rotterdam model versus AIDS model) is based on statistical tests (Brown et al. 1994).
Economic theory does not suggest a criterion to choose ex ante between demand models.
Barten (1993) demonstrates that the Rotterdam and AIDS models are special cases of a
general demand model so that nested tests can be applied to choose either the Rotterdam
or AIDS model or the hybrid of these two models (Central Statistical Bureau (CBS) and
National Bureau of Research (NBR)).
In the field of consumer demand analysis, the issue of selecting among competing
functional forms has been addressed in a number of recent studies (Eales et al. 1997; Lee
et al. 1994, Barten; 1993.; Schmitz and Seale, 2002; Weatherspoon and Seale, 1995).
They have demonstrated that a family of competing systems can be generated through
alternative parameterizations of Theil’s differential system (Theil 1980).
However, separability is an issue in estimating system-wide models (Seale, 1996).
The AIDS model is not globally separable and only becomes separable locally under
stringent conditions (Lee et al. 1994). This will render multi-stage demand estimation
difficult. However, it is not uncommon to find the application of the AIDS model in a
two-stage budgeting framework (Heien and Pick, 1991; Soshnin, et al. 1999). In these
two studies, the AIDS model was used for both the first and second stages. Other studies
have specified a two-stage demand system by applying the LES model for the first stage
and the AIDS model for the second stage (Fan, et al. 1995; Han and Wahl, 1998;
32
Michalek and Keyzer, 1992; Ma and Rae, 2003). Gao et al. (1996) specified a two-stage
demand by applying the extension of the AIDS model for the first stage and Generalized
Linear Expenditure System for the second stage.
The Rotterdam model, which is globally separable, has been applied in several
studies to specify a two-stage demand system. These include Duffy (1986); Clements
and Johnson (1983), Clements and Selvanathan (1988), Brown and Lee (1997), Xao et al.
(1998); E. Selvanathan and A- Selvanatha (2004). All of these studies have used the
Rotterdam model for both the first and second stage in a block independent framework
for different applications, mostly of advertising.
The present study prefers to use the Rotterdam model because of its global
separability. Unlike the previous studies which have applied the Rotterdam model, the
present study tests different separability hypotheses. The hypotheses will be discussed in
the next sections.
The Rotterdam Model
Following Theil (1980a, 1980b), the Rotterdam model is derived from the
maximization of a general utility function or total differentiation of a general demand
function.
Totally differentiating (3.3) yields
(3.5) j
N
j j
iii dp
pq
dmmq
dq ∑=
⎟⎟⎠
⎞⎜⎜⎝
⎛
∂∂
+∂∂
=1
.
Expressing (3.5) in log form ( )iii qdqqd =log yields
(3.6) ( ) ( ) ( ) ( )jj
N
j j
iiii pdp
pq
mdmmq
qdq logloglog1∑=
⎟⎟⎠
⎞⎜⎜⎝
⎛
∂∂
+∂∂
=
33
where ( )iqd log is the log change in quantity demanded of product i and ( )jpd log is the
log change in price of product j .
Based on Barten’s Fundamental matrix, the total substitution effects ⎟⎟⎠
⎞⎜⎜⎝
⎛
∂∂
j
i
pq
in
(3.6) can be decomposed into specific and general substitution terms as
(3.7) jijiij
j
i qmq
mq
mq
mu
pq
∂∂
−∂
∂
∂∂
∂∂−=
∂∂
λλλ
where iju is the ( )thji, element of 1−U the inverse of the Hessian; ijuλ is the specific
substitution effect, which shows that the corresponding component depends upon the
specific relation, in terms of iju between i and j . In other words, the utility obtained
from product i is a function of the consumption level of product j ; mq
mq
mji
∂
∂
∂∂
∂∂−
λλ is
the general substitution effect, which shows that all products are competing for the
consumer’s budget, and ji q
mq∂∂
− is the income effect of the price change jdp on the
demand for the thi product. Therefore, the total substitution effect of a price change can
be expressed as the sum of the substitution effect ⎟⎟⎠
⎞⎜⎜⎝
⎛∂
∂
∂∂
∂∂−
mq
mq
mu jiij
λλλ and income
effect ⎟⎠⎞
⎜⎝⎛
∂∂
− ji q
mq
and is known as the Slutsky equation. The component ijuλ of
ji pq ∂∂ is the effect on iq of a change in jp when the change is accompanied by an
income change so that the marginal utility of income remains unchanged.
34
Substituting (3.7) into (3.6) and multiplying both sides by mpi / , we find
(3.8)( ) ( )
( )jjijiij
N
j
ji
iiii
pdqmq
mq
mq
mu
mpp
mdmq
pqdw
log
loglog
1⎟⎟⎠
⎞⎜⎜⎝
⎛∂∂
−∂
∂
∂∂
∂∂−⎟⎟
⎠
⎞⎜⎜⎝
⎛
+∂∂
=
∑= λ
λλ
where iw is the expenditure share of product i defined as mqp
w iii = .
Multiplying out the second terms of the right-hand expression of (3.8) yields
(3.9)
( ) ( ) ( )
( ) ( ).loglog
logloglog
11
1
∑∑
∑
==
=
⎟⎟⎠
⎞⎜⎜⎝
⎛∂∂
−⎟⎟⎠
⎞⎜⎜⎝
⎛∂
∂
∂∂
∂∂
−⎟⎠⎞
⎜⎝⎛+
∂∂
=
N
jjj
ijiN
jj
jiji
j
N
jj
iji
iiii
pdqmq
mpp
pdmq
mq
mmpp
pdpupm
mdmq
pqdw
λλ
λ
The first term of the right-hand side expression of (3.9) is the marginal value
share defined as
(3.9.1) mq
p iii ∂∂
=θ .
The second term of the right-hand side expression of (3.9) is the relative price
coefficient ijv defined as
(3.9.2) jij
iij pupm
v λ= .
The third terms of the right hand-side expression of (3.9) can be rearranged to
yield the general substitution effect as
35
(3.9.3) ( ) ( )
( )j
N
jji
j
N
j
jj
ii
N
jj
jiji
pd
pdmq
pmmm
qppd
mq
mq
mmpp
log
loglog
1
11
∑
∑∑
=
==
=
⎟⎟⎠
⎞⎜⎜⎝
⎛∂
∂⎟⎠⎞
⎜⎝⎛
∂∂
∂∂
=⎟⎟⎠
⎞⎜⎜⎝
⎛∂
∂
∂∂
∂∂
θφθ
λλ
λλ
where φλ
λ=⎟
⎠⎞
⎜⎝⎛
∂∂ −1m
m, which is the reciprocal of the income elasticity of the marginal
utility of incomeλ .
The fourth terms of the right hand-side expression of (3.9) can be rearranged to
yield the income effect of a price change as
(3.9.4) ( ) ( ) ( )j
N
jjij
N
j
jjii
N
jjj
iji pdwpdmqp
mq
ppdqmq
mpp
logloglog111∑∑∑===
=∂∂
=⎟⎟⎠
⎞⎜⎜⎝
⎛∂∂
θ .
Substituting (3.9.1) through (3.9.4) and rearranging them yields
(3.10) ( ) ( ) ( ) ( ) ( )∑∑∑===
−+⎟⎟⎠
⎞⎜⎜⎝
⎛−=
N
jjjij
N
jij
N
jjjiii pdpdvqdwmdqdw
111logloglogloglog θφθθ .
Rearranging (3.10) and using the constraint that the sum of the relative price
coefficients is proportional to the marginal value share i
N
jijv φθ=∑
=1, we find the relative
price version of the Rotterdam model (3.11) and the absolute price version of the
Rotterdam model (3.12) as
(3.11) ( ) ( ) ⎟⎟⎠
⎞⎜⎜⎝
⎛+= ∑
= Pp
dvQdqdw jN
jijiii logloglog
1θ .
(3.12) ( ) ( ) ( )j
N
jijiii pdQdqdw logloglog
1∑=
+= πθ .
where ( ) ( ) ( )⎟⎟⎠
⎞⎜⎜⎝
⎛−= ∑
=
N
jjj qdwmdQd
1
logloglog is the real income term; jiijij v θφθπ −=
are the Slutsky price coefficients; ( )∑=
=N
jjj pdP
1log)(log θ is the Frisch price index.
36
In order to identify the market structure underlying the importation of fruit juices
into Japan, four different versions of the Rotterdam model are derived from the relative
price version of the Rotterdam Model under different hypotheses. The hypotheses which
represent different market structures are block independence, block-wise dependence,
and uniform substitutes. The block independence and block-wise dependence hypotheses
are applied to products that belong to different product groups while the uniform
hypothesis is applied to products within the same product group.
The models derived under these hypotheses in this study are block independent
non-uniform substitute-Rotterdam model, block independent uniform-substitute-
Rotterdam model, block-wise dependent non-uniform substitute-Rotterdam model and
block-wise dependent uniform substitute-Rotterdam model.
Block Independence
Block independence is a special case of strong separability where one can group
commodities into different blocks depending on some tangible criterion. Separability is a
relative concept whose frame of reference is some partition of a product set into mutually
exclusive and exhaustive subsets. Blundell and Robin (2000) indicate that the idea
behind separability in consumer preferences is the existence of “natural” groupings of
related commodities that reflect the budgeting decisions consistent with the true
preference ordering of the representative consumer. Otherwise, empirical estimates of
structural demand parameters are invalid. The usefulness of separability depends on the
ability to classify products into groups which are empirically valid (Barten, 1977).
The grouping of commodities into blocks is of paramount significance from a
statistical point of view since it increases the degrees of freedom. However, the blocking
37
has to be theoretically consistent and empirically plausible. Suppose that we have
NG < blocks or groups denoted as GSS ,...,1 such that each product belongs to exactly
one group, the consumer’s preferences under block independence is represented by the
sum of G sub-utility functions, each involving the quantities of only one group given as
(3.13) ( ) ( ) ( )( )mg mnmmgnggn qquqquqquuu ,...,,...,,...,,...,,..., 111111 1
++++=
where gu is a sub-utility function; gq is a sub-vector of q which consists of the sqi' that
fall under ( )GgS g ,...,1= . The groups run from 1 to m ; the number of commodities in
group 1 is 1n ; the number of commodities in group g is gn . The total number of products
is mnn ++ ...1 .
Under (3.13), the utility obtained from the products in group g is independent of
the utility of products in group h. That is 02
=∂∂
∂
ji qqu . However, for i and j in the same
group 02
≠∂∂
∂
ji qqu . That is, the consumption of an extra unit of product j has an effect
on the utility of product i and vice versa.
Formally, the hypothesis of block independence ( )0H states that the change in the
marginal utility of a dollar spent on the thi product ( )gSi∈ caused by an extra dollar
spent on the thj product which belongs to a different group ( )hgSj h ≠∈ , equals zero.
The alternative hypothesis states that the change in the marginal utility of a dollar spent
on the thi product ( )gSi∈ caused by an extra dollar spent on the thj product which
belongs to a different group ( )hgSj h ≠∈ , is different from zero.
38
( ) ( ) 0:2
0 =∂∂
∂
jjii qpqpuH for gSi∈ and hSj∈ ; hg ≠ ,
( ) ( ) 0:2
≠∂∂
∂
jjiiA qpqp
uH for gSi∈ and hSj∈ ; hg ≠ .
Under this hypothesis, the Hessian ( )ji qqu ∂∂∂ 2 and its inverse ( ) 12 −∂∂∂ ji qqu
becomes a block diagonal. The marginal utility of each product depends only on the
quantities consumed of the products that belong to the same group (Theil, 1975).
Following Theil (1975), the changes in the marginal utilities can be related to
demand parameters as jij
iij pupm
v λ= where ijv are the relative price coefficients. When
i and j belong to different product groups, ijv can be set equal to zero because iju
equals zero under the assumption of block independence.
This implies that the assumption of block independence represents a market
structure whereby the change in the relative price of a product in one product group does
not affect the demand for another product in another product group. For instance, under
this market structure, we are hypothesizing that the change in the price of U.S. grapefruit
juice does not affect the demand for Brazilian orange juice. Orange juice and grapefruit
juices are in different product groups.
Block independent non-uniform substitute-Rotterdam Model. Following
Theil (1980a), the block independent non-uniform substitute-Rotterdam model can be
derived from (3.11) by setting ijv equal to zero for i and j that belong to different groups
as
(3.14) ( ) ( ) ∑∈
⎟⎟⎠
⎞⎜⎜⎝
⎛+=
gSj
jijiii P
pdvQdqdw logloglog θ .
39
Since all ijv with i and j in different groups vanished, the number of free
parameters is obviously reduced. However, no product is a specific substitute or
complement of any product that belongs to a different group. The demand equation of
the thi product contains gN relative prices when it belongs to set gS . The number of free
parameters depends on the number of blocks and the number of commodities in each
block. Theil (1980) shows that with G blocks having N commodities in total and an
equal number of commodities in each block, the number of free parameters is
( )GNN /15.0 + .
Block-wise Dependence
In the previous section, we have assumed that the consumer’s utility function can
be additively separated into group utility functions. A weaker assumption is that the
consumer utility function ( )qu equals some function ( )f rather than the sum of the
group utility functions.
(3.15) ( ) ( ) ( )( )mg MnMMGnGGn qquqquqquuu ,...,,...,,...,,...,,..., 111111 1
= .
Unlike the case with (3.13), the utility obtained from a product in one group under
(3.15) is not independent of the consumption of another product in another group. That is,
for i and j in different groups, 02
≠∂∂
∂
ji qqu . Since we are dealing with products in each
group on a block-wise basis, we are assuming that the effect of the consumption of an
extra unit of product j ( )hSj∈ on the marginal utility of product i ( )hgSi g ≠∈ ; is the
same for all pairs of products in the two product groups; i.e., this effect is independent of
i and j .
40
Formally, the hypothesis of block-wise dependence ( )0H states that the change in
marginal utility of a dollar spent on the thi product ( )gSi∈ caused by an extra dollar
spent on the thj product which belongs to a different group ( )hgSj h ≠∈ , equals some
constant gha ; i.e., this effect is independent of i and j and hence, the same for all pairs
of products in the two product groups. For instance, in the orange and apple juice groups,
an extra dollar spent on either U.S. orange juice or Brazilian orange juice in the orange
juice group has the same effect on the marginal utility of a dollar spent on Chinese apple
juice or Austrian apple juice in the apple juice group. The utility interaction of two
products of different groups in a block-wise dependence framework is a matter of the
groups rather than the individual products (Theil, 1980a).
( ) ( ) ghjjii qpqpuH α=∂∂∂ 20 : for all )(, hghjgi ≠∈∈ .
( ) ( ) ghjjiiA qpqpuH α≠∂∂∂ 2: for all )(, hghjgi ≠∈∈ .
Following Theil (1975), the changes in marginal utilities can be related to the
relative price coefficients ( )ijv as
(3.16) ( ) ( ) hghgjj
h
ii
g
hg
ij
uuuuuuum
qpu
qpu
uuumv
∂∂∂∂∂∂∂
=∂∂
∂
∂
∂∂∂
=λλ
λλφφ
22
,
where gSi∈ ; hSj∈ ; and hg ≠ .
Equation (3.16) shows that the cross-group term is the same for all pairs of
products from different groups. Following Theil (1975) and Brown (1993), the relative
price coefficients corresponding to (3.16) can be given as
(3.17) jighijv θθφΦ−=
41
where ijv is the relative price coefficient; ghΦ is a factor of proportionality which is the
same for all gSi∈ and hSj∈ ; iθ is the marginal expenditure share.
This implies that the assumption of block-wise dependence represents a market
structure whereby the change in the price of a product in one group would affect the
demand for another product in another product group in the same fashion. In other words,
the effect of a change in the price of a product in group A on the demand for another
product in group B is the same for all pairs of products in the two groups. For instance,
under this market structure, we are hypothesizing that the effect of a change in the price
of U.S. orange juice on the demand for Israelis grapefruit juice is the same as that of
Brazilian orange juice on the demand for U.S. grapefruit juice.
Block-wise dependent non-uniform substitute-Rotterdam model. Following
Theil (1980a), the block-wise dependent non-uniform substitute-Rotterdam model can be
derived from (3.11) as
(3.18) ( ) ( ) ⎟⎠⎞
⎜⎝⎛+⎟⎟
⎠
⎞⎜⎜⎝
⎛+= ∑∑
≠∈ P
PdV
Pp
dQdqdw h
ghghiSj
jijiii
gloglogloglog 'θνθ .
where iθ is the marginal expenditure share; ijv is the relative price coefficient,
which applies for the products within a group; 'iθ is the conditional marginal expenditure
share; ghV is the group relative price coefficient defined as ∑∑∈ ∈
=gi hj
ijgh vV , where hg ≠ .
Substituting equation (3.17) for ijv , we can write the group price coefficient as
hgghghV ΘΘΦ−= φ where ∑∈
=Θgi
ig θ and ∑∈
=Θhj
jh θ are the gΘ are the group marginal
expenditure shares of group g and h , respectively.
42
Uniform Substitute Hypothesis
In the previous two sections, no restriction was imposed within the groups of
commodities, but weak separability prevails between groups. Now, we impose a testable
restriction (uniform hypothesis) on products within a given group. A group of closely-
related products are uniform substitutes when the cross effect of an additional dollar
spent on one product on the marginal utility of another dollar spent on another product is
the same for all pairs of products in the group (Brown, 1993). The uniform substitute
hypothesis was initially proposed by Theil (1980a) to deal with the demand for closely
related products such as different brands of a product. Since the same products that are
imported from different countries can be treated like different brands of the same product,
the application of this hypothesis to the same product differentiated by country of origin
is relevant. We consider this hypothesis given block independence and block-wise
dependence framework discussed earlier. In other words, we impose the uniform
substitute hypothesis on (3.14) and (3.18).
Uniform substitute given block independence
Suppose that we have a product group gS that consists of the same product
differentiated by country of origin of production. The consumer’s preferences for a
uniform product given block independence can be represented by the sum of G sub-
utility functions, each involving the quantities of only one group given as
(3.19) ( ) ( ) ( )( )mg mnmmgnggn qquqquqquuu ,...,,...,,...,,...,,..., 111111 1
++++= .
Under (3.19), the utility that a consumer obtains from the products in one group is
independent of the utility of products in another group. That is, 02
=∂∂
∂
ji qqu for gSi∈
43
and hSj∈ . However, for the thi and ths products in the same group, we are
hypothesizing that the marginal utility of a dollar spent on the thi product ( )gSi∈ caused
by an extra dollar spent on the ths product which belongs to the same group ( )gSs∈ is
the same because i and s are the same products differentiated by country of origin of
production.
Theil (1980a) writes the submatrix of the Hessian of the utility function ijθ in
expenditure terms, multiplied by the scalar λφ m. as
[ ] ( ) ( )⎥⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢⎢
⎣
⎡
=⎥⎦
⎤⎢⎣
⎡∂∂
∂=
nnssii
ij
kk
kkkk
qpqpum
θ
θθ
λφθ
L
MOMM
L
L22
11
2. ,
where all the off-diagonal elements(ij
ij
θθ 1
= , ji ≠ ) are equal to a positive constant k .
Since λφ m. is negative, this type of preference structure implies that the marginal utility
of a dollar spent on each product in gS is affected negatively and by the same amount
mk .φλ when an additional dollar is on any other product in the group. Thus, all
products in gS are affected uniformly by the additional consumption of any other
products in the group.
Since we have between-group (block independence) in addition to the within
group restrictions (uniform substitute), we have two null hypotheses. The block
independence hypothesis has to do with the products between two product groups, and
the uniform substitute has to do with products within the same group. Note that the
44
uniform substitute hypothesis in this study is applied to the same product differentiated
by country of origin.
Formally, the null hypothesis of a uniform substitute relationship states that the
marginal utility of a dollar spent on the thi product ( )gSi∈ caused by an extra dollar
spent on the ths product which belongs to the same group ( )gSs∈ equals some positive
constant k except when si = , i.e., this effect is independent of i and s and hence, the
same for all pairs of products in the same group.
( ) ( ) g
2
0 Ss i,for : ∈=∂∂
∂ kqpqp
uHssii
.
The block independence hypothesis states that the marginal utility of a dollar
spent on the thi product ( )gi∈ caused by an extra dollar spent on the thj product which
belongs to a different group ( )hghj ≠∈ , equals zero.
( ) ( ) 0: H2
0 =∂∂
∂
jjii qpqpu for gSi∈ and hSj∈ ; hg ≠ .
Combining the two null hypotheses corresponding to the uniform substitute and
block independence, the new null hypothesis which corresponds to the uniform substitute
hypothesis given block independence can be restated as
( ) ( ) ( ) ( ) 0 ; Ss i,for :2
g
2
0 =∂∂
∂∈=
∂∂∂
jjiissii qpqpuk
qpqpuH for gSi∈ and hSj∈ ;
hg ≠ .
( ) ( ) ( ) ( ) 0 ; Ss i,for :2
g
2
=∂∂
∂∈≠
∂∂∂
jjiissiiA qpqp
ukqpqp
uH for gSi∈ and
gSj∈ ; hg ≠ .
Theil (1980a) derives the relative price coefficients of a block independent
uniform substitute model as
45
( )
⎪⎪⎩
⎪⎪⎨
⎧
≠Θ−
−=
=Θ−
−=
jik
kv
jik
kv
g
jiij
g
iiij
1
11
)20.3(θθ
φ
θθφ
where ijv is the relative price coefficients; iθ is the marginal value share, k is a
constant; gΘ is the group marginal value share; φ is the income flexibility.
Block independent Uniform Substitute-Rotterdam Model. Substituting the
price substitution terms (3.20) in the block independent non-uniform substitute-
Rotterdam model (3.14), the block independent uniform substitute-Rotterdam model can
be derived as
(3.21) ( ) ( ) ( )⎥⎥⎦
⎤
⎢⎢⎣
⎡⎟⎟⎠
⎞⎜⎜⎝
⎛Θ−
−+⎟
⎠
⎞⎜⎝
⎛Θ−
−+= ∑
∈≠ gSij
j
g
jii
g
iiiii P
pd
kk
Pp
dk
kQdqdw log
1log
11
loglogθθθθ
φθ .
where iθ is the unconditional marginal value share; φ is the income flexibility; k is a
constant; gΘ is the group marginal value share;
Uniform substitute given block-wise dependence
The consumer’s preferences for a uniform product given block-wise dependence
can be represented by consumer utility function ( )qu equals some function ( )f rather
than the sum of the group utility functions.
(3.22) ( ) ( ) ( )( )mg MnMMGnGGn qquqquqquuu ,...,,...,,...,,...,,..., 111111 1
= .
Under (3.24), the utility that a consumer obtains from the products in one group is
not independent of the utility of products in another group. That is, for thi and thj
products in two different groups, 02
≠∂∂
∂
ji qqu . The consumption of product i has an
46
effect on that of product j and vice versa. For the thi and ths products in the same
group, 02
≠∂∂
∂
si qqu .
Since we have between-group (block-wise dependence) and within group
(uniform substitute) restrictions, we have two null hypotheses. The block-wise
dependence hypothesis has to do with the products between product groups, and the
uniform substitute has to with products within the same group.
Formally, the null hypothesis of a uniform substitute relationship states that the
marginal utility of a dollar spent on the thi product ( )gSi∈ caused by an extra dollar
spent on the ths product which belongs to the same group ( )gSs∈ equals some constant
k , i.e., this effect is independent of i and s and hence, the same for all pairs of products
in the same group.
( ) ( ) sikqpqp
uHssii
≠∈=∂∂
∂ ;Ss i,for : g
2
0 .
The block-wise dependence hypothesis states that the marginal utility of a dollar
spent on the thi product ( )gSi∈ caused by an extra dollar spent on the thj product which
belongs to a different group ( )hgSj h ≠∈ , equals some constant gha .
( ) ( ) ghjjii
aqpqp
u=
∂∂∂ 2
0 : H for gSi∈ and hSj∈ ; hg ≠ ,
Combining the two null hypotheses corresponding to the uniform substitute and
block-wise dependence, the new null hypothesis which corresponds to the uniform
substitute hypothesis given block-wise dependence can be restated as
47
( ) ( ) ( ) ( ) ghjjiissii
aqpqp
ukqpqp
uH =∂∂
∂∈=
∂∂∂ 2
G
2
0 ; Ss i,for : for gSi∈ and
hSj∈ ; hg ≠ ,
( ) ( ) ( ) ( ) ghjjiissii
A aqpqp
ukqpqp
uH =∂∂
∂∈≠
∂∂∂ 2
g
2
; Ss i,for : for gSi∈ and hSj∈ ;
hg ≠ .
Seale (2003) derives the relative price coefficients for a block-wise dependent
uniform substitute Rotterdam model as
( )
⎪⎪
⎩
⎪⎪
⎨
⎧
≠Θ−
ΘΘ−=
=Θ−
Θ−Θ=
jik
kv
jik
kv
g
ggjggiij
gg
ggiggiij
1
11
)23.3(''
''
θθφ
θθφ
where ijv is the relative price coefficients; φ is the income flexibility; 'iθ is the
conditional marginal value share, k is a constant; ggΘ is the group marginal value share.
Block-wise dependent Uniform Substitute-Rotterdam Model Substituting the
price substitution terms (3.23) in the block-wise-Rotterdam model (3.18), the block-wise
dependent uniform substitute-Rotterdam model can be derived as
(3.24)( ) ( )
⎟⎠⎞
⎜⎝⎛
+⎟⎟⎠
⎞⎜⎜⎝
⎛Θ−
−+⎟
⎠⎞
⎜⎝⎛
⎟⎟⎠
⎞⎜⎜⎝
⎛
Θ−−
+=
∑
∑
≠
∈≠
PP
dV
Pp
dk
kPp
dkk
Qdqdw
h
ghghi
Sij
j
gg
jiii
ggi
iiiiii
g
log
log1
log11
loglog
'θ
θθφ
θφθθ
The uniform substitute restriction results in a substantial reduction in the
parameter space and can be useful for obtaining more precise parameter estimates and
maintaining sufficient degrees of freedom (Brown, 1993).
48
CHAPTER 4 EMPIRICAL MODELS AND ESTIMATION PROCEDURES
Since the differential approach to consumption theory discussed in chapter 3 does
not postulate constancy for the coefficients of its demand equations, we are not entitled
yet to talk about empirical estimation. In this chapter, we discuss the ways in which the
theoretical models in chapter 3 are parameterized so that they can be applied to statistical
data. Since the nature of data forces us to work with finite rather than infinitesimal
changes, we replace the infinitesimal changes by finite changes. Furthermore, we
postulate that the coefficients are constant to make the models operational. Finally,
estimation procedures are presented for the different versions of the Rotterdam model.
Empirical Models
The Relative Price Version of the Rotterdam Model
Following Theil (1975), the relative price version of the Rotterdam model (3.11)
can be written in finite changes as
(4.1) itt
jtN
jijtiitit dP
dpvdQdqw εθ +⎟⎟
⎠
⎞⎜⎜⎝
⎛+= ∑
=1
.
where ( ) 212, −+= tiitit www is the average expenditure share ; ( )12,log −= tiitit qqdq is the
finite change in quantity imported of product i ; iθ is the marginal expenditure share of
product i ; ttttt dqwdqwdQ 181811 ...++= is the finite change version of the Divisia price
index (real income) ; ijv is the relative (Frisch-deflated) price coefficients;
49
( )12,log −= tjjtjt ppdp is the finite change in price of product j ;
ttt dpdpdP 181811 ... θθ ++= is the finite change version of the Frisch price index; Note
that the lower case p is for prices of individual products and the upper case P is for
Divisia price indices. itε is the demand disturbance, which is regarded as the random
effect of all variables other than income and prices. It is assumed that it has zero
expectation, that the variances and contemporaneous covariances are constant over time,
and that all other covariances vanish as
(4.2) ( )
tsiftsifijjtis
≠=
==
0ϖεεξ
The coefficients of (4.1) are subject to the adding-up constraint∑ =i i 1θ and the
symmetry constraint jiij vv = and negative definiteness of the matrix ijv . Furthermore,
the sum of the relative price coefficients is proportional to the marginal expenditure
shares ij
ijv φθ=∑=
18
1, where φ is the income flexibility.
In this study, six fruit juice groups (orange, grapefruit, other citrus, apple,
pineapple and grape juices) imported from 18 countries with three countries for each
juice group are included (Table 4.1).
In order to estimate (4.1), three steps are followed. First, one of the 18 demand
equations is deleted in order to eliminate singularity. Second, the constraint on the price
coefficient ij
ijv φθ=∑=
18
1 is imposed and third, the adding up constraint 1
18
1=∑
=kkθ is
imposed on the income coefficients.
50
Table 4.1 Codes for countries exporting fruit juice to Japan Product
Exporting country
Quantity log changes
Price log changes
Budget shares
Code
U.S. dq1 dp1 w1 1 Brazil dq2 dp2 w2 2
Orange juice
ROW dq3 dp3 w3 3
U.S. dq4 dp4 w4 4 Israel dq5 dp5 w5 5
Grapefruit juice
ROW dq6 dp6 w6 6
U.S. dq7 dp7 w7 7 Austria dq8 dp8 w8 8
Apple juice
ROW dq9 dp9 w9 9
Thailand dq10 dp10 w10 10 Philippines dq11 dp11 w11 11
Pineapple juice
ROW dq12 dp12 w12 12
U.S. dq13 dp13 w13 13 Argentina dq14 dp14 w14 14
Grape juice
ROW dq15 dp15 w15 15
Italy dq16 dp16 w16 16 Israel dq17 dp17 w17 17
Other citrus juice
ROW dq18 dp18 w18 18 aROW means rest of the world
Imposing the constraint ij
ijv φθ=∑=
18
1 on the price coefficients, we write the own
price coefficient iiv in terms of the other price coefficients as ∑≠
−=18
ijijiii vv φθ so that the
price term of (4.1) becomes
(4.3) ( ) ( )∑∑≠≠
−+−⎟⎟⎠
⎞⎜⎜⎝
⎛−=
1818
ijjiji
ijiji dPdpvdPdpvφθ
( ) ( )∑≠
−+−=ij
ijijii dpdpvdPdpφθ
51
Imposing the adding-up constraint ∑=
−=17
118 1
ikθθ on the income coefficients, the
price substitution term of (4.3) becomes
(4.4) ( ) ( )∑∑≠=
−+⎟⎠
⎞⎜⎝
⎛−−−=
1817
118 18
ijijij
kkkii dpdpvdpdpdpdp θφθ .
Substituting (4.4) into (4.1), we obtain
(4.5) ( ) ( ) iij
ijijiiiii dpdpvAdqdqw εθφθ +−++= ∑≠
where ( ) ( )⎟⎠
⎞⎜⎝
⎛−−−= ∑
=
17
118 18
kkkiii dpdpdpdpA θθθ .
Equation (4.1) is still not estimable unless conditions are imposed on the matrix of
price coefficients in addition to symmetry and negative definiteness, such as preference
independence and/or block independence (Theil, 1980a). As soon as there is one
constraint on the price coefficients such as 012 =ν (preference independence between
product 1 and product 2) in addition to symmetry and negative definiteness, it is possible
to estimate (4.1). The reason that (4.1) is still not estimable is that the income flexibility
φ is not identified because of its invariance under monotone transformation of the
consumer’s utility function (Theil, 1980) in which case there may not be unique demand
functions.
Equation (4.1) can be estimated using the following system of symmetry-
constrained equations.
( ) ( ) ( )( ) ( ) ( )( ) ( ) ( )⎪
⎩
⎪⎨
⎧
−++−+−++=−++−+−++=
−++−+−++=
)(...)(...
)(...
318318322331133333
218218232321122222
118118131312121111
dpdpvdpdpvdpdpvAdQdqwdpdpvdpdpvdpdpvAdQdqw
dpdpvdpdpvdpdpvAdQdqw
θφθθφθθφθ
52
( ) ( ) ( )( ) ( ) ( )( ) ( ) ( )⎪
⎩
⎪⎨
⎧
−++−+−++=−++−+−++=−++−+−++=
)(...)(...)(...
618618622661166666
518518522551155555
418418422441144444
dpdpvdpdpvdpdpvAdQdqwdpdpvdpdpvdpdpvAdQdqwdpdpvdpdpvdpdpvAdQdqw
θφθθφθθφθ
( ) ( ) ( )( ) ( ) ( )( ) ( ) ( )⎪
⎩
⎪⎨
⎧
−++−+−++=−++−+−++=−++−+−++=
)(...)(...)(...
918918922991199999
818818822881188888
718718722771177777
dpdpvdpdpvdpdpvAdQdqwdpdpvdpdpvdpdpvAdQdqwdpdpvdpdpvdpdpvAdQdqw
θφθθφθθφθ
( ) ( ) ( )( ) ( ) ( )( ) ( ) ( )⎪
⎩
⎪⎨
⎧
−++−+−++=−++−+−++=−++−+−++=
)(...)(...)(...
1218121812221212111212121212
1118111811221111111111111111
1018101810221010111010101010
dpdpvdpdpvdpdpvAdQdqwdpdpvdpdpvdpdpvAdQdqwdpdpvdpdpvdpdpvAdQdqw
θφθθφθθφθ
( ) ( ) ( )( ) ( ) ( )( ) ( ) ( )⎪
⎩
⎪⎨
⎧
−++−+−++=−++−+−++=−++−+−++=
)(...)(...)(...
1518151815221515111515151515
1418141814221414111414141414
1318131813221313111313131313
dpdpvdpdpvdpdpvAdQdqwdpdpvdpdpvdpdpvAdQdqwdpdpvdpdpvdpdpvAdQdqw
θφθθφθθφθ
( ) ( ) ( )( ) ( ) ( )⎩
⎨⎧
−++−+−++=−++−+−++=
)(...)(...
1718171817221717111717171717
1618161816221616111616161616
dpdpvdpdpvdpdpvAdQdqwdpdpvdpdpvdpdpvAdQdqw
θφθθφθ
The above system of equations provides the specific price substitution effect. The
specific substitution effect accounts for the n price changes on the demand for the
thi product (Theil, 1980). The specific substitution effect is one component of the effect
of a change in price. In order to estimate the total price substitution effect, one needs to
estimate the absolute price version of the Rotterdam model. The total substitution effect
is the sum of the specific and general substitution effect. The general substitution effect is
concerned with the competition of all products for an extra dollar of the consumer’s
income.
The Absolute Price Version of the Rotterdam Model
The absolute price version of the Rotterdam model (3.12) can be written in finite
changes as
53
(4.7) ij
N
jijiii dpdQdqw επθ ++= ∑
=1
where jiijij v θφθπ −= .
The Slutsky price coefficients ijπ are symmetric negative semi definite of rank n-
1 and satisfy the homogeneity property 01
=∑=
n
jijπ . A major convenience of the absolute
price version is its linearity in the parameters, thus implying that a least-square regression
estimation yields best linear unbiased parameter estimates when (1) the explanatory
variables take non-stochastic values; (2) the disturbances have zero means and a constant
contemporaneous covariance matrix and are serially uncorrelated; and (3) the
homogeneity condition and the Slutsky symmetry condition are ignored.
A disadvantage of the absolute price version of the Rotterdam model is that the
number of the Slutsky price coefficients ijπ grows rapidly when the number of
commodities N increases. The number of free parameters in (4.7) is given by 2NN + ,
where N is the number of commodities. Even after imposing the restrictions of adding
up, homogeneity, and symmetry, the number of free parameters is reduced
to ( )( )125.0 −+ NN .
The absolute price version of the Rotterdam model with symmetry and
homogeneity imposed can be estimated using the following system of equations.
( ) ( ) ( )[ ]( ) ( ) ( )[ ]( ) ( ) ( )[ ]⎪
⎩
⎪⎨
⎧
−++−+−+=−++−+−+=−++−+−+=
18173181822318113333
18172181822218112222
18171181821218111111
......
...)8.4(
dpdpdpdpdpdpdQdqwdpdpdpdpdpdpdQdqw
dpdpdpdpdpdpdQdqw
πππθπππθπππθ
54
( ) ( ) ( )[ ]( ) ( ) ( )[ ]( ) ( ) ( )[ ]⎪
⎩
⎪⎨
⎧
−++−+−+=−++−+−+=−++−+−+=
18176181822618116666
18175181822518115555
18174181822418114444
...
......
dpdpdpdpdpdpdQdqwdpdpdpdpdpdpdQdqwdpdpdpdpdpdpdQdqw
πππθπππθπππθ
( ) ( ) ( )[ ]( ) ( ) ( )[ ]( ) ( ) ( )[ ]⎪
⎩
⎪⎨
⎧
−++−+−+=−++−+−+=−++−+−+=
18179181822918119999
18178181822818118888
18177181822718117777
.........
dpdpdpdpdpdpdQdqwdpdpdpdpdpdpdQdqwdpdpdpdpdpdpdQdqw
πππθπππθπππθ
( ) ( ) ( )[ ]( ) ( ) ( )[ ]( ) ( ) ( )[ ]⎪
⎩
⎪⎨
⎧
−++−+−+=−++−+−+=−++−+−+=
18171218182212181112121212
18171118182211181111111111
18171018182210181110101010
.........
dpdpdpdpdpdpdQdqwdpdpdpdpdpdpdQdqwdpdpdpdpdpdpdQdqw
πππθπππθπππθ
( ) ( ) ( )[ ]( ) ( ) ( )[ ]( ) ( ) ( )[ ]⎪
⎩
⎪⎨
⎧
−++−+−+=−++−+−+=−++−+−+=
18171518182215181115151515
18171418182214181114141414
18171318182213181113131313
...
......
dpdpdpdpdpdpdQdqwdpdpdpdpdpdpdQdqwdpdpdpdpdpdpdQdqw
πππθπππθπππθ
( ) ( ) ( )[ ]( ) ( ) ( )[ ]⎩
⎨⎧
−++−+−+=−++−+−+=
18171718182217181117171717
18171618182216181116161616
...
...dpdpdpdpdpdpdQdqwdpdpdpdpdpdpdQdqw
πππθπππθ
Block Independent Non-uniform Substitute-Rotterdam Model
The block independent non-uniform substitute-Rotterdam model (3.14) can be
written in finite changes as
(4.9) iSj
jijiii
gdPdp
vdQdqw εθ +⎟⎟⎠
⎞⎜⎜⎝
⎛+= ∑
∈
.
The estimation procedure of (4.9) is similar to that of (4.1) presented earlier. In
order to estimate (4.9), one of the 18 demand equations is deleted to eliminate singularity.
Using the constraint iSj
ijg
v φθ=∑∈
, we write the own price coefficient iiv in terms of
the other price coefficients as ∑≠
−=ij
ijiii vv φθ so that the price term of (4.9) becomes
55
(4.10) ( ) ( )∑∑∈≠∈≠
−+−⎟⎟⎠
⎞⎜⎜⎝
⎛−=
gg Sijjiji
Sijiji dPdpvdPdpvφθ .
( ) ( )∑∈≠
−+−=gSji
ijijii dpdpvdPdpφθ .
Now, using the adding up constraint ∑=
−=18
118 1
kkθθ ,(4.10) becomes
(4.11) ( ) ( )∑∑∈=
−+⎟⎠
⎞⎜⎝
⎛−−−=
gSjijij
kkkii dpdpvdpdpdpdp
17
118 18θφθ .
Substituting (4.11) into (4.9) yields
(4.12) ( ) ( ) iij
ijijiiiiii dpdpvAdqdqw εθφθθ +−++= ∑≠
Equation (4.9) can thus be estimated using the following symmetry-constrained
system of equations.
( ) ( ) ( )( ) ( ) ( )( ) ( ) ( )⎪
⎩
⎪⎨
⎧
−+−++=−+−++=−+−++=
322331133333
232321122222
131312121111
dpdpvdpdpvAdQdqwdpdpvdpdpvAdQdqw
dpdpvdpdpvAdQdqw
θφθθφθθφθ
( ) ( ) ( )( ) ( ) ( )( ) ( ) ( )⎪
⎩
⎪⎨
⎧
−+−++=−+−++=−+−++=
555664466666
565654455555
464645454444
dpdpvdpdpvAdQdqwdpdpvdpdpvAdQdqwdpdpvdpdpvAdQdqw
θφθθφθθφθ
( ) ( ) ( )( ) ( ) ( )( ) ( ) ( )⎪
⎩
⎪⎨
⎧
−+−++=−+−++=−+−++=
988997799999
898987788888
797978787777
dpdpvdpdpvAdQdqwdpdpvdpdpvAdQdqwdpdpvdpdpvAdQdqw
θφθθφθθφθ
56
( ) ( ) ( )( ) ( ) ( )( ) ( ) ( )⎪
⎩
⎪⎨
⎧
−+−++=−+−++=−+−++=
121111121012101212121212
111211121110101111111111
101210121011101110101010
dpdpvdpdpvAdQdqwdpdpvdpdpvAdQdqwdpdpvdpdpvAdQdqw
θφθθφθθφθ
( ) ( ) ( )( ) ( ) ( )( ) ( ) ( )⎪
⎩
⎪⎨
⎧
−+−++=−+−++=−+−++=
151414151513131515151515
141514151413131414141414
131513151314131413131313
dpdpvdpdpvAdQdqwdpdpvdpdpvAdQdqwdpdpvdpdpvAdQdqw
θφθθφθθφθ
( ) ( ) ( )( ) ( ) ( )⎩
⎨⎧
−+−++=−+−++=
171817181716161717171717
161816181617161716161616
dpdpvdpdpvAdQdqwdpdpvdpdpvAdQdqw
θφθθφθ
Block-wise Dependent Non-uniform Substitute-Rotterdam Model
The block-wise dependent non-uniform substitute-Rotterdam model (3.18,) can be
written in finite changes as
(4.14) ih
ghgh
Sji
iSj
jijiii dP
dPV
dPdp
dQdqw
g
gε
θθ
νθ +⎟⎠⎞
⎜⎝⎛+⎟⎟
⎠
⎞⎜⎜⎝
⎛+= ∑∑∑
≠∈
∈
6
where ijv is the specific price coefficients of products within in a group; ghV is group
relative price coefficients; hdP is the Frisch price index of a group, and iε is the error
term.
The estimation procedure of equation (4.14) is also similar to that of (4.1). In
order to estimate (4.14), one of the 18 demand equations is deleted to eliminate
singularity.
Using the constraint that the sum of the price substitution terms is proportional to
the marginal value share ih
gh
Sjj
i
Sjij V
g
g
φθθ
θν =+ ∑∑∑
≠∈
∈ 1, we write the own price coefficient
57
iiv in terms of the other price coefficients as ∑∑∑≠
∈∈
−−=6
ghgh
Sjj
i
Sjijiii Vv
g
gθ
θνφθ and
substitute it in equation (4.14) so that the first price term of equation (4.14) which
corresponds to the within group demand becomes
(4.15) ( ) ( )∑∑∑∑∈≠≠
∈∈
−+−⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜
⎝
⎛
−−=g
g
g Sijjiji
ghgh
Sjj
i
Sjiji dPdpvdPdpV
6
θθ
νφθ
( ) ( )∑∑∑ ∈≠≠∈
−+−⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜
⎝
⎛
−=g
g
Sijijiji
ghgh
Sji
ii dpdpvdPdpV
6
θθ
φθ
Now collecting equation (4.15) and the second price term of equation (4.14), the
price substitution terms of equation (4.14) becomes
(4.16) ( ) ( ) ( )dPdPVdpdpvdPdpV hgh
gh
Sji
i
Sijijiji
ghgh
Sji
ii
g
g
g
−+−+−⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜
⎝
⎛
−= ∑∑∑∑∑ ≠∈
∈≠≠∈
66
θθ
θθ
φθ .
Substituting (4.16) into (4.14), we obtain
(4.17)
( ) ( )
( ) ihgh
gh
Sji
i
Sijijiji
ghgh
Sji
iiiii
dPdPV
dpdpvdPdpVdQdqw
g
g
g
εθ
θ
θθ
φθθ
+−
+−+−⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜
⎝
⎛
−+=
∑∑
∑∑∑
≠∈
∈≠≠∈
6
6
Equation (4.14) can thus be estimated using the following symmetry-constrained
system of equations (equation 4.18).
58
( ) ( )
( ) ( )
( ) ( )[ ])(...
...
616313212321
1
13131212
1161312321
11111
dPdPVdPdPVdPdPV
dpdpvdpdpv
dPdpVVVdQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+−+−
+−⎥⎥⎦
⎤
⎢⎢⎣
⎡+++⎟⎟
⎠
⎞⎜⎜⎝
⎛++
−+=
θθθθ
θθθθ
φθθ
( ) ( )
( ) ( )
( ) ( )[ ])(...
...
626323212321
2
23232112
2161312321
22222
dPdPVdPdPVdPdPV
dpdpVdpdpv
dPdpVVVdQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
−+−
+−⎥⎦
⎤⎢⎣
⎡+++⎟⎟
⎠
⎞⎜⎜⎝
⎛++
−+=
θθθθ
θθθθ
φθθ
( ) ( )
( ) ( )
( ) ( )[ ])(...
...
616313212321
3
32233113
3262312321
33333
dPdPVdPdPVdPdPV
dpdpVdpdpv
dPdpVVVdQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
−+−
+−⎥⎦
⎤⎢⎣
⎡+++⎟⎟
⎠
⎞⎜⎜⎝
⎛++
−+=
θθθθ
θθθθ
φθθ
( ) ( )
( ) ( )
( ) ( )[ ])(...
...
626323112654
4
46464545
4262312654
44444
dPdPVdPdPVdPdPV
dpdpVdpdpv
dPdpVVVdQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+−+−
+−⎥⎥⎦
⎤
⎢⎢⎣
⎡+++⎟⎟
⎠
⎞⎜⎜⎝
⎛++
−+=
θθθθ
θθθθ
φθθ
( ) ( )
( ) ( )
( ) ( )[ ])(...
...
626323112654
5
56565445
5262312654
55555
dPdPVdPdPVdPdPV
dpdpVdpdpv
dPdpVVVdQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+−+−
+−⎥⎥⎦
⎤
⎢⎢⎣
⎡+++⎟⎟
⎠
⎞⎜⎜⎝
⎛++
−+=
θθθθ
θθθθ
φθθ
( ) ( )
( ) ( )
( ) ( )[ ])(...
...
626323112654
6
65566446
6262312654
66666
dPdPVdPdPVdPdPV
dpdpVdpdpv
dPdpVVVdQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+−+−
+−⎥⎥⎦
⎤
⎢⎢⎣
⎡+++⎟⎟
⎠
⎞⎜⎜⎝
⎛++
−+=
θθθθ
θθθθ
φθθ
59
( ) ( )
( ) ( )
( ) ( )[ ])(...
...
636223113987
7
79797878
7362313987
77777
dPdPVdPdPVdPdPV
dpdpVdpdpv
dPdpVVVdQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+−+−
+−⎥⎥⎦
⎤
⎢⎢⎣
⎡+++⎟⎟
⎠
⎞⎜⎜⎝
⎛++
−+=
θθθθ
θθθθ
φθθ
( ) ( )
( ) ( )
( ) ( )[ ])(...
...
636223113987
8
89898778
8362313987
88888
dPdPVdPdPVdPdPV
dpdpVdpdpv
dPdpVVVdQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+−+−
+−⎥⎥⎦
⎤
⎢⎢⎣
⎡+++⎟⎟
⎠
⎞⎜⎜⎝
⎛++
−+=
θθθθ
θθθθ
φθθ
( ) ( )
( ) ( )
( ) ( )[ ])(...
...
636223113987
9
98899779
9362313987
99999
dPdPVdPdPVdPdPV
dpdpVdpdpv
dPdpVVVdQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+−+−
+−⎥⎥⎦
⎤
⎢⎢⎣
⎡+++⎟⎟
⎠
⎞⎜⎜⎝
⎛++
−+=
θθθθ
θθθθ
φθθ
( ) ( )
( ) ( )
( ) ( )[ ])(...
...
646224114121110
10
1012101210111011
10462414121110
1010101010
dPdPVdPdPVdPdPV
dpdpVdpdpv
dPdpVVVdQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+−+−
+−⎥⎥⎦
⎤
⎢⎢⎣
⎡+++⎟⎟
⎠
⎞⎜⎜⎝
⎛++
−+=
θθθθ
θθθθ
φθθ
( ) ( )
( ) ( )
( ) ( )[ ])(...
...
646224114121110
11
1112111211101011
11462414121110
1111111111
dPdPVdPdPVdPdPV
dpdpVdpdpv
dPdpVVVdQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+−+−
+−⎥⎥⎦
⎤
⎢⎢⎣
⎡+++⎟⎟
⎠
⎞⎜⎜⎝
⎛++
−+=
θθθθ
θθθθ
φθθ
( ) ( )
( ) ( )
( ) ( )[ ])(...
...
646224114121110
12
1211111212101012
11462414121110
1212121212
dPdPVdPdPVdPdPV
dpdpVdpdpv
dPdpVVVdQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+−+−
+−⎥⎥⎦
⎤
⎢⎢⎣
⎡+++⎟⎟
⎠
⎞⎜⎜⎝
⎛++
−+=
θθθθ
θθθθ
φθθ
60
( ) ( )
( ) ( )
( ) ( )[ ])(...
...
656225115151413
13
1315131513141314
13562515151413
1313131313
dPdPVdPdPVdPdPV
dpdpVdpdpv
dPdpVVVdQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+−+−
+−⎥⎥⎦
⎤
⎢⎢⎣
⎡+++⎟⎟
⎠
⎞⎜⎜⎝
⎛++
−+=
θθθθ
θθθθ
φθθ
( ) ( )
( ) ( )
( ) ( )[ ])(...
...
656225115151413
14
1415141514131314
14562515151413
1414141414
dPdPVdPdPVdPdPV
dpdpVdpdpv
dPdpVVVdQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
−+−
+−⎥⎥⎦
⎤
⎢⎢⎣
⎡+++⎟⎟
⎠
⎞⎜⎜⎝
⎛++
−+=
θθθθ
θθθθ
φθθ
( ) ( )
( ) ( )
( ) ( )[ ])(...
...
656225115151413
15
1514141515131315
15562515151413
1515151515
dPdPVdPdPVdPdPV
dpdpVdpdpv
dPdpVVVdQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+−+−
+−⎥⎥⎦
⎤
⎢⎢⎣
⎡+++⎟⎟
⎠
⎞⎜⎜⎝
⎛++
−+=
θθθθ
θθθθ
φθθ
( ) ( )
( ) ( )
( ) ( )[ ])(...
...
556226116181716
16
1618161816171617
16562616181716
1616161616
dPdPVdPdPVdPdPV
dpdpvdpdpv
dPdpVVVdQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+−+−
+−⎥⎦
⎤⎢⎣
⎡+++⎟⎟
⎠
⎞⎜⎜⎝
⎛++
−+=
θθθθ
θθθθ
φθθ
( ) ( )
( ) ( )
( ) ( )[ ])(...
...
556226116181716
17
1718171817161617
17562616181716
1717171717
dPdPVdPdPVdPdPV
dpdpvdpdpv
dPdpVVVdQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+−+−
+−⎥⎦
⎤⎢⎣
⎡+++⎟⎟
⎠
⎞⎜⎜⎝
⎛++
−+=
θθθθ
θθθθ
φθθ
Block Independent Uniform Substitute-Rotterdam Model
The block independent uniform substitute-Rotterdam model (3.21) can be written
in finite changes as
61
(4.20) ( )⎥⎥⎦
⎤
⎢⎢⎣
⎡
Θ−
−+
Θ−−
+= ∑∈≠ gSij
j
g
jii
g
iiiii dP
dpk
kdPdp
kk
dqdqw11
1 θθθθφθ
Equation (4.20) can be estimated using the following system of equations.
( )
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=)(
)(1
)()(1
)()(1
1
33211
311
23211
2111
3211
111
111
dPdpk
k
dPdpk
kdPdp
kk
dQdqw
θθθθθ
θθθθθ
θθθθθ
φθ
( )
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=)(
)(1
)()(1
1)(
)(1
33211
321
23211
2121
3211
211
222
dPdpk
k
dPdpk
kdPdp
kk
dQdqw
θθθθθ
θθθθθ
θθθθθ
φθ
( )⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=)(
)(11
)()(1
)()(1
33211
313
23211
3211
3211
311
333
dPdpk
k
dPdpk
kdPdp
kk
dQdqw
θθθθθ
θθθθθ
θθθθθ
φθ
( )
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=)(
)(1
)()(1
)()(1
1
66542
642
56542
5424
6542
424
444
dPdpk
k
dPdpk
kdPdp
kk
dQdqw
θθθθθ
θθθθθ
θθθθθ
φθ
( )
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=)(
)(1
)()(1
1)(
)(1
66542
652
56542
5254
6542
542
555
dPdpk
k
dPdpk
kdPdp
kk
dQdqw
θθθθθ
θθθθθ
θθθθθ
φθ
( )⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=)(
)(11
)()(1
)()(1
66542
626
56542
6524
6542
642
666
dPdpk
k
dPdpk
kdPdp
kk
dQdqw
θθθθθ
θθθθθ
θθθθθ
φθ
( )
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=)(
)(1
)()(1
)()(1
1
99873
973
89873
8737
9873
737
777
dPdpk
k
dPdpk
kdPdp
kk
dQdqw
θθθθθ
θθθθθ
θθθθθ
φθ
62
( )
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=)(
)(1
)()(1
1)(
)(1
99873
983
89873
8387
9873
873
888
dPdpk
k
dPdpk
kdPdp
kk
dQdqw
θθθθθ
θθθθθ
θθθθθ
φθ
( )⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=)(
)(11
)()(1
)()(1
99873
939
89873
9837
9873
973
999
dPdpk
k
dPdpk
kdPdp
kk
dQdqw
θθθθθ
θθθθθ
θθθθθ
φθ
( )
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=)(
)(1
)())(1
)()(1
1
121211104
12104
111211104
1110410
1211104
10410
101010
dPdpk
k
dPdpk
kdPdp
kk
dQdqw
θθθθθ
θθθθθ
θθθθθ
φθ
( )
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=)(
)(1
)()(1
1)()(1
121211104
12114
111211104
1141110
1211104
11104
111111
dPdpk
k
dPdpk
kdPdpk
k
dQdqw
θθθθθ
θθθθθ
θθθθθ
φθ
( )⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=)(
)(11
)()(1
)()(1
121211104
12412
111211104
1211410
1211104
12104
121212
dPdpk
k
dPdpk
kdPdpk
k
dQdqw
θθθθθ
θθθθθ
θθθθθ
φθ
( )
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=)(
)(1
)()(1
)()(1
1
151514135
15135
141514135
1413513
1514135
13513
131313
dPdpk
k
dPdpk
kdPdp
kk
dQdqw
θθθθθ
θθθθθ
θθθθθ
φθ
( )
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=)(
)(1
)()(1
1)(
)(1
151514135
15145
141514135
1451413
1514135
14135
141414
dPdpk
k
dPdpk
kdPdp
kk
dQdqw
θθθθθ
θθθθθ
θθθθθ
φθ
63
( )⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=)(
)(11
)()(1
)()(1
151514135
15515
141514135
1514513
1514135
15135
151515
dPdpk
k
dPdpk
kdPdp
kk
dQdqw
θθθθθ
θθθθθ
θθθθθ
φθ
( )
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=)(
)(1
)()(1
)()(1
1
181817166
18166
171817166
1716616
1817166
16616
161616
dPdpk
k
dPdpk
kdPdp
kk
dQdqw
θθθθθ
θθθθθ
θθθθθ
φθ
( )
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=)(
)(1
)()(1
1)(
)(1
181817166
18176
171817166
1761716
1817166
17166
171717
dPdpk
k
dPdpk
kdPdp
kk
dQdqw
θθθθθ
θθθθθ
θθθθθ
φθ
Block-wise Dependent Uniform Substitute-Rotterdam Model
The block-wise dependent uniform substitute-Rotterdam model (3.24) can be
written in finite changes as
(4.22) dPdP
VdPdp
kk
dPdp
kk
dQdqw h
ghghi
Sij
j
g
jiii
gi
iiiiii
g
∑∑≠∈≠
+Θ−
−+⎟
⎟⎠
⎞⎜⎜⎝
⎛
Θ−−
+= '
111
θθθ
φθ
φθθ .
Equation (4.22) can be estimated using the following system of equations.
( ) ( ) ( )
( )
( ) ( )[ ])(...
)(1
)(1)(11
616313212321
1
33211
311
23211
2111
3211
111
111
dPdPVdPdPVdPdPV
DPdpk
k
dPdpk
kdPdp
kk
dQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=
θθθθ
θθθθθ
θθθθθ
θθθθθ
φθ
( ) ( ) ( )
( )
( ) ( )[ ])(...
)(1
)(11
)(1
616313212321
2
33211
321
23211
2121
3211
211
222
dPdPVdPdPVdPdPV
DPdpk
k
dPdpk
kdPdp
kk
dQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=
θθθθ
θθθθθ
θθθθθ
θθθθθ
φθ
64
( ) ( )
( ) ( )
( ) ( )[ ])(...
)(11
)(1)(1
616313212321
3
33211
313
23211
3211
3211
311
333
dPdPVdPdPVdPdPV
DPdpk
k
dPdpk
kdPdp
kk
dQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=
θθθθ
θθθθθ
θθθθθ
θθθθθ
φθ
( ) ( ) ( )
( )
( ) ( )[ ])(...
)(1
)(1)(11
626323112654
4
66542
642
56542
5424
6542
424
444
dPdPVdPdPVdPdPV
DPdpk
k
dPdpk
kdPdp
kk
dQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=
θθθθ
θθθθθ
θθθθθ
θθθθθ
φθ
( ) ( ) ( )
( )
( ) ( )[ ])(...
)(1
)(11
)(1
626323112654
5
66542
652
56542
5254
6542
542
555
dPdPVdPdPVdPdPV
DPdpk
k
dPdpk
kdPdp
kk
dQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=
θθθθ
θθθθθ
θθθθθ
θθθθθ
φθ
( ) ( )
( ) ( )
( ) ( )[ ])(...
)(11
)(1)(1
626323112654
6
66542
626
56542
6524
6542
642
666
dPdPVdPdPVdPdPV
DPdpk
k
dPdpk
kdPdp
kk
dQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=
θθθθ
θθθθθ
θθθθθ
θθθθθ
φθ
( ) ( ) ( )
( )
( ) ( )[ ])(...
)(1
)(1)(11
636223113987
7
99873
973
89873
8737
9873
737
777
dPdPVdPdPVdPdPV
DPdpk
k
dPdpk
kdPdp
kk
dQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=
θθθθ
θθθθθ
θθθθθ
θθθθθ
φθ
65
( ) ( ) ( )
( )
( ) ( )[ ])(...
)(1
)(11
)(1
636223113987
8
99873
983
89873
5357
9873
873
888
dPdPVdPdPVdPdPV
DPdpk
k
dPdpk
kdPdp
kk
dQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=
θθθθ
θθθθθ
θθθθθ
θθθθθ
φθ
( ) ( )
( ) ( )
( ) ( )[ ])(...
)(11
)(1)(1
636223113987
9
99873
939
89873
9827
9873
973
999
dPdPVdPdPVdPdPV
DPdpk
k
dPdpk
kdPdp
kk
dQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=
θθθθ
θθθθθ
θθθθθ
θθθθθ
φθ
( ) ( ) ( )
( )
( ) ( )[ ])(...
)(1
)(1)(11
646224114121110
10
121211104
12104
111211104
1110410
1211104
10410
101010
dPdPVdPdPVdPdPV
DPdpk
k
dPdpk
kdPdp
kk
dQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=
θθθθ
θθθθθ
θθθθθ
θθθθθ
φθ
( ) ( ) ( )
( )
( ) ( )[ ])(...
)(1
)(11
)(1
646224114121110
11
121211104
12114
111211104
1141110
1211104
11104
111111
dPdPVdPdPVdPdPV
DPdpk
k
dPdpk
kdPdp
kk
dQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=
θθθθ
θθθθθ
θθθθθ
θθθθθ
φθ
( ) ( )
( ) ( )
( ) ( )[ ])(...
)(11
)(1)(1
646224114121110
12
121211104
12412
111211104
1211410
1211104
12104
121212
dPdPVdPdPVdPdPV
DPdpk
k
dPdpk
kdPdpk
k
dQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=
θθθθ
θθθθθ
θθθθθ
θθθθθ
φθ
66
( ) ( ) ( )
( )
( ) ( )[ ])(...
)(1
)(1)(11
656225115151413
13
151514135
15135
141514135
1413413
1514135
13513
131313
dPdPVdPdPVdPdPV
DPdpk
k
dPdpk
kdPdp
kk
dQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=
θθθθ
θθθθθ
θθθθθ
θθθθθ
φθ
( ) ( ) ( )
( )
( ) ( )[ ])6(...
)(1
)(11
)(1
56225115151413
14
151514135
15145
141514135
1451413
1514135
14135
141414
dPdPVdPdPVdPdPV
DPdpk
k
dPdpk
kdPdp
kk
dQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=
θθθθ
θθθθθ
θθθθθ
θθθθθ
φθ
( ) ( )
( ) ( )
( ) ( )[ ])(...
)(11
)(1)(1
656225115151413
15
151514135
15515
141514135
1514513
1514135
15135
151515
dPdPVdPdPVdPdPV
DPdpk
k
dPdpk
kdPdp
kk
dQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=
θθθθ
θθθθθ
θθθθθ
θθθθθ
φθ
( ) ( ) ( )
( )
( ) ( )[ ])(...
)(1
)(1)(11
556226116181716
16
181817166
18166
171817166
1716616
1817166
16616
161616
dPdPVdPdPVdPdPV
DPdpk
k
dPdpk
kdPdp
kk
dQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=
θθθθ
θθθθθ
θθθθθ
θθθθθ
φθ
( ) ( ) ( )
( )
( ) ( )[ ])(...
)(1
)(11
)(1
556326216181716
17
181817166
18176
171817166
1761716
1817166
17166
171717
dPdPVdPdPVdPdPV
DPdpk
k
dPdpk
kdPdp
kk
dQdqw
−++−+−⎟⎟⎠
⎞⎜⎜⎝
⎛++
+
⎥⎥⎥⎥
⎦
⎤
⎢⎢⎢⎢
⎣
⎡
−++−
−
+−++−
−+−
++−−
+=
θθθθ
θθθθθ
θθθθθ
θθθθθ
φθ
Data Sources
The sources of data for this study are the Statistics Bureau of Japan and Japan’s
Ministry of Finance as well as the Food and Agricultural Organization. Monthly
population data from December 1995 to May 2005 came from the web page
67
(http://www.stat.go.jp/english/data/jinsui/2-2.htm) maintained by the Statistics Bureau of
Japan’s Ministry of Internal Affairs and Communications. Import data came from the
Trade Statistics of Japan that are published by the Ministry of Finance and the Customs
under the provision of the Customs Law and the relevant international conventions. It is
available on the web page http://www.customs.go.jp. The monthly imports and
expenditures on imports of orange, grapefruit, other citrus, apple, pineapple and grape
juices were obtained for the period December, 1995 to May, 2005. The values of imports
are on a cost, insurance and freight (CIF) basis, which include costs of the product,
insurance and transportation. Unit import values, which proxy commodity prices, were
obtained by dividing import values by import quantities. Data on the production,
consumption and trade of fruit juices came from the webpage
http://faostat.fao.org/faostat/ maintained by the Food and Agricultural Organization.
Analytical Methods
The method used to estimate the model is the non-linear least square (LSQ) in the
Time Series Processor Program (TSP 4.5). This method is based on the entire system of
equations, and estimates all parameters jointly. When estimating the system of demand
equations, one of the equations has to be deleted or the covariance matrix will be singular.
However, parameter estimates of the deleted equation can be recovered by re-estimating
the system with another equation in the system. Parameter estimates are invariant to the
deleted equation when using maximum likelihood estimation (Barten, 1969).
The LSQ command computes maximum likelihood estimates if it is specified with
no instruments and more than one equation (Cummins, 1999). Since the parameter
estimates in this study are generated from a system of demand equations without
specifying instruments, they can be taken as maximum likelihood estimates. With
68
normally distributed disturbances ( itu ), the ML method has all the desirable asymptotical
properties of Maximum Likelihood (ML) estimators and, therefore, is asymptotically
efficient among all estimators (Greene, 2000). The likelihood ratio test is used to test for
autocorrelation.
69
CHAPTER 5 RESULTS AND DISCUSSION
Descriptive Results
Since Japan’s deregulation of imports in the 1990s, the imports of fruit juices
have increased with the exception of U.S. apple juice (Table 5.1). Over the period
January, 1995 to May, 2005, the imports of U.S. apple juice has decreased by 17% while
that of U.S. orange, grapefruit and grape juices increased by 4%, 12% and 5%,
respectively. The highest increase was attained by the ROW grapefruit juice (51%)
followed by the Chinese apple juice (31%) and the Israelis grapefruit fruit juice (26%).
The analysis of import stability as measured by the coefficient of variation shows that the
imports of fruit juices in Japan over the given period have exhibited a significant
fluctuation. The fluctuation of imports varies from country to country. U.S. orange and
grape juices have experienced the highest fluctuation among U.S. fruit juices.
Over the same period, Japan’s import price of all fruit juices has decreased (Table
5.1). On average, Japan’s import price of U.S. orange, grapefruit, apple and grape juices
has decreased by 12%, 10%, 7% and 6% per month over the period December, 1995 to
May, 2005. Over the same period, apple juice imported from the rest of the world has
witnessed the largest price decrease (13%). Among U.S. products, prices of orange and
grapefruit juices are relatively more stable than those of the respective competitors’
products. The prices of apples are less stable compared to their respective rival products.
Except for Brazilian orange juice (25%) and the ROW apple juice (19%), the
average expenditure share of fruit juices in Japan is below 10% (Table 5.1). Expenditure
70
share of U.S. juices, expressed as a percentage of total fruit juice expenditure, ranges
from 6% for apple juice to 8% for grapefruit juice.
Table 5.1 Fruit juice quantity and price log-changes, and expenditure shares, Japan, December 1995 to May 2005
Quantity log-changes ( )1,log −= tiiti qqdq
Price log-changes ( )1,log −= tiiti ppdp
Expenditure shares ( )iw
Imports
Mean SD Mean SD Mean SD U.S. oranges 0.0410 0.6701 -0.1155 0.2803 0.0724 0.0335 Brazilian. oranges 0.0982 0.9847 -0.1033 0.2683 0.2542 0.0895 ROW oranges 0.0959 0.8876 -0.0083 0.4210 0.0324 0.0205 U.S. grapefruits 0.1200 0.4909 -0.0979 0.2907 0.0808 0.0302 Israelis grapefruits 0.2617 1.0503 -0.0720 0.5821 0.0259 0.0168 ROW grapefruits 0.5078 1.3739 -0.1149 0.8360 0.0111 0.0104 U.S. apples -0.1694 0.9249 -0.0690 0.2847 0.0567 0.0422 Chinese apples 0.3176 0.6891 -0.1405 0.2798 0.0727 0.0372 ROW apples 0.0760 0.4059 -0.0946 0.1958 0.1652 0.0510 Thai pineapples 0.1549 1.0317 -0.0572 0.3934 0.0109 0.0058 Philippines pineapples 0.1578 1.7814 -0.0606 0.3713 0.0075 0.0037 ROW pineapples 0.1109 1.5452 -0.0414 0.5171 0.0089 0.0062 U.S. grapes 0.0529 0.5942 -0.0647 0.2890 0.0621 0.0249 Argentinean grapes 0.2792 1.1260 -0.0969 0.3346 0.0091 0.0058 ROW grapes 0.1717 0.4728 -0.0802 0.2584 0.0648 0.0235 Israelis other citrus 0.0861 0.6349 -0.0924 0.3138 0.0220 0.0064 Italian other citrus 0.1756 0.7744 -0.0902 0.2412 0.0172 0.0069 ROW other citrus 0.2032 0.8238 -0.1031 0.5923 0.0250 0.0118 (Source: Study data)
Test for First-order Autocorrelation
A test for first order autocorrelation AR (1) was carried out for five different
versions of the Rotterdam model. These are block independent non-uniform substitute
Rotterdam model (4.9), block-wise dependent non-uniform substitute Rotterdam model
(4.14), block independent uniform substitute Rotterdam model (4.20) and block-wise
dependent uniform substitute Rotterdam model (4.22). The test was done considering
each model with and without autocorrelation as the unrestricted and restricted model,
respectively. The null hypothesis ( )0:0 =ρH was tested using the likelihood ratio test
given as ( ) ( )( ) [ ]JLLLR d 2ˆ~2 χθθ ⎯→⎯−−= where θ~ is a vector of restricted parameter
71
estimates and θ̂ is a vector of parameter estimates associated with the unrestricted model.
The restricted model is the one with first order serial correlation while the unrestricted
model is the one without first order autocorrelation.
Under the null hypothesis ( )0H , the LR has an asymptotic chi-square distribution
with the degrees of freedom equal to the number of restrictions J . Since symmetry was
imposed as part of the estimation procedure, the coefficient of autocorrelation was taken
to be common across equations.
The result of the test indicates that the null hypothesis of no autocorrelation was
rejected in all of the models (Table 5.2), implying that the data is serially correlated. The
value of ρ , which is common across equations in each system, ranges from 0.31 for (4.1)
to 0.36 for (4.9), and is significantly different from zero (P<0.001). The Hildreth and Lu
(1960) approach was used for the correction.
Table 5.2: Test for first-order autocorrelation Model Coefficient Log Likelihood
value ( ) ( )( )θθ ~ˆ2 LL − a
Rho = 0.00 4710.26 Equation (4.20) Rho = 0.35 4785.67
150.82***
Rho = 0.00 4716.24 Equation (4.9) Rho = 0.36 4789.90
147.32***
Rho = 0.00 4748.99 Equation (4.22) Rho = 0.33 4813.17
128.36***
Rho = 0.00 4757.74 Equation (4.14) Rho = 0.35 4826.09
136.70***
Rho = 0.00 4892.99 Equation (4.1) Rho = 0.31 4934.79
83.60***
a Twice the difference between the log likelihood value for the unconstrained model, ( )θ̂L , and the
log likelihood value for the constrained model, ( )θ~L . *** The chi-square critical value is at the 1% significance level.
72
Hypothesis Testing for Model Selection
Following the correction for first-order autocorrelation, the study tests two
hypotheses (block independence/uniform substitute hypothesis and block-wise
dependence/uniform substitute hypothesis) to select the model that best describes the
import data of fruit juices in Japan. The hypotheses of block independence and block-
wise dependence have to do with the relationship between products that belong to two
different product groups while that of the uniform substitute has to do with a relationship
between products that belong to the same product group. Therefore, the block-
independence/uniform substitute hypotheses and block-wise dependence/uniform
substitute hypothesis involve between-group (block independence or block-wise
dependence) and within-group (uniform substitute) relationships. Recall that the uniform
substitute hypothesis is applied to the same product differentiated by country of
production.
The result of these tests enables us to select the model that best describes the
import data of fruit juices in Japan. In light of these hypotheses, two restricted models
were derived from the relative price version of the Rotterdam model (4.1). The restricted
models are block independent uniform substitute-Rotterdam model (4.20), and block-
wise dependent uniform substitute-Rotterdam model (4.22). Since these two restricted
versions (4.20) and (4.22) are nested in the unrestricted version (4.1), the likelihood ratio
test is used.
Block Independence and Uniform Substitute Hypothesis
The hypothesis of block independence states that there is no specific cross price
effect ( )ijν between any two products in any two different product groups. The uniform
73
substitute hypothesis states that the specific cross price effect ( )ijν between any two
products in the same product group is the same for all pairs of goods in that group.
Combining the two null hypotheses, the null hypothesis of block independence
and uniform substitute relationship can be restated as
0:0 =ijH ν , gSi∈ , hSj∈ and hg ≠ ; kis =ν for any gSsi ∈, .
0: ≠ijAH ν , gSi∈ , hSj∈ and hg ≠ ; kis ≠ν for any gSsi ∈, .
The test for the hypothesis of block independence and uniform substitution
involves a comparison between the uniform substitute block independent Rotterdam
model (4.20) and the relative price version of the Rotterdam model (4.1).
Since (4.20) is a restricted function, we expect its likelihood value to be smaller
than that of (4.1). The likelihood value of (4.20) is 4785.67 with 24 degrees of freedom
while the value of (4.1) is 4934.79 with 171 degrees of freedom (Table 5.3). The value
of the model chi-square is 298.24 which is greater than the critical chi-square value at 1%
significance level. Therefore, we reject the null hypothesis, and conclude that there is
competition between products that belong to different product groups since there is a
change in marginal utility of a dollar spent on a product in one product group caused by
an extra dollar spent on another product in another product group. For example, an extra
dollar spent on U.S. orange juice i , affects the marginal utility of another dollar spent on
the Chinese apple juice j .
Furthermore, the change in marginal utility of a dollar spent on a product caused
by an extra dollar spent on another product is not the same for all pairs of products within
the same group. The country of origin makes a difference in one’s decision to buy a
certain fruit juice. For example, the change in marginal utility of a dollar spent on the
74
U.S. orange juice i , caused by an extra dollar spent on Brazilian orange juice r is not the
same as that of the change in marginal utility of a dollar spent on the U.S. orange juice
caused by an extra dollar spent on the ROW orange juice, s . This suggests that
consumers decide to buy orange juice based on the country of origin. In summary,
consumers are influenced by the country of origin when they choose between products
that belong to the same group.
Table 5.3 Hypothesis testing for model selection Model Log likelihood
value Free parameters ( ) ( )( )θθ ~ˆ2 LL − a
Equation (4.20) 4785.67 24 298.24***
Equation (4.22) 4813.17 39 243.24***
Equation (4.1) 4934.79 171
a Twice the difference between the log likelihood value for the unconstrained model, ( )θ̂L , and the
log likelihood value for the constrained model, ( )θ~L . *** The chi-square critical value is at the 1% significance level.
Block-wise Dependence and Uniform Substitute Hypothesis
The hypothesis of block-wise dependence states that the specific cross price effect
( )ijν between any two products in two different product groups is the same for all pairs of
products in the two groups. The uniform substitute hypothesis states that the specific
cross price effect ( )ijν between any two products in the same product is the same for all
pairs of products within that group.
Combining the two null hypotheses, the new null hypothesis which corresponds to
the block-wise dependence uniform substitute relationship can be restated as
ghij avH =:0 , gSi∈ , hSj∈ and hg ≠ ; and kvir = for any gSri ∈, .
75
ghijA avH ≠: , gSi∈ , hSj∈ and hg SS ≠ ; and kvir ≠ for any gSri ∈, .
The test for the hypothesis of block-wise dependence and uniform substitution
involves a comparison between the block-wise uniform substitute-Rotterdam model
(4.22) and the relative price version of the Rotterdam model (4.1). The likelihood value
of (4.22) is 4813.17 with 39 degrees of freedom while the value of (4.1) is 4934.79 with
171 degrees of freedom (Table 5.3). The value of the model chi-square is 243.24 which
is greater than the critical chi-square value at 1% probability level. Therefore, we reject
the null hypothesis, and conclude that the competition between products in different
groups is not the same for all pairs of products in the two groups the change in marginal
utility of a dollar spent on a product in one product group caused by an extra dollar spent
on another product in another product group is not the same for all pairs of products in
the two groups. In other words, the competition between product i of group g and
product j of group h is not the same as that of product i and product l of group h since
the change in marginal utility of product i , caused by an extra dollar spent on product j
is not the same as that of the change in marginal utility of a dollar spent on product i
caused by an extra dollar spent on product l . For example, an extra dollar spent on U.S.
orange juice j , affects the marginal utility of another dollar spent on Thai pineapple
juice i differently than does it affect the marginal utility of a dollar spent on the
Philippines pineapple juice. In other words, the effect of the change in price of U.S.
orange juice on the demand for Thai pineapple juice is not the same as that of the effect
on the demand for the Philippines pineapple juice. This implies that the country of origin
of the pineapple juice makes a difference when a consumer decides to buy orange and
pineapple juices.
76
Furthermore, we can conclude that the change in marginal utility of a dollar spent
on a product caused by an extra dollar spent on another product is not the same for all
pairs of products within the same group. The country of origin makes a difference in
one’s decision to choose between products that belong to the same product group. For
example, the change in marginal utility of a dollar spent on U.S. orange juice i , caused
by an extra dollar spent on Brazilian orange juice r is not the same as that of the change
in marginal utility of a dollar spent on U.S. orange juice i caused by an extra dollar spent
on the ROW orange juice, s . This implies that consumers are influenced by the country
of origin and thus decide to buy orange juice based on the country of origin. In summary,
the country of origin is taken into account by consumers when they choose between
products that belong to different product groups and also when they choose between
products that belong to the same product group.
Therefore, based on the results of the likelihood ratio test which rejected both
restricted models (Table 5.3) the relative price version of the Rotterdam model (4.1) is
chosen to best describe the import data of Japan’s import of fruit juices.
The relative Price Version of the Rotterdam Model
Since the relative price version of the Rotterdam model does not have any
restriction within or across the price coefficients, individual products are competing with
each other based on the country of origin. In other words, it allows investigating the
relationship between individual products based on the country of origin of the product.
For example, we can investigate the relationship between apple juice from China and
orange juice from U.S.
77
Parameter Estimates
Marginal expenditure shares Results indicate that the marginal expenditure
shares are all positive except for those of the Israel’s grapefruit juice and the ROW
pineapple juice (Table 5.4). However, the coefficient of the Israelis grapefruit juice is not
statistically significant. The largest share of the increase in marginal expenditure on
imported fruit juices goes to Brazilian orange juice (70%) followed by that of the ROW
apple juice (8%). This is consistent with the average expenditure shares since the
Brazilian orange juice (25%) and the ROW apple juice (17%) have the first and second
highest average expenditure shares (Table 5.1). Japanese imports of U.S. orange juice
grapefruit and apple juice receive only 3-4% of the increase in marginal expenditures.
Except for apple juice and grape juice, imports of fruit juices from ROW receive less than
one percent of each dollar increase in expenditures.
Table 5.4 Marginal expenditure shares of imported fruit juices in Japan Product Estimates SE U0.S0. oranges 0.0337*** 0.0100 Brazilian oranges 0.6997*** 0.0373 ROW oranges 0.0033 0.0058 U0.S0. grapefruits 0.0441*** 0.0078 Israelis grapefruits -0.0016 0.0059 ROW grapefruits 0.0051 0.0034 U0.S0. apples 0.04686*** 0.0124 Chinese apples 0.0473** 0.0092 ROW apples 0.0800*** 0.0159 Thai pineapples 0.0044* 0.0023 Philippine p0. apples 0.0024 0.0022 ROW pineapples -0.0073** 0.0036 U0.S0. grapes 0.0080 0.0076 Argentinean grapes 0.0017 0.0024 ROW grapes 0.0188*** 0.0066 Israelis other citrus 0.0045 0.0032 Italian other citrus 0.0019 0.0027 ROW other citrus 0.0064 0.0041 *** (**)* significance at 1%, 5% and 10%
78
Price effects. Price effects are described here by both relative and Slutsky price
coefficients. The Slutsky price coefficients ijπ can be derived from relative (Frisch-
deflated) price coefficients ijv and marginal value shares iθ using jiijij v θφθπ −= ,
where φ is the coefficient of income flexibility.
The Slutsky price coefficients ( )ijπ are the sum of the specific ijv and general
substitution effects ( )jiθφθ− . The Slutsky price coefficients ijπ measure the total
substitution effect of a change in the thj price on the demand for the thi product or,
equivalently, the effect of such a change when real income remains constant. The
relative price coefficients measure the specific substitution effect which accounts for the
n price changes on the demand for the thi product, or equivalently, the effect of such a
change when the marginal utility of income remains constant. The general substitution
effect ( )jiθφθ− , which serves as a deflator of the specific substitution effect by
transforming the absolute prices into relative prices, accounts for the Frisch price index
changes on the demand for the thi product.
If the relative price coefficients ijv and jiv are both positive, it means that an
increase in the relative price of either product raises the demand for the other, and thus
the two products are called specific substitutes. Similarly, if ijv and jiv are both negative,
it means that an increase in the relative price (opportunity cost) of either product reduces
the demand for the other, or thus the two products are called specific complements.
The Hicks’s definitions of net substitutes and net complements are based on the
signs of ijπ . The sign of the parameter ijπ determines if products i and j are net
79
complements 0<ijπ or net substitutes 0>ijπ . In terms of the Slutsky
equation, ( ) ( )mqqpqpq jisijij ∂∂−∂∂=∂∂ , if the substitution term, ( ) 0>∂∂sij pq for
net (or Hicksian) substitutes, and ( ) 0<∂∂sij pq for net (or Hicksian) complements.
While the Slutsky price coefficients provide the net substitution effects when real
income remains constant, the relative price coefficients provide the same effects when the
marginal utility of income remains unchanged. The concepts of net substitutes and
complements focus solely on the substitution effects.
The statistically significant relative and Slutsky cross price coefficients are
presented in Table 5.5. Most of these products are substitutes. The difference between
the coefficients of the relative and absolute price coefficients in terms of magnitude is
small. This implies that the general substitution effect is small. The general substitution
effect is concerned with the competition of all products for an extra dollar of the
consumer’s income.
Contrary to expectation, the cross price effects of products that belong to the same
group are not necessarily greater than the cross effects of products that belong to different
product groups. For example, the cross price effect of U.S. grapefruit/ROW grapefruit
juice is smaller than that of U.S. grapefruit /U.S apple juice. Furthermore, products that
belong to the same product group are not necessarily substitutes. For example, U.S.
apple/ROW apple that belong to the same product group are complements.
Based on the cross price effects of substitute products, we can identify the market
structure of fruit juice in Japan (Figure 5.6), showing that there are both direct and
indirect competitions based on the country of origin. Recall that the direct competition
refers to the competition between products within the same juice group (e.g., orange juice
80
group) while the indirect competition refers to the competition between products in
different juice groups (e.g., orange juice and apple juice). Except for grape juice, there is
no direct competition in Japan’s fruit juice market. The indirect competition appears to
be more important than the direct competition in Japan’s fruit juice market.
Table 5.5: Parameter estimates of cross prices of fruit juices in Japan Relative price coefficients Slutsky coefficients Products
Estimates SE Estimates SEU.S. orange/Brazilian orange 0.0395 0.0296 0.0822*** 0.0259U.S. orange/ROW grapefruit -0.0089** 0.0040 -0.0086** 0.0040U.S. orange/U.S. apple 0.0309** 0.0126 0.0338*** 0.0128U.S. orange/Philippines pineapple 0.0087** 0.0044 0.0088** 0.0044U.S. orange/Israelis citrus -0.0158** 0.0062 -0.0155** 0.0062U.S. orange/ROW citrus -0.0107** 0.0052 -0.0103** 0.0052Brazilian. orange/Chinese apple -0.0701*** 0.0236 -0.0101 0.0207Brazilian orange/ROW apple -0.1769*** 0.0402 -0.0754** 0.0354Brazilian orange/ROW p. apple 0.0304*** 0.0099 0.0211** 0.0085Brazilian orange/Israelis citrus 0.0076 0.9335 0.0134* 0.0080ROW orange/U.S apple 0.0129** 0.6936 0.0132* 0.0069ROW orange/ROW apple 0.0216** 0.0103 0.0221** 0.0102ROW orange/Argentinean grape 0.0036* 0.0022 0.0036* 0.0022U.S. grapefruit/ROW grapefruit 0.0102*** 0.0030 0.0106*** 0.0030U.S. grapefruit/U.S. apple 0.0230** 0.0096 0.0267*** 0.0096U.S. grapefruit/Thai. pineapple -0.0188*** 0.0035 -0.0185*** 0.0035U.S. grapefruit/Philippines pineapple -0.0146*** 0.0034 -0.0144*** 0.0034U.S. grapefruit/U.S. grape -0.0161* 0.0091 -0.0155* 0.0091U.S. grapefruit/ROW grape 0.0194** 0.0095 0.0209** 0.0095Israelis grapefruit/Italian citrus 0.00467** 0.0020 0.0046** 0.0020ROW grapefruit/Italian citrus -0.0043*** 0.0013 -0.0043*** 0.0013U.S. apple/ROW apple -0.0445** 0.0178 -0.0377** 0.0176U.S. apple/Philippines pineapple 0.00649** 0.0031 0.0066** 0.0031U.S. apple/ROW pineapple -0.0143*** 0.0046 -0.0149*** 0.0046U.S. apple/Argentinean grape -0.0177*** 0.0035 -0.0176*** 0.0035U.S. apple/Israelis citrus -0.0127*** 0.0045 -0.0123*** 0.0045U.S. apple/ROW citrus 0.0097** 0.0047 0.0103** 0.0047Chinese apple/ROW pineapple -0.0066 0.0043 -0.0072* 0.0042Chinese apple/U.S. grape 0.0211** 0.0087 0.0218** 0.0087ROW apple/Israelis citrus 0.0127* 0.0080 0.0133* 0.0079Philipp. pineapple/Argentinean grape 0.0050*** 0.0019 0.0050*** 0.0019Philipp. pineapple/ROW grape 0.0069* 0.0039 0.0070* 0.0039Philipp. pineapple/Israelis citrus 0.00477** 0.0021 0.0047** 0.0021ROW pineapple/Argentinean grape 0.0055*** 0.0017 0.0054*** 0.0017U.S. grape/Argentinean grape 0.0108** 0.0043 0.0108** 0.0043U.S. grape/ROW citrus 0.0085** 0.0039 0.0086** 0.0039*** (**)* significance at 1%, 5% and 10%
81
Table 5.6 Market structure of fruit juices in Japan Orange Grapefruit Apple Pineapples Grapes Other citrus Product Country
U.S. Brazil U.S. Israel U.S. China Thailand
Philippine
U.S. Argentina
Israel Italy
U.S.
SS SS CC
Orange Brazil
CC
U.S. SS CC CC CC
Grapefruit Israel SS
U.S. SS SS SS CC CC
Apple China
CC SS
Thailand
CC
Pineapple Philippines
SS CC SS SS SS
U.S.
CC SS S
grape Argentina
CC SS SS S
Israel CC CC CC SS Other citrus Italy SS
82
Results also indicate that the relative price coefficients and the Slutsky own price
coefficients are all negative and significantly different from zero except for the ROW
apple juice (Table 5.7). Contrary to expectation, the own price coefficient of the ROW
apple juice is positive but not statistically significant. The negative signs are consistent
with demand theory since they ensure the negativity of the own substitution effect.
Table 5.7 Parameter estimates of own prices of fruit juices in Japan Relative price coefficients ( )iiv
Slutsky price coefficients ( )iiπ
Juice
Estimate SE Estimate SE
U.S. oranges -0.1139*** 0.0222 -0.1118*** 0.0224 Brazilian oranges -0.9667*** 0.1275 -0.0791 0.0861 ROW oranges -0.0469*** 0.0053 -0.0469*** 0.0053 U.S. grapefruits -0.0471*** 0.0131 -0.0436*** 0.0131 Israelis grapefruits -0.0142*** 0.0045 -0.0142*** 0.0045 ROW grapefruits -0.0079*** 0.0015 -0.0078*** 0.0016 U.S. apples -0.0308* 0.0177 -0.0268 0.0180 Chinese apples -0.0439*** 0.0135 -0.0398*** 0.0134 ROW apples 0.0116 0.0383 0.0233 0.0371 Thai pineapples -0.0095*** 0.0021 -0.0095*** 0.0021 Philippine pineapples -0.0231*** 0.0020 -0.0231*** 0.0020 ROW pineapples -0.0057** 0.0027 -0.0057** 0.0027 U.S. grapes -0.0523*** 0.0126 -0.0522*** 0.0126 Argentinean grapes -0.0059 0.0038 -0.0058 0.0038 ROW grapes -0.0409*** 0.0149 -0.0402*** 0.0149 Israelis other citrus -0.0221*** 0.0042 -0.0220*** 0.0042 Italian other citrus -0.0202*** 0.0044 -0.0202*** 0.0044 ROW other citrus -0.0239*** 0.0026 -0.0238*** 0.0026
*** (**)* significance at 1%, 5% and 10% Expenditure Elasticities
The value of income flexibility is estimated to be φ =-1.8126. The reciprocal of
this coefficient, which is the value of the income elasticity of the marginal utility of
income is φ1 = -0.5517. This estimate is consistent with the estimates of Frisch (1959)
for the richest section of the population. According to Frisch (1959), a value of φ1 =-0.7
83
is for the better off part of the population. Since Japanese consumers are among the
richest in the world, a value of φ1 = -0.5517 obtained in this study is a reasonable
estimate for Japan.
The expenditure elasticities are calculated at the sample means of expenditure
shares of the respective imported fruit juices using the equation ii wθη = where iθ is
the marginal value share of product i and iw is the average value share of the same
product. Expenditure elasticities of imported products are useful to provide guidance for
marketing strategies and policy making in exporting countries.
The estimates of the expenditure elasticities are positive except for those of the
Israelis grapefruit juice and the ROW pineapple juice (Table 5.8). However, the
expenditure elasticity of Israel’s grapefruit juice is statistically insignificant while that of
the ROW pineapple juice is statistically significant. Thus, we can conclude that the
Israelis grapefruit juice is not an inferior product while that of the ROW pineapple juice
is an inferior product.
Among the 18 fruit juices, only the demand for Brazilian orange juice is
expenditure elastic (2.7522). All four major fruit juices (orange, grapefruit, apple and
grape juices) that the U.S. exports to Japan are expenditure inelastic, implying that there
is less preference for the U.S. juices. The expenditure elasticities of U.S. exports range
from 0.1302 for grape juice to 0.8252 for apple juice. The demand for these products
exported by the rest of the world is also expenditure inelastic.
The high expenditure elasticity of Brazilian orange juice and low expenditure
elasticities of U.S. and the ROW products is not surprising given that Brazil’s share of
the total import expenditure is very high compared to that of other countries. The
84
average expenditure share of Brazilian orange juice is 25% while that of U.S. ranges from
5% for apple juice to 8% for grapefruit juice (Table 5.1). The average expenditure share
of fruit juices imported from the ROW is the smallest except for that of apple juice,
which accounts for about 17% of the total import expenditure on imported fruit juices.
The major exporting country of apple juice in the category of the ROW is Austria.
Table 5.8 Expenditure elasticity estimates of fruit juices in Japan Product Estimate SE
USA orange 0.4654*** 0.1390 Brazil orange 2.7525*** 0.1467 ROW orange 0.1047 0.1789 USA grapefruit 0.5463*** 0.0967 Israel grapefruit -0.0630 0.2300 ROW grapefruit 0.4603 0.3115 USA apple 0.8252*** 0.2189 Chinese apple 0.6504*** 0.1267 ROW apple 0.4842*** 0.0963 Thailand pineapple 0.4048* 0.2158 Philippines pineapple 0.3212 0.2954 ROW pineapple -0.8262** 0.4060 USA grape 0.1301 0.1226 Argentina grape 0.1921 0.2670 ROW grape 0.2912*** 0.1031 Israel other citrus 0.2065 0.1491 Italy other citrus 0.1153 0.1582 ROW other citrus 0.2578 0.1649
*** (**)* significance at 1%, 5% and 10%
The high expenditure elasticity may imply that there is a strong preference for
Brazilian orange juice, and that it is a luxury product. It also implies that as expenditures
on imported fruit juice increases, consumers change their consumption of Brazilian
orange juice more, in terms of percentage, than they change their consumption of the
same juice imported from the U.S. or the rest of the world. Furthermore, these results
have important implications for exporting countries in terms of making export decisions
in light of the expansion and contraction of the Japanese market for imported fruit juices
85
because of the change in expenditure. Under a situation where the Japanese market for
imported fruit juices expands because of an increase in expenditure, Brazil will become
much better off. This is because as the Japanese market for imported fruit juices expands
because of increasing expenditure, Brazilian orange juice market share will increase more
than proportionately. Other Exporters will not be as well off since they are expenditure
inelastic.
Given that Brazilian orange juice makes up the larger proportion of the total
imports of fruit juices in Japan, a one percent increase in expenditure on imported fruit
juices results in a far greater increase in actual imports; and, its market share would
increase further upon the expansion of the Japanese market of imported fruit juices over
time. However, under conditions in which the economy goes to recession, or expenditure
growth slows down, Brazil will be worse off because, a given percentage decrease in
expenditure on imported fruit juices results in a far greater decrease in actual imports; and
its market share would decrease further upon the contraction of the market of imported
fruit juices over time because of its larger expenditure elasticity. The fact that recession
has been more frequent in Japan over the past few years requires Brazil to devise an
effective export strategy which takes account of the performance of the economy.
In addition to recession, the growth of population is another major factor
anticipated to affect the demand for imported fruit juices in Japan as a result of its aging
population. The population growth of Japan has turned negative in 2006 (Statistics
Bureau of Japan). With per capita income growing at 2% per annum and assuming that it
will remain constant until 2020, and population growth starting to take negative rate since
2006, the growth of demand for fruit juices imported into Japan is projected (Table 5.9).
86
The growth of demand for fruit juice in Japan is positive except for that of Israelis
grapefruit juice over the over the period 2006 through 2014. The demand for Israelis
grapefruit is negative not only due to the population growth but also negative expenditure
elasticity. Products which have positive expenditure elasticity will continue to grow at a
declining rate regardless of the negative growth of population except for U.S. grape juice
and Israelis and Italian other citrus juices. From the result of the simulation, it appears
that grape and other citrus juice will be more affected than the other juices. The demand
for Brazilian orange juices declined from 5.53% in 2005 when the growth of population
was 0.3% to 5.49% in 2006 when the growth of population turned negative. It will
continue to shrink over the period 2006 through 2020 while the demand for U.S. orange
is projected to shrink at 1.12.9 to 0.66% over the same period.
Among U.S. products, apple juice will grow at a higher rate (more than 1%) while
grape juice will grow at the smallest rate (less than 0.25%). These simulations were
made under the assumption that the growth of per capita income will remain constant at
2% per annum over the period 2006 through 2020. The increase in the growth of per
capita income will offset the decrease in population growth so that the decline in the
growth of demand may be checked. If income grows at more than 2%, demand may
increase, though population growth slows down. The prospect of the growth of demand
for fruit juices will depend on the growth of per capita income relative to the decline in
growth of the population. If both move in the same direction, the decline of the growth
rate of demand for fruit juices will be greater.
87
Table 5.9 Projected estimates of the growth of demand for fruit juices in Japan Year Population
growth rate
U.S. oranges
Brazilian Oranges
U.S. grape fruits
Israelis grape- fruit
U.S. apples
Chinese apples
Thai pine apple
Philipp. p. apples
U.S grapes
Argenti- nean grapes
Israelis citrus
Italian citrus
2005 0.03 0.96 5.53 1.12 -0.09 1.68 1.33 0.83 0.67 0.29 0.41 0.44 0.26 2006 -0.01 0.92 5.49 1.08 -0.13 1.64 1.29 0.79 0.63 0.25 0.37 0.4 0.22 2007 -0.04 0.89 5.46 1.05 -0.16 1.61 1.26 0.76 0.6 0.22 0.34 0.37 0.19 2008 -0.07 0.86 5.43 1.02 -0.19 1.58 1.23 0.73 0.57 0.19 0.31 0.34 0.16 2009 -0.10 0.83 5.40 0.99 -0.22 1.55 1.20 0.7 0.54 0.16 0.28 0.31 0.13 2010 -0.13 0.80 5.37 0.96 -0.25 1.52 1.17 0.67 0.51 0.13 0.25 0.28 0.10 2011 -0.16 0.77 5.34 0.93 -0.28 1.49 1.14 0.64 0.48 0.10 0.22 0.25 0.07 2012 -0.19 0.74 5.31 0.90 -0.31 1.46 1.11 0.61 0.45 0.07 0.19 0.22 0.04 2013 -0.22 0.71 5.28 0.87 -0.34 1.43 1.08 0.58 0.42 0.04 0.16 0.19 0.01 2014 -0.25 0.68 5.25 0.84 -0.37 1.40 1.05 0.55 0.39 0.01 0.13 0.16 -0.01 2015 -0.28 0.65 5.22 0.81 -0.40 1.37 1.02 0.52 0.36 -0.01 0.10 0.13 -0.04 2016 -0.31 0.62 5.19 0.78 -0.43 1.34 0.99 0.49 0.33 -0.04 0.07 0.10 -0.07 2017 -0.35 0.58 5.15 0.74 -0.47 1.30 0.95 0.45 0.29 -0.08 0.03 0.06 -0.11 2018 -0.38 0.55 5.12 0.71 -0.50 1.27 0.92 0.42 0.26 -0.11 0.00 0.03 -0.14 2019 -0.4 0.53 5.10 0.69 -0.52 1.25 0.90 0.40 0.24 -0.13 -0.01 0.01 -0.16 2020 -0.43 0.50 5.07 0.66 -0.55 1.22 0.87 0.37 0.21 -0.16 -0.04 -0.01 -0.19
Own-price Elasticities
In order to assess the responsiveness of Japan’s imports to changes in prices, two
types of own-price elasticities (uncompensated and compensated) are calculated using
equation ⎟⎟⎠
⎞⎜⎜⎝
⎛−=
i
ij
i
ijuij w
ww
θπε and equation ⎟⎟
⎠
⎞⎜⎜⎝
⎛=
i
ijcij w
πε , respectively, where ijπ is the
Slutsky price coefficient.
Uncompensated (Marshallian) own price elasticity of imported juice i indicate
the percentage change in quantity demanded of imported juice i resulting from a one
percent change in its own price, holding nominal expenditures on imported juices
constant. Compensated (Slutsky/Hicksian) own price elasticity of juices i indicates the
percentage change in quantity demanded of juice i resulting from a one percent change
in its own price, holding real expenditures on imported juices constant. The
uncompensated price elasticities provide the responsiveness of demand resulting from
both substitution and income effects of a price change while the compensated price
elasticities would provide the responsiveness of demand resulting from the substitution
effect of a price change net of the income effect.
Results indicate that both uncompensated and compensated own price elasticities
of the demand for fruit juices in Japan are all negative and statistically different from zero
except for the ROW apple juice (Table 5.10). Among the 18 fruit juices, four fruit juices
are price elastic and two are unitary price elastic. These are Philippine pineapple juice,
U.S. orange juice, the ROW orange juice, Italian other citrus juice, Israelis other citrus
juice and Brazilian orange juice. Of these, the demand for the Philippines pineapple juice
is the most price elastic (-3.0543) followed by that of the U.S. orange juice (-1.5774), the
ROW orange juice (-1.4521), and Italian other citrus juice (-1.1745). The demand for
89
Brazilian orange juice (-1.0109) and Israelis other citrus juice (-1.0039) are unitary price
elastic.
Table 5.10 Own price elasticities of fruit juices in Japan Uncompensated own
price elasticities Compensated own price
elasticities Product
Estimate SE Estimate SE USA orange -1.577*** 0.3080 -1.5437*** 0.3100 Brazil orange -1.010*** 0.3404 -0.3112 0.3387 ROW orange -1.452*** 0.1649 -1.4487*** 0.1648 USA grapefruit -0.5835*** 0.1625 -0.5394*** 0.1624 Israel grapefruit -0.5453*** 0.1771 -0.5469*** 0.1762 ROW grapefruit -0.7108*** 0.1429 -0.7056*** 0.1435 USA apple -0.5191* 0.3150 -0.4722 0.3171 Chinese apple -0.5948*** 0.1848 -0.5474*** 0.1842 ROW apple 0.0609 0.2284 0.1410 0.2247 Thailand pineapple -0.8758*** 0.1945 -0.8714*** 0.1946 Philippines P. Apple -3.054*** 0.2731 -3.051*** 0.2730 ROW pineapple -0.6296** 0.3100 -0.6370** 0.3103 USA grape -0.8484*** 0.2029 -0.8404*** 0.2029 Argentina grape -0.6447 0.4223 -0.6430 0.4226 ROW grape -0.6403*** 0.2301 -0.6215*** 0.2303 Israel other citrus -1.003*** 0.1928 -0.9994*** 0.1925 Italy other citrus -1.1745*** 0.2567 -1.1725*** 0.2562 ROW other citrus -0.9584*** 0.1061 -0.9520*** 0.1060 *** (**)* significance at 1%, 5% and 10%
Although the absolute value of uncompensated price elasticities of most of the
fruit juices are higher than that of the respective compensated price elasticities, the
magnitude of difference between the two elasticities is very small. Some notable
exceptions are Brazilian orange juice, U.S. orange juice, U.S. grapefruit juices, and U.S.
apple juice. These products have a relatively larger income effect. Suffice to mention
the high income elasticity of the Brazilian orange juice. The uncompensated price
elasticity of Brazilian orange juice is -1.01096 while that of compensated price elasticity
is -.311240. This large difference is due to a large income effect. This is apparent in the
large income elasticity of the Brazilian orange juice (2.7522). Based on the magnitude of
90
differences between the two elasticities, one can see which products have a relatively
larger income effect.
When commodities have elastic demand, price discounting can be an effective
tool to expand exports and increasing total revenue for the exporters. Countries which
can benefit from expanding their exports through reducing prices are U.S. (orange juice),
Brazil (orange juice), ROW (orange juice), Philippines (pineapples), Israel (other citrus)
and Italy (other citrus). Since the demand for other juices from other countries including
U.S. (grapefruit, apple, grape juices) are price inelastic, export supply expansion through
price-oriented promotional measures, trade negotiations or other marketing activities that
involve reduction of prices will result in a reduction of total revenue and a relatively
smaller increase in quantity demand than countries with an elastic demand.
Cross-price Elasticities
Like the case with own price elasticities, two types of cross-price elasticities,
uncompensated and compensated, were calculated. The uncompensated (Marshallian)
cross-price elasticity of product i with respect to product j provides the percentage
change in the quantity of product i demanded resulting from a one percent change in the
price of product j , holding nominal expenditures on imported juices constant. The
Slutsky (compensated) cross-price elasticity of product i with respect to product j
indicates the percentage response in the quantity of product i demanded resulting from a
one percent change in the price of product j , holding real expenditures on imported
fruits constant. The elasticities were determined at the mean values of expenditure shares
over the period December, 1995 to May, 2005.
91
Results indicate that more compensated cross price elasticities are statistically
significant than uncompensated price elasticities, and most products are substitutes
(Tables B-1 and B-2). This is consistent with Hick’s second law of demand.
Since the uncompensated price elasticities are not net of income effects, they
don’t reflect the true substitution relationship. The uncompensated cross price elasticities
include both substitution and income effects and determines if two products are gross
substitutes and complements. Net substitutes and complements satisfy symmetry in the
sense that it is not possible for 1q to be a substitute to 2q and for 2q to be a complement to
1q at the same time. Symmetry of sign and magnitude does not hold in gross substitutes
and complements since it is possible for 1q to be a substitute to 2q and for 2q to be a
complement to 1q at the same time. Notice that net refers to compensated demand (and
thus confines itself to the substitution effect), while gross refers to uncompensated
demand. Net substitute products become gross complements if the income effect is both
adverse and large enough. The undesirable characteristic of the gross substitutes and
complements is that they are asymmetric.
The compensated price elasticities, which are of income effects, are chosen to
discuss the substitution and complementary relationships. They determine if two
products are net substitute and complements. Net substitutes and complements are
symmetric. The whole set of the estimates of uncompensated and compensated price
elasticities are presented Tables B-1 and B-2 which provide different categories of
products-gross and net substitutes, gross and net complements, gross complements and
net substitutes and gross substitutes and net complements.
92
Gross and net substitutes have positive and statistically significant uncompensated
and compensated cross price elasticities. Gross and net complements are negative and
statistically significant uncompensated and compensated cross price elasticities. Gross
complements and net substitutes are those products whose uncompensated cross price
elasticities are negative and compensated cross price elasticities are positive while gross
substitutes and net complements are those products whose uncompensated cross price
elasticities are positive and compensated cross price elasticities are negative. Results
indicate that Brazilian orange juice/U.S. grapefruit juice, Brazilian orange juice/U.S.
apple juice, Brazilian orange juice/Thai pineapple juice, Brazilian orange juice/ROW
grape juice, and Brazilian orange juice/ROW other citrus juice are gross complements
and net substitutes (Table B-1). Normally, we expect these products to be net substitutes.
However, they are also gross complements because of the strong income effect of the
Brazilian orange juice. In other words, when the prices of U.S. grapefruit juice, U.S.
apple juice, Thai pineapple juice, ROW grape juice and ROW other citrus juice fall, the
substitution effect may be so small that the consumer purchases more of Brazilian orange
juice and less of the other juices.
The estimates of the cross price elasticities of net substitutes and net complements
are presented in Tables 5.11 and 5.12. Products which are net substitutes and thus belong
to the same market structure include U.S. orange juice/Brazilian orange juice and U.S.
orange juice/Philippines pineapple juice, U.S. apple juice/Philippines pineapple juice
(Table 5.11). Products which have complementary relationship include Israelis grapefruit
juice and Thai pineapple juice (Table 5.12).
93
A decrease in the price of Brazilian orange juice has a larger negative effect on
the quantity demanded of U.S. orange juice. However, the decrease in the price of U.S.
orange juice has a very small negative effect on the demand for Brazilian orange juice.
This is not unexpected given that Brazil has the highest market share (25%) in Japan’s
market. Because of its high market share, it can influence the juice market in Japan.
However, since the demand for Brazilian orange juice is compensated price inelastic and
less uncompensated price elastic than that of U.S. orange juice, Brazil does not have a
reason to decrease the price of its orange juice. The benefit to Brazil comes mainly from
the increase in the level of income because of its high income elasticity.
Another important product to which U.S. orange juice is a substitute is the
Philippines pineapple juice. A decrease in the price of U.S. orange juice has a larger
negative effect on the quantity demanded of the Philippines pineapple juice while a
decrease in the price of the Philippines pineapple juice has a very small negative effect on
the demand for U.S. orange juice. This implies that the U.S. may take some market share
from the Philippines pineapple juice should the price of the Philippines pineapple juice
remain constant. Nonetheless, given that the demand for both U.S. orange juice and the
Philippines pineapple juice are price elastic, both have reasons to decrease price to raise
total sales. The move by both countries to decrease price will positively impact the
demand for their respective products.
Similarly, the U.S. orange juice is also a substitute to U.S. apple juice, which, in
turn, is a substitute to U.S. grapefruit juice, and vice versa. U.S. grapefruit juice is also a
substitute to the ROW grapefruit juice. Since all of them are price inelastic, the change in
price of one product will not have a big impact on the demand for other product.
94
Table 5.11 Cross-price elasticity estimates of substitutes Uncompensated cross price
elasticity Compensated cross price
elasticity Products
Estimate SE Estimate SE US orange/Brazilian orange 1.0173*** 0.3634 1.1356*** 0.3581 US orange/U.S. apple 0.4404*** 0.1756 0.4668*** 0.1768 US orange/Philipp. pineapple 0.1186** 0.0612 0.1221** 0.0612 Brazilian orange/U.S. oranges 0.1242 0.1021 0.3236*** 0.1020 Brazilian orange/ROW pineapple 0.0583* 0.0334 0.0830** 0.0334 ROW orange/U.S. apple 0.4038* 0.2144 0.4097* 0.2157 ROW orange/ROW apple 0.6669** 0.3216 0.6842** 0.3160 ROW orange/Argent. grape 0.1121* 0.0696 0.1131* 0.0696 U.S. grapefruit/ROW grapefruit 0.1254*** 0.0380 0.1315*** 0.0382 U.S. grapefruit/U.S. apple 0.3001*** 0.1184 0.3311*** 0.1191 U.S. grapefruit/ROW grapes 0.2235* 0.1183 0.2589** 0.1184 Israelis g. fruit/Italian citrus 0.1808** 0.0779 0.1797** 0.0776 ROW grapefruit/U.S. grapefruit 0.9143*** 0.2781 0.9515*** 0.2769 U.S. apples/U.S. oranges 0.5358** 0.2246 0.5955*** 0.2256 U.S. apples/ROW. orange 0.2070* 0.1235 0.2338* 0.1231 U.S. apples/U.S. grapefruit 0.4047** 0.1706 0.4714*** 0.1695 U.S. apples/Philippines pineapple 0.1116** 0.0558 0.1179** 0.0557 U.S. apples/ROW citrus 0.1614* 0.0847 0.1821** 0.0844 Chinese apple/U.S. grape 0.2601** 0.1198 0.3005** 0.1195 ROW apple/ROW orange 0.1184* 0.0619 0.1341** 0.0619 ROW apple/Israelis other citrus 0.0703 0.0480 0.0810* 0.0479 Philippines pineapple /U.S. orange 1.1434** 0.5806 0.0810* 0.0479 Philippines pineapple /U.S. apple 0.8645** 0.4145 1.166** 0.5846 Philippines pineapple/Argentinean grape 0.6659*** 0.2540 0.8828** 0.4171 Philippines pineapple/ROW grape 0.9049* 0.5209 0.6688*** 0.2541 Philippines pineapple/Israelis citrus 0.6247** 0.2798 0.9257* 0.5210 ROW pineapple/Brazilian orange 2.5687*** 0.9605 2.3586** 0.9501 ROW pineapple/Argentinean grape 0.6218** 0.1958 0.6143*** 0.1959 U.S. grape/Chinese apple 0.3425** 0.1406 0.3520** 0.1400 U.S. grape/Argentinean grape 0.1731** 0.0696 0.1743** 0.0696 U.S. grape/ROW citrus 0.1362** 0.0636 0.1395** 0.0636 Argentinean grape/ROW orange 0.3934* 0.2461 0.3996* 0.2461 Argentinean grape/Philipp. pineapple 0.5516*** 0.2103 0.5531*** 0.2102 Argent. grape/ROW pineapple 0.5975*** 0.1909 0.5992*** 0.1911 Argent. grape/U.S. grape 1.1695** 0.4713 1.1815** 0.4718 ROW grape/U.S. grapefruit 0.2995** 0.1479 0.3231** 0.1478 ROW grape/Philipp. pineapple 0.1061** 0.0610 0.1083* 0.0609 Israelis other citrus /Brazilian orange 0.5567 0.3710 0.6092* 0.3666 Israelis other citrus /ROW apple 0.5721 0.3642 0.6062* 0.3585 Israelis other citrus /Philipp. pineapple 0.2154** 0.0960 0.2169** 0.0959 Italian other citrus /Israelis grapefruit 0.2672** 0.1171 0.2702** 0.1167 ROW other citrus /U.S. apple 0.3982** 0.1902 0.4129** 0.1914 ROW other citrus /U.S. grape 0.3300** 0.1583 0.3460** 0.1579 *** (**)* significance at 1%, 5% and 10%
95
A decrease in the price of Brazilian orange juice has a larger negative effect on
the quantity demanded of the ROW pineapple juice. However, the decrease in price of
the ROW pineapple juice has a very small negative effect on the demand for Brazilian
orange juice. Nonetheless, given that the demand for Brazilian orange juice is
compensated price inelastic, the decrease of Brazilian orange juice is disadvantageous to
both Brazil and the ROW. This is because consumers don’t significantly increase the
consumption of Brazilian orange juice in spite of price decrease. The best option for both
countries is to increase price. This will, however, benefit the ROW more if Brazil
increases the price of its orange juice.
Similarly, a decrease in the price of U.S. grape juice has a larger negative effect
on the quantity demanded of Argentinean grape juice. However, the decrease in price of
the Argentinean grape juice has a very small negative effect on the demand for U.S.
grape juice. Nonetheless, given that the demand for grape juice from both countries is
price inelastic, the move by either country to decrease the price of its grape juice is
disadvantageous to both of them. This is because consumers don’t significantly increase
the consumption of grape juice in spite of a decrease in the price. Hence, the product
option for both countries is to raise price. This will, however, benefit Argentina more if
the U.S. increases the price of its grape juice.
Given that most of the imported juices are price inelastic, most exporters can’t
increase their market share at the expense of their rivals through reducing prices. Some
notable exceptions are the Philippines pineapple juice, U.S. and ROW orange juice. In
conclusion, product differentiation/promotion appears to be a better option to increase
96
market share. Product differentiation will provide exporters with some monopolistic
power over their products.
Table 5.12 Cross-price elasticity estimates of complements Uncompensated cross price
elasticity Compensated cross price
elasticity Products
Estimates SE Estimates SE U.S. orange/ROW grapefruit -0.1246** 0.0560 -0.1194** 0.0563 U.S. orange/Israelis citrus -0.2249*** 0.0865 -0.2146** 0.0862 U.S. orange/ROW citrus -0.1548** 0.0730 -0.1431** 0.0729 Brazilian orange/ROW orange -0.1598*** 0.0506 -0.0706 0.0503 Brazilian orange/Chinese apple -0.2401*** 0.0823 -0.0398 0.0815 Brazilian orange/ROW apple -0.7516*** 0.1433 -0.2968** 0.1395 Brazilian orange/U.S. grape -0.2100*** 0.0720 -0.0389 0.0712 Brazilian orange/ROW citrus -0.0835** 0.0360 -0.0145 0.0358 U.S. grapefruit/Thai pineapple -0.2349*** 0.0436 -0.2289*** 0.0437 U.S. grapefruit/Philippines pineapple -0.1826*** 0.0430 -0.1785** 0.0430 U.S. grapefruit/U.S. grape -0.2259** 0.1127 -0.1920* 0.1126 ROW. grapefruit/U.S. orange -0.8079** 0.3634 -0.7745** 0.3652 ROW. grapefruit/Italian citrus -0.3978*** 0.1204 -0.3898*** 0.1200 U.S. apple/ROW apple -0.8014** 0.3183 -0.6650** 0.3111 U.S. apple/ROW pineapple -0.2705*** 0.0821 -0.2631*** 0.0822 U.S. apple/Argentinean grape -0.3180*** 0.0629 -0.3105*** 0.0629 U.S. apple/Israelis citrus -0.2350*** 0.0803 -0.2168*** 0.0799 Chinese. apple/ROW p. apple -0.1052* 0.0588 -0.0994* 0.0588 Chinese. apple/ROW grape -0.2106* 0.1205 -0.1684 0.1206 ROW apple/Brazilian orange -0.5797*** 0.2162 -0.4566** 0.2146 ROW apple/U.S. apple -0.2560** 0.1061 -0.2285** 0.1069 Thai pineapple/U.S. grapefruit -1.7249*** 0.3230 -1.6922*** 0.3229 Philippines pineapple/U.S. grapefruit -1.9287*** 0.4587 -1.9027*** 0.4587 ROW pineapple/U.S. apple -1.6233*** 0.5187 -1.6702*** 0.5219 ROW pineapple/Chinese apple -0.7491* 0.4808 -0.8093* 0.4788 U.S. grape/U.S. grapefruit -0.2603* 0.1468 -0.2497* 0.1465 Argentinean grape/U.S. apple -1.9336*** 0.3870 -1.9227*** 0.3894 Israelis other citrus/U.S. orange -0.7189** 0.2810 -0.7040** 0.2829 Israelis other citrus/U.S. apple -0.5692*** 0.2041 -0.5575*** 0.2054 Italian other citrus/ROW grapefruit -0.2533*** 0.0771 -0.2521*** 0.0776 ROW other citrus/U.S. orange -0.4326** 0.2100 -0.4139** 0.2110 *** (**)* significance at 1%, 5% and 10%
97
CHAPTER 6 MARKET STRUCTURES AND STRATEGY OPTIONS
Market Structures
This chapter discusses four market structure scenarios:(1) block independence
(direct) with non-uniform substitution; (2) block independence (direct) with uniform
substitution; (3) block-wise dependence with non-uniform substitution; and (4) block-
wise dependence with uniform substitution. They are identified based on the structure of
competition (block independence or block-wise dependence) and degree of product
substitutability (uniform or non-uniform) consistent with the assumptions of consumer
preferences.
Block Independence (Direct) with Non-uniform Substitution
Block independence with non-uniform substitution is a case where competition
between products occurs within the same product group such that the effect of a change
in price of a given product on the demand for another product varies from product to
product. This means, for example, that the effect of a change in price of the ROW orange
juice on the demand for U.S. orange juice is different from that on the demand for
Brazilian orange juice.
In this market structure, consumers care about the country of origin of the product
because the change in marginal utility of a dollar spent on product i caused by an extra
dollar spent on product j is different from the change in the marginal utility of a dollar
spent on product k caused by an extra dollar spent on product j . This means, for
example, that the change in marginal utility of a dollar spent on Brazilian orange juice
98
caused by an extra dollar spent on the ROW orange juice is different from the change in
marginal utility of a dollar spent on the U.S. orange juice caused by an extra dollar spent
on the ROW orange juice. This implies that consumers may pay a different price for
products of the same group since they perceive one product as differentiated from the
other.
Block Independence (Direct) with Uniform Substitution
Block independence with uniform substitution is similar to the block
independence with non-uniform substitution in the fact that the competition is occurring
within the same product group. However, unlike the case with the block independence
with non-uniform substitution, the effect of a change in price of a given product in this
market structure is the same for all products in the group. This means, for example, that
the effect of a change in the price of the ROW orange juice on the demand for Brazilian
orange juice is the same as that on the demand for U.S. orange juice because the change
in marginal utility of a dollar spent on Brazilian orange juice caused by an extra dollar
spent on the ROW orange juice is the same as that on U.S. orange juice caused by an
extra dollar spent on the ROW orange juice. Since the country of origin does not make a
difference in this market structure, the competition between products is so high that a
slight change in the price of one product will significantly affect the demand for another
product. Consumers only care about prices. They buy a certain product when its price is
lower than the price of the rival product since they perceive all products in the same
group as homogenous.
Block-wise Dependence with Non-uniform Substitution
Block-wise dependence with non-uniform substitution is the case where a product
is competing with another product outside its product group such that the effect of a
99
change in price of a product in one group on the demand for another product which
belongs to a different group is the same for all pairs of products in the two groups. This
implies that the competition between any two products in two different groups is so high
that a slight change in the price of one product in one group will significantly affect the
demand for another product in another group. This is because consumers don’t care
about the country of origin of the product when they choose between products that belong
to different product groups. However, when they have to choose between products
within the same group, they consider the country of origin of the product. The effect of a
change in the price of a product in one group on the demand for another product within
the same group is different for all pairs of products within that group.
The competition in the block-wise dependent non-uniform substitution market
structure can be described in such a way that the effect of, for example, a change in price
of U.S. orange juice on the demand for Thai pineapple juice is the same as that of a
change in the price of Brazilian orange juice on the demand for the Philippines pineapple
juice. This implies that a slight decrease in the price of the U.S. orange juice will cause
consumers to quickly switch to the Philippines pineapple juice. Conversely, a slight
decrease in the price of Brazilian orange juice causes consumers to switch to the Thai
pineapple juice. It is assumed here that the price of Brazilian orange juice does not
change when the price of the U.S. orange juice decreases or vice versa.
The reason that consumers switch from the Thai pineapple juice to the Philippines
juice or vice versa is that they don’t care about the country of origin of the pineapple
juice. They only care about prices. They can choose to buy one juice if its price is less
than that of the competitors’ price. However, when they choose between products that
100
belong to the same group, they consider the country of origin. Consequently, the
competition between products in the same group (either within orange juice or pineapple
juice group) is not as high as is between products in different groups (orange juice versus
pineapple juice). For example, they may consider whether the orange juice is from the
U.S. or Brazil or the ROW when they choose between orange juices because the change
in marginal utility of a dollar spent on Brazilian orange juice caused by another dollar
spent on the ROW orange juice is not the same as the change in marginal utility of a
dollar spent on the U.S. orange juice caused by an extra dollar spent on the ROW orange
juice. The same is true with the pineapple juice in that the change in marginal utility of a
dollar spent on the Philippines pineapple juice caused by another dollar spent on the
ROW pineapple juice is not the same as the change in marginal utility of a dollar spent on
the Thai pineapple juice caused by an extra dollar spent on the ROW pineapple juice.
Consumers are focused not only on price changes. They also consider other physical
attributes. Therefore, they may pay a different price for the same product differentiated
by country of origin since they perceive one product as differentiated from the other.
Block-wise dependence with Uniform Substitution
Block-wise dependence with uniform substitution is similar to the block-wise
dependent with non-uniform substitution market structure in the fact that the competition
occurs between any two products in two different groups such that the effect of a change
in price of a product in one group on the demand for another product in another group is
the same for all pairs of products in the two groups. However, in this market structure,
the effect of a change in price of a product in one group on the demand for another
product within the same group is the same for all pairs of products within that group.
This implies that consumers don’t care about the country of origin of the product when
101
they choose between products within the same group. Recall that the county of origin is
important when consumers choose between products within the same group in the case of
block-wise dependent with non-uniform substitution market structure.
Table 6.1 Importance of country of origin in five market structures Country of origin Market structures Models
Between groups Within a group
Product-wise dependence with non-uniform substitution 4.1 Important Important
Block independence with non-uniform substitution
4.9
NA Important
Block independence with uniform substitution
4.20
NA Not important
Block-wise dependence with non-uniform substitution
4.14
Not important Important
Block-wise dependence with uniform substitution
4.22
Not important Not important
Note: NA not applicable
Parameter and Elasticity Estimates in Five Market Structures
Changes in marginal utilities are related to cross price effects as mpupv jij
iij =
where ijv is the specific cross price effects; ip is the price of product i ; iju is the rate of
change of marginal utilities; jp is the price of product j and m is total expenditure.
Hence, the variation in cross price effects between models is explained by the variation in
the rate of changes of marginal utilities. Based on the assumptions of preference
structure associated with the models, we can hypothesize which models will provide
higher cross price effects. Theoretically, the cross price effects of the uniform
substitution models (4.20 and 4.22) should be higher than those of the non-uniform
substitute models (4.9 and 4.14) because the former models subject the products to be
102
close substitutes such that extra consumption of one product will have a larger effect on
the marginal utility of another product.
Given that price elasticities are defined as ( )( )ijji qppq ∂∂ where is iq and jp
are the quantity and prices of products, the variation in elasticities between models is
explained by the variation in the response of demand to changes in price ( )ji pq ∂∂ .
Based on the nature and degree of substitutability that the different models impose on
products, we can hypothesize which models will result in higher demand response to
changes in price. In light of this, the uniform substitution models (4.20 and 4.22) should
theoretically result in a higher demand response to changes in prices than the non-
uniform substitute-models (4.9 and 4.14).
Parameter estimates
Results indicate that while only U.S/Argentinean grape juices are substitutes in
the selected market structure discussed in chapter 5, which is (4.1), U.S. orange/Brazilian
orange juice; U.S. grapefruit/Israelis grapefruit; and U.S. grape/Argentinean grape are
substitutes in all the four market structure scenarios (Table 6.2). Though U.S. apple juice
and Chinese apple juice are the same products differentiated by country of origin, there is
no substitution between them in all the five market structures.
Thai/Philippines pineapple juices are substitutes only in the block independence
with non-uniform substitution and block independence with uniform substitution market
structures. Israelis/Italian other citrus juices are substitutes only in the block
independence with uniform substitution market structure.
Grapefruit/apple juices are substitutes only in the block-wise dependent with
uniform substitution market structure. Apples/pineapples, pineapple/other citrus, and
103
grape/other citrus juices are substitutes in both the block-wise dependent non-uniform
substitution and the block-wise dependent uniform substitution market structures. The
competition between any two of these products (e.g. U.S. apple juice and Thai pineapple
juice) is so high that a slight change in the price of either product has a significant impact
on the demand for other product. They are competing with each other regardless of the
country of origin. For example, it does not matter whether apple juices are from the U.S.
or China or pineapple juices are from Thailand or the Philippines. If, for example, the
price of U.S. apple juice decreases, the demand for Thai pineapple juice will significantly
decrease. This implies that exporters of these substitute products need to watch the prices
of the respective substitute products. The competition between the same products in the
selected market structure (4.1) is not as high as the case in the block-wise dependent
market structures because consumers consider the country of origin in that market
structure. They don’t switch from product to product for some slight price reduction.
They also value the product attributes.
Orange/apple and grapefruit/pineapple are complements in both the block-wise
dependent non-uniform substitution and the block-wise dependent uniform substitution
market structures.
The above discussion implies that if we happened to pick one of the four market
structure scenarios, we would come up with wrong results leading to wrong conclusions.
This is because most of the results obtained from the unselected models (4.9, 4.14, 4.20
and 4.22) are inconsistent with that of the selected model (4.1). Unlike this study, most
studies pick one of these unselected models and estimate demand parameters. As we can
104
see here, we found relationships between products in the unselected models while there
was none in the selected model. Such results will lead to wrong conclusions.
Table 6.2 Relative price coefficients of fruit juices in five market structures Products Equation
(4.1) Equation
(4.9) Equation
(4.20) Equation
(4.14) Equation
(4.22) U.S. orange/Brazilian orange .0395 .0579** .0620*** .0482* .0614** U.S. orange/ROW orange .0043 .0055 .0007*** .0079 .0006*** Brazilian orange/ROW orange -.0222 .0120 .0206*** .0210* .01967*** U.S. grapefruit/Israelis grapefruit .0033 .0081* .0030*** .0121*** .0090*** U.S. grapefruit/ROW grapefruit .0102*** .0058** .0056*** .0047** .0057*** Israelis. grapefruit/ROW grapefruit .0022 .0026* .0006*** .0024* .0005*** U.S. apple/Chinese apple -.0094 -.0099 -.0135** -.0001 -.0159** U.S. apple/ROW apple -.0445** -.0393*** -.0181*** -.0250* -.0206*** Chinese apple/ROW apple .0156 -.0157 -.0195*** -.0016 -.0160*** Thai. pineapple/Philippines pine apple .0003 .0022* .0025* -.0006 .0009 Thai. pineapple/ROW pineapple -.0005 -.0005 .0003 -.0008 .0001 Philippines pineapple/ROW p. apple .0008 .0017 .0006 .0005 .0001 U.S. grapes/Argentinean grapes .0108** .0071* .0024* .0066* .0022* U.S. grapes/ROW grapes .0081 .01492* .0171** .0106 .01513** Argentinean grapes/ROW grapes .0023 .0015 .0028* .00085 .0032* Israelis citrus/Italian citrus .0013 .0039 .0018* .0022 .0017 Israelis citrus/ROW citrus .0024 .0034* .0017* .0024 .0010 Italian citrus/ROW citrus -.0010 -.0003 .0017* -.0019 .0011 Product groups Equation
(4.14) Equation
(4.22) Orange/grapefruit NA NA -.0178 -.0175 Orange/apple NA NA -.2276*** -.1197*** Orange/pineapple NA NA .0134 -.0118 Orange/grape NA NA -.0186 -.0487** Orange/other citrus NA NA -.0174 -.0011 Grapefruit/apple NA NA .0081 .0406** Grapefruit/pineapple NA NA -.0309*** -.0393*** Grapefruit/grape NA NA .0154 .0128 Grapefruit/other citrus NA NA -.0094 -.0129* Apple/pineapple NA NA .0293*** .0282*** Apple/grape NA NA -.0119 -.0055 Apple/other citrus NA NA .0194 .0088 Pineapple/grape NA NA .0107 .0053 Pineapple/other citrus NA NA .0095** .0112** Grape/other citrus NA NA .0270*** .0153* Note: NA: Not applicable; all the cross price effects of equation (4.1) are not presented here since they don’t have counterparts in the other models for comparison. Expenditure elasticities
The estimates of the expenditure elasticities in all four market structure models
(4.9, 4.14, 4.20 and 4.22) are less than one except for that of Brazilian orange juice
105
(Table 6.3). This is consistent with the estimates of (4.1). However, the expenditure
elasticities of Argentinean grape juice and Israelis grapefruit juice are negative in (4.9)
and (4.14), respectively, though none of them are statistically significant. The
expenditure elasticity of Israelis grapefruit is also negative and statistically insignificant.
The negative expenditure elasticity of the ROW pineapple juice in (4.1) turns positive in
the four models. However, in both cases, it is not statistically different from zero.
Table 6.3 Expenditure elasticity estimates of fruit juices in Japan in five market structures Product Equation
(4.1) Equation (4.9)
Equation (4.20)
Equation (4.14)
Equation (4.22)
U.S. orange 0.4654*** 0.3859*** 0.3756*** 0.4763*** 0.3303*** Brazilian orange 2.7522*** 2.9331*** 2.9254*** 2.8040*** 3.0748*** ROW orange 0.1047 0.3316** 0.2800*** 0.1249 0.2298*** U.S. grapefruit 0.5463*** 0.3681*** 0.3563*** 0.4501*** 0.3714*** Israelis grapefruit -0.0630 0.0911 0.1199 -.07707 0.0935* ROW grapefruit 0.4603 0.0463 0.1695* 0.2792* 0.1382* U.S. apple 0.8252*** 0.7092*** 0.6275*** 0.9190*** 0.6885*** Chinese apple 0.6504*** 0.5079*** 0.5274*** 0.6800*** 0.1124*** ROW apple 0.4842*** 0.2847*** 0.3110*** 0.4053*** 0.2607*** Thai pineapple 0.4048* 0.4213*** 0.3795*** 0.0882 0.6196*** Philippines pineapple 0.3212 1.0363*** 1.0774*** 0.1986 0.8107*** ROW pineapple -0.8262** 0.0953 0.1150 0.0146 0.0405 U.S. grape 0.1301 0.1417* 0.1460** 0.1246* 0.1480** Argentinean grape 0.1921 -0.0091 0.1662 0.0690 0.2152* ROW grape 0.2912*** 0.1764** 0.1673** 0.2022** 0.1991** Israelis other citrus 0.2065 0.3238*** 0.3954*** 0.2400*** 0.4390*** Italian other citrus 0.1153 0.5177*** 0.5075*** 0.3645*** 0.5500*** ROW other citrus 0.2578 0.3604*** 0.3344*** 0.2647*** 0.2361*** *** (**)* Significance at 1%, 5% and 10%
Among the 18 fruit juices, only the demand for Brazilian orange juice is
expenditure elastic in all four models. This is also the case in (4.1). The expenditure
elasticity estimate of Brazilian orange in the four models is higher compared to the
estimates of (4.1). It has the highest value in the (4.22). Brazilian orange juice, the ROW
orange juice, Israelis and Italian other citrus juices have consistently higher expenditure
elasticity in the four models compared to (4.1). All four fruit juices (orange, grapefruit,
apple and grape juices) that the U.S. exports to Japan are expenditure inelastic in all
106
models. Similarly, the demand for these products exported by ROW is also expenditure
inelastic.
In spite of some pattern, results indicate no systematic association of estimates of
particular models with particular products. In other words, a particular model does not
have the highest expenditure elasticities for all products or the lowest estimates for all
products. Some products have the highest value in some models and some other products
have their highest values in other models, implying that the magnitude of the expenditure
elasticity is not theoretically associated with a particular market structure. Given that the
same price holds for each product across the five models, the variation in expenditure
elasticity across models is expected to result from the responsiveness of the quantity
demanded of the product to a unit change in price.
Price elasticities
Results indicate that both uncompensated and compensated price elasticities are
all negative and statistically different from zero except for the ROW apple juice (Table
6.4 and 6.5). Among the 18 fruit juices, five fruit juices are consistently uncompensated
price elastic in all four models (4.9, 4.14, 4.20 and 4.22). These are the U.S. orange juice,
Brazilian orange juice, ROW orange juice, the Philippines pineapple juice and Italian
other citrus juice. These products were also price elastic in (4.1).
Similarly, the demand for the Philippines pineapple juice is consistently the most
price elastic in all models. However, it is less elastic in the four models (4.9, 4.14, 4.20
and 4.22) as compared to (4.1). Other products which are less price inelastic in the four
models include U.S. grape juice, Israelis other citrus, ROW pineapple and other citrus
juice (Table 6.4). Most of the products are more price elastic in the four models than in
(4.1). Such products include Brazilian orange juice, U.S. grapefruit juice, ROW
107
grapefruit juice, Thai pineapple juice and Argentinean grape juice. The ROW apple juice
has turned its negative sign in (4.1) to positive in the other four models. Generally,
products which were uncompensated price inelastic in (4.1) are also price inelastic in all
four models.
Table 6.4 Uncompensated own price elasticity estimates of fruit juices in Japan Product Equation
(4.1) Equation
(4.9) Equation
(4.20) Equation
(4.14) Equation
(4.22) U.S. orange -1.5774*** -1.5840*** -1.5483*** -1.4733*** -1.5022*** Brazilian orange -1.0109*** -2.3729*** -2.4111*** -1.4440*** -2.2496*** ROW orange -1.4521*** -1.1494*** -1.1678*** -1.0803*** -1.0559*** U.S. grapefruit -0.5835*** -0.8491*** -0.8333*** -0.6174*** -0.8185*** Israelis grapefruit -0.5453*** -0.5808*** -0.5989*** -0.4937*** -0.5330*** ROW grapefruit -0.7108*** -0.8408*** -0.8658*** -0.8974*** -0.8012*** U.S. apple -0.5191* -0.4047*** -0.5602*** -0.1928*** -0.4825*** Chinese apple -0.5948*** -0.5705*** -0.4911*** -0.4602*** -0.0213*** ROW apple 0.0609 -0.2055*** -0.3512*** -0.1519*** -0.2287*** Thai pineapple -0.8758*** -0.9244*** -0.9458*** -1.1098*** -1.0693*** Philippines p. apple -3.0543*** -2.405*** -2.3447*** -2.8050*** -1.4017*** ROW pineapple -0.6296** -0.3087*** -0.3143*** -0.1755*** -0.0726*** U.S. grape -0.8484*** -0.6185*** -0.5819*** -0.6434*** -0.5573*** Argentinean grape -0.6447 -0.9261*** -0.8741*** -1.0261*** -1.0159*** ROW grape -0.6403*** -0.5811*** -0.6154*** -0.7682*** -0.6475*** Israelis citrus -1.0039*** -0.9253*** -0.8710*** -0.9789*** -0.9159*** Italian other citrus -1.1745*** -1.1452*** -1.1151*** -1.1805*** -1.1478** ROW other citrus -0.9584*** -0.7813*** -0.7407*** -0.8660*** -0.5187*** *** (**)* Significant at 1%, 5% and 10%
However, this is not the case with regard to the compensated price elasticity
(Table 6.5). Only three products are consistently compensated price elastic in all four
models. These are U.S. orange juice, the ROW orange juice and the Philippines
pineapple juice. Brazilian orange juice is not consistently compensated price inelastic
while Italian other citrus juice is not consistently compensated price elastic in all four
models. Results indicate that some products are more price elastic in some models than
in others. For example, Brazilian orange juice is more price elastic in (4.22) while U.S.
and Israelis grapefruit juices are more price elastic in (4.9) and (4.20), respectively.
108
Table 6.5 Compensated own price elasticity estimates of fruit juices in Japan Product Equation
(4.1) Equation
(4.9) Equation
(4.20) Equation
(4.14) Equation
(4.22) U.S. Orange -1.5437*** -1.5561*** -1.5211*** -1.4388*** -1.4783*** Brazilian orange -0.3112 -1.6272*** -1.6674*** -0.7311*** -1.4679*** ROW orange -1.448*** -1.1386*** -1.1587*** -1.0762*** -1.0484*** U.S. grapefruit -0.5394*** -0.8193*** -0.8045*** -0.5810*** -0.7885*** Israelis grapefruit -0.5469*** -0.5784*** -0.5958*** -0.4957*** -0.5306*** ROW grapefruit -0.7056*** -0.8402*** -0.8639*** -0.8943*** -0.7996*** U.S. apple -0.4722 -0.3644*** -0.5246*** -0.1406*** -0.4434*** Chinese apple -0.5474*** -0.5335*** -0.4527*** -0.4106*** -0.0131*** ROW apple 0.1410 -0.1584*** -0.2998*** -0.0849*** -0.1856*** Thai pineapple -0.8714** -0.9198*** -0.9416*** -1.1088*** -1.0626*** Philippines pineapple -3.0519*** -2.3973*** -2.3365*** -2.8035*** -1.3955*** ROW pineapple -0.6370** -0.3079*** -0.3133*** -0.1754*** -0.5481*** U.S. grape -0.8404*** -0.6097*** -0.5729*** -0.6356*** -1.013*** Argentinean grape -0.6430 -0.9261*** -0.8725*** -1.0255*** -0.6346*** ROW grape -0.6215*** -0.5697*** -0.6045*** -0.7551*** -0.9062*** Israelis other citrus -0.9994*** -0.9182*** -0.8623*** -0.9736*** -1.1383*** Italian other citrus -1.1725*** -1.1362*** -1.1063*** -1.1742*** -0.5128*** ROW other citrus -0.9520*** -0.7723*** -0.7323*** -0.8594*** -1.4783*** *** (**)* Significant at 1%, 5% and 10%
Since the models represent different market structures, and that competitiveness is
associated with the degree of responsiveness, it can be implied that a product is more
competitive in the market structure that is represented by the model in which it is more
price elastic. In light of this, Brazilian orange juice is more competitive in the block-wise
dependent with uniform substitution market structure while U.S. grapefruit juice and
Israelis grapefruit juice are more competitive in the block independent with non-uniform
substitution and block independence with uniform substitution market structure,
respectively. Similarly, Thai pineapple and Argentinean grape juices are more
competitive in the block-wise dependent with non-uniform substitution market structure
(Tables D.1-D.4).
109
Market Strategy Options
Market strategies are associated with market structure (Table 6.6). Price
reduction is appropriate when the relationship between products is uniform while product
differentiation (product promotion) is appropriate when the relationship between products
is non-uniform.
Table 6.6 Market strategies by market structures Country of origin Market structure Models
Between groups Within a group
Product-wise dependence with non-uniform substitution 4.1 Product
differentiation Product differentiation
Block independence with non-uniform substitution
4.9
NA Product differentiation
Block independence with uniform substitution
4.20
NA Price reduction
Block-wise dependence with non-uniform substitution
4.14
Price reduction Product differentiation
Block-wise dependence with uniform substitution
4.22
Price reduction Price reduction
Note: NA not applicable
110
CHAPTER 7 SUMMARY, CONCLUSIONS AND IMPLICATIONS
Summary and Conclusions
The main theme of the study is to assess the competitiveness of the world’s
largest exporters of fruit juice to Japan. As a background to this theme, the study has
assessed the trend and pattern of global fruit production, trade and consumption. Results
indicate that there has been a sustained increase in global production and trade of both
citrus and non-citrus fruits over the last four decades. Most of the growth was accounted
for by developing countries, primarily in South America but also in Asia and to a lesser
extent in Africa.
In South America, the volume of production expanded considerably in Brazil and
Mexico. In Asia, production expanded significantly in China, India and Iran. Over the
last four decades, Brazil’s and China’s citrus production grew at an average rate of 4.5%
and 3.0%, respectively while that of the U.S. grew at 0.6% (FAO, 2005). The growth of
production of both citrus and non-citrus fruits in the U.S. has slowed since the 1980/90s
compared to that of Brazil, China, Mexico and Thailand, which have increased their
production significantly through expansion of cultivation.
Consequently, these countries have emerged as the world’s largest producers of
oranges, apples, lemons and limes and pineapples in the 1980/90s, respectively. The U.S.
was the world’s largest producer of both fruit types before the 1980/90s. The U.S. is still
the world’s largest producer of grapefruit. Although the yield per ha of both citrus and
non-citrus fruits in the U.S. is relatively large, the growth of production in the U.S. has
111
not been keeping up with the faster growth of fruit cultivation in the rest of the world.
This may have an implication for its competitiveness in the world market.
In terms of consumption, fruits are consumed mainly in industrialized countries,
not only because consumers in these countries have high income levels but also because
they have increasing concerns about healthy eating. However, the per capita
consumption of fruits in Japan is small compared to that of other industrialized countries,
implying that there is a potential to increase exports into Japan.
In order to assess the competitiveness of exporters, which is the main theme of the
study, a differential consumer demand approach has been applied. Since competitiveness
is associated with the type of market structure, two hypotheses (block
independent/uniform substitute hypothesis and block-wise dependent/uniform substitute)
are tested to identify the market structure of Japan’s fruit juice market. The block
independence /block-wise dependence involves the relationship among products that
belong to different product groups while that of uniform substitute hypothesis involves
the relationship among products that belong to the same product group. Thus, the above
two hypotheses have both between-group and within-group relationships.
The analysis of market structure in this study has involved the estimation of three
different versions of the Rotterdam model derived in light of the above mentioned
hypotheses (block independence/uniform substitute hypothesis and block-wise
dependence/uniform substitute hypothesis). The three versions of the Rotterdam model
are the block independent uniform substitute-Rotterdam model, the block-wise dependent
uniform substitute-Rotterdam model, and the relative price version of the Rotterdam
model.
112
The block independent uniform substitute-Rotterdam model describes a market
structure characterized by competition between products within the same group while the
block-wise dependent uniform substitute-Rotterdam model describes a market structure
characterized by competition between product groups. Block independent with uniform
substitution is a case where one product is uniformly competing with another product
which belongs to the same product group. Block-wise dependent with uniform
substitution is a case where products in one group are competing with products in another
group in a similar fashion. The relative price version of the Rotterdam model describes a
market structure whereby individual products compete with one another based on the
country of origin.
The three models were estimated for six fruit juices (orange, apple, grapefruit,
pineapple, grape, and other citrus) imported from 18 countries on data compiled over the
period January, 1995 to May, 2005.
Based on the likelihood ratio tests, both hypotheses are rejected, leading to the
selection of the relative price version of the Rotterdam demand model. This model
explains the allocation decisions better compared with the other versions.
Based on the parameter estimates of the selected model (the relative price version
of the Rotterdam model) and average expenditure shares, both income and price
elasticities are calculated. Results indicate that the expenditure elasticities for all fruit
juices are positive and statistically significant except for Israelis grapefruit and the ROW
pineapple juice. The demand for Brazilian orange juice is income elastic while the
demand for U.S. and the ROW products is income inelastic. The low income elasticities
in the face of a negative population growth indicate that the growth of demand for fruit
113
juices will be slow. At a zero population growth rate and 2% annual growth rate of per
capita income, the demand for Brazilian orange juices grows at about 5.5% while the
demand U.S. orange juice is 1.1%. With a declining population growth, the growth of
demand for these fruit juices and other fruit juices from the ROW is projected to decline
in the years to come.
The demands for U.S. orange juice and Philippines pineapple juice are price
elastic while the demand for other juices from other countries including Brazil and U.S.
(grapefruit, apple, grape juices) are price inelastic. Furthermore, the cross price
elasticities of most of the juices imported into Japan are below one. Some notable
exceptions are the U.S/Brazilian orange juice and U.S./Philippines pineapple juice.
In order to identify a marketing strategy consistent with the market structure, the
study has tested the plausibility of price reduction and product promotion given two
different market structures-block independent market competition and block-wise
dependent market competition. In this study, block independent (direct) market
competition is defined as a competition between the same fruit juice with different
countries of origin (example, U.S. orange juice versus Brazilian orange juice). It is based
on product group type in which case a change in the price of one product in a given
product group does not affect the demand for another product in another product group.
The block-wise dependent market competition is defined here as a competition between
different product groups (orange juice with apple juice). It indicates that a change in the
price of one product in a given product group affect the demand for another product in
another group in a similar fashion.
114
The choice between the price reduction and product promotion in the block
independent market competition involves a likelihood ratio test of two models (block
independent non-uniform substitute Rotterdam model versus and block independent
uniform substitute-Rotterdam model). Likewise, the choice between the same policy
options in the block-wise dependent market competition involves a likelihood ratio test of
two other models (block-wise dependent non-uniform substitute Rotterdam model and
block-wise dependent uniform substitute Rotterdam model).
The study finds that the plausibility of the market strategies depends on the
market structure. The price reduction is plausible in a market structure characterized by
block independent market competition model while product promotion is more plausible
in a market structure characterized by a block-wise dependent market competition.
Implications
The results of the study have important implications to countries exporting fruit
juices to Japan for making marketing strategies such as price reduction, product
differentiation as well as export supply plan in light of the expansion and contraction of
the Japanese market for imported fruit juices because of the change in income. The
effectiveness of a supply plan in raising market share through export expansion depends
on the estimates of expenditure and price elasticities. In light of this, the country which
benefits the most from the growth of income in Japan is Brazil. Brazilian orange juice
has the highest income elasticity and market share in Japan’s market.
Given that Brazilian orange juice is income elastic and makes up the largest
proportion of the total imports of fruit juices in Japan (25%), an increase in expenditure
on imported fruit juices results in a far greater increase in actual imports. Consequently,
its market share will increase upon the expansion of the Japanese market of imported fruit
115
juices over time. However, under conditions in which the economy goes to recession, or
income growth slows down, Brazil will be worse off because, a decrease in expenditure
on imported fruit juices results in a far greater decrease in actual imports; and, its market
share will decrease upon the contraction of the market of imported fruit juices over time.
The fact that recession has been more frequent in Japan over the past few years requires
Brazil to devise an effective export strategy which takes account of the performance of
Japan’s economy.
In addition to recession, the growth of population is another major factor
anticipated to affect the demand for imported fruit juices in Japan. The Japanese
population growth turned negative in 2006. Consequently, the growth of demand for fruit
juices will be slow in the years to come.
Supply plan (export supply expansion or contraction) also depends on the price
elasticity of demand. Given that the demand for the U.S. orange juice and the Philippines
pineapple juice is price elastic, price discounting can be an effective tool for the U.S.
citrus industry and the Philippines fruit industry in expanding their exports to Japan.
Since the demand for other juices from other countries including Brazil and U.S.
(grapefruit, apple, grape juices) are price inelastic, export supply expansion through
price-oriented marketing strategies, trade negotiations or other marketing activities that
involve reduction of prices will negatively impact the exporting country of the respective
product. These countries should reduce their cost of production, processing, and
marketing so that they can stay more competitive in Japan’s import market.
The degree of competition or market structure depends on the magnitude of cross
price elasticities. Given that the cross price elasticities of most of the juices imported into
116
Japan are below one, an exporter can’t take market share from another exporter quickly
through price reductions. Some notable exceptions are the U.S/Brazilian orange juice
and U.S. apple/Philippines pineapple juice. A decrease in the price of Brazilian orange
juice has a significant negative impact on the demand for U.S. orange juice but not vice
versa. However, since the demand for Brazilian orange juice is price inelastic, Brazil
does not have a reason to decrease price under the current market structure. Nonetheless,
if the current market structure changes to other market structures such as block
independent competition, Brazil may a have a reason to decrease its price since the
demand for its product becomes price elastic in the block independent market structure.
Therefore, the U.S. citrus industry should pay close attention to the development of the
Brazilian citrus industry. Assume, for example, that Brazil becomes more competitive by
introducing new technologies that reduce costs. Unless there is a similar response by the
U.S. citrus industry, there may be adverse effects on the demand for U.S. orange juice.
Similarly, the Philippines fruit industry should pay close attention to the
development of the U.S. orange and apple sector. In particular, further reductions in the
cost of production, processing or marketing activities of the U.S. orange and apple juices,
if not matched by decreases in the Philippines pineapple juice can have adverse effects on
the demand for the Philippines pineapple juice. Generally, because of the low cross price
elasticities of fruit juices in Japan, product promotion and further product differentiation
seems to be a more plausible option than a price reduction option for most countries to
stay more competitive in Japan’s fruit juice market.
The decision to use a particular marketing strategy depends on the degree of
competition or market structure. Price reduction is plausible in a market structure
117
characterized by block independent market competition while product promotion is
plausible in the market structure characterized by block-wise dependent market
competition. This implies that exporters can only compete through price reduction in the
block independent market competition while they can compete through product
promotion in the block-wise dependent competition. This means that if the competition
among fruit juices is restricted to within the product group (for example, orange juice
from one country is competing only with orange juice from another country), competition
through price reduction can be a plausible option since consumers are not influenced by
the origin of the product under such circumstances. However, when the competition
among fruits juices transcends beyond the product group (i.e. block-wise dependent
market competition), price reduction is not a plausible option, since consumers are
influenced by the origin of the product and base their buying decisions on the country of
origin of the product. In this case, product promotion is a better option than a price
reduction strategy.
118
APPENDIX A PRICE COEFFICIENTS OF FRUIT JUICES IN JAPAN
Table A-1 Relative price coefficients of fruit juices in Japan Products Estimates SE t-statistics p-value U.S. orange/Brazilian orange .039517 .029610 1.33459 .182 U.S orange/ROW orange .433277E-02 .753276E-02 .575191 .565 U.S orange./US grapefruit -.419052E-02 .012167 -.344408 .731 U.S. orange/Israelis grapefruit .619232E-03 .663284E-02 .093359 .926 U.S. orange/ROW grapefruit -.897070E-02 .404618E-02 -2.21708 .027 U.S orange/US apple .030962 .012648 2.44788 .014 U.S orange/Chinese apple .306733E-02 .011940 .256886 .797 U.S orange/ROW apple .018597 .020765 .895618 .370 U.S orange/Thai pineapple .633658E-02 .450683E-02 1.40600 .160 U.S orange /Philippines pineapple .870130E-02 .441177E-02 1.97229 .049 U.S orange/ROW pineapple -.396189E-02 .554243E-02 -.714829 .475 U.S orange/US grapes .013335 .011726 1.13718 .255 U.S orange/Argentinean grapes -.776343E-02 .520873E-02 -1.49046 .136 U.S orange/ROW grapes -.018846 .012377 -1.52258 .128 U.S orange/Israelis -.015829 .621954E-02 -2.54507 .011 U.S orange/Italian citrus -.233757E-02 .555697E-02 -.420656 .674 U.S orange/ROW citrus -.010770 .526588E-02 -2.04517 .041 Brazilian orange/ROW orange -.022273 .014888 -1.49601 .135 Brazilian orange/U.S grapefruit -.026268 .020834 -1.26081 .207 Brazilian orange/Israelis grapefruit .694081E-02 .014814 .468523 .639 Brazilian orange/ROW grapefruit -.781334E-02 .877812E-02 -.890093 .373 Brazilian orange/U.S apple -.045471 .030751 -1.47867 .139 Brazilian orange/Chinese apple -.070171 .023659 -2.96594 .003 Brazilian orange/ROW apple -.176993 .040216 -4.40110 .000 Brazilian orange/Thai pineapple -.494161E-02 .671348E-02 -.736073 .462 Brazilian orange/Philippines pineapple -.535420E-02 .643771E-02 -.831693 .406 Brazilian orange/ROW pineapple .030486 .991511E-02 3.07471 .002 Brazilian orange/U.S grape -.020175 .020691 -.975063 .330 Brazilian orange/Argentinean grape -.445004E-02 .729550E-02 -.609971 .542 Brazilian orange/ROW grape .232805E-02 .019210 .121192 .904 Brazilian orange/Israelis citrus .767426E-02 .933543E-02 .822057 .411 Brazilian orange/Italian citrus .717348E-02 .800765E-02 .895829 .370 Brazilian orange/ROW citrus -.011894 .010566 -1.12570 .260 ROW orange/U.S grapefruit .447447E-02 .571291E-02 .783221 .433 ROW orange/Israelis grapefruit .293407E-02 .351068E-02 .835756 .403 ROW orange/ROW grapefruit .254609E-02 .206505E-02 1.23294 .218 ROW orange/U.S apple .012993 .693677E-02 1.87306 .061 ROW orange/Chinese apple .121044E-02 .602607E-02 .200867 .841 ROW orange/ROW apple .021687 .010393 2.08664 .037 ROW orange/Thai pineapple .154743E-02 .200063E-02 .773470 .439 ROW orange/Philippines pineapple -.822573E-03 .195193E-02 -.421414 .673 ROW orange/ROW pineapple -.178218E-02 .269732E-02 -.660720 .509 ROW orange/U.S grape .819075E-03 .552104E-02 .148355 .882 ROW orange/Argentinean grape .365527E-02 .225755E-02 1.61913 .105
119
Table A-1 Continued Products Estimates SE t-statistics p-value ROW orange/ROW grape .278362E-02 .557101E-02 .499662 .617 ROW orange/Israelis citrus .223843E-02 .280532E-02 .797922 .425 ROW orange/Italian citrus .240343E-02 .243674E-02 .986331 .324 ROW orange/ROW citrus .207073E-02 .268926E-02 .770001 .441 U.S grapefruit/Israelis grapefruit .333785E-02 .499656E-02 .668030 .504 U.S grapefruit/ROW grapefruit .010222 .309575E-02 3.30180 .001 U.S grapefruit/U.S Apple .023026 .961601E-02 2.39453 .017 U.S grapefruit/Chinese Apple -.224555E-02 .914467E-02 -.245559 .806 U.S grapefruit/ROW Apple -.012776 .016320 -.782836 .434 U.S grapefruit/Thai pineapple -.018868 .353346E-02 -5.33991 .000 U.S grapefruit/Philippines pineapple -.014627 .348184E-02 -4.20103 .000 U.S grapefruit/ROW pineapple -.250641E-02 .424153E-02 -.590921 .555 U.S grapefruit/U.S grape -.016172 .912613E-02 -1.77205 .076 U.S grapefruit/Argentinean grape .258171E-02 .420159E-02 .614460 .539 U.S grapefruit/ROW grape .019431 .958208E-02 2.02785 .043 U.S grapefruit/Israelis citrus .299584E-02 .490923E-02 .610246 .542 U.S grapefruit/Italian citrus -.555928E-02 .446225E-02 -1.24585 .213 U.S grapefruit/R citrus .422583E-02 .401890E-02 1.05149 .293 Israelis grapefruit/ROW grapefruit .223771E-02 .191149E-02 1.17066 .242 Israelis grapefruit/U.S apple -.548759E-02 .656412E-02 -.835997 .403 Israelis grapefruit/Chinese apple .749554E-02 .548065E-02 1.36764 .171 Israelis grapefruit/ROW apple -.384568E-02 .899250E-02 -.427654 .669 Israelis grapefruit/Thai. pineapple -.518240E-03 .172109E-02 -.301112 .763 Israelis grapefruit/Philippines pineapple .218158E-02 .164435E-02 1.32671 .185 Israelis grapefruit/ROW pineapple -.184889E-03 .240440E-02 -.076896 .939 Israelis grapefruit/U.S grape -.212274E-02 .491121E-02 -.432225 .666 Israelis grapefruit/Argentinean grape .166582E-02 .185501E-02 .898013 .369 Israelis grapefruit/ROW grape -.341507E-03 .471477E-02 -.072433 .942 Israelis grapefruit/Israelis citrus -.161237E-02 .235387E-02 -.684987 .493 Israelis grapefruit/Italian citrus .467622E-02 .201741E-02 2.31794 .020 Israelis grapefruit/ROW citrus -.782211E-03 .245829E-02 -.318193 .750 ROW grapefruit/U.S apple .456180E-03 .376387E-02 .121200 .904 ROW grapefruit/Chinese apple -.600853E-03 .337186E-02 -.178197 .859 ROW grapefruit/ROW apple .661544E-03 .574249E-02 .115202 .908 ROW grapefruit/Thai. pineapple -.622186E-03 .109468E-02 -.568374 .570 ROW grapefruit/Philippines pineapple .357053E-05 .106516E-02 .003352 .997 ROW grapefruit/ROW pineapple .105967E-02 .147464E-02 .718598 .472 ROW grapefruit/U.S grape .266439E-02 .300608E-02 .886336 .375 ROW grapefruit/Argentinean grape .486089E-03 .123008E-02 .395170 .693 ROW grapefruit/ROW grape .275034E-02 .302456E-02 .909338 .363 ROW grapefruit/Israelis citrus -.253563E-03 .153248E-02 -.165459 .869 ROW grapefruit/Italian citrus -.437547E-02 .133541E-02 -3.27649 .001 ROW grapefruit/ROW citrus -.183852E-02 .146474E-02 -1.25519 .209 U.S Apple/Chinese apple -.946959E-02 .010773 -.878992 .379 U.S Apple/ROW apple -.044572 .017847 -2.49741 .013 U.S Apple/Thai pineapple .308132E-02 .323551E-02 .952346 .341 U.S Apple/Philippines pineapple .649085E-02 .313969E-02 2.06736 .039 U.S Apple/ROW pineapple -.014317 .464092E-02 -3.08495 .002 U.S Apple/U.S grape .935256E-02 .942572E-02 .992238 .321 U.S Apple/Argentinean grape -.017787 .354220E-02 -5.02142 .000 U.S Apple/ROW grape -.114664E-02 .908601E-02 -.126199 .900 U.S Apple/Israelis citrus -.012704 .450465E-02 -2.82024 .005 U.S Apple/Italian citrus .647074E-03 .379601E-02 .170461 .865
120
Table A-1 Continued Products Estimates SE t-statistics p-value U.S Apple/ROW citrus .979529E-02 .476176E-02 2.05707 .040 Chinese Apple/ROW apple .015667 .016813 .931881 .351 Chinese Apple/Thai. pineapple .225042E-02 .321070E-02 .700913 .483 Chinese Apple/Philippines pineapple .380558E-02 .308706E-02 1.23275 .218 Chinese Apple/ROW pineapple -.660875E-02 .431613E-02 -1.53117 .126 Chinese Apple/U.S grape .021190 .875605E-02 2.42002 .016 Chinese Apple/Argentinean grape -.122730E-02 .353006E-02 -.347671 .728 Chinese Apple/ROW grape -.013886 .881970E-02 -1.57448 .115 Chinese Apple/Israelis citrus .699838E-03 .446597E-02 .156704 .875 Chinese Apple/Italian citrus .246934E-02 .378790E-02 .651903 .514 Chinese Apple/ROW citrus .444569E-02 .425468E-02 1.04489 .296 ROW Apple/Thai pineapple .681398E-02 .572009E-02 1.19124 .234 ROW Apple/Philippines pineapple .162458E-04 .564016E-02 .002880 .998 ROW Apple/ROW pineapple .956209E-02 .733031E-02 1.30446 .192 ROW Apple/U.S grape -.456384E-02 .015306 -.298180 .766 ROW Apple/Argentinean grape .696273E-02 .661634E-02 1.05235 .293 ROW Apple/ROW grape -.014143 .016106 -.878080 .380 ROW Apple/Israelis citrus .012733 .803583E-02 1.58456 .113 ROW Apple/Italian citrus .227829E-02 .709473E-02 .321124 .748 ROW Apple/ROW citrus .516621E-02 .720261E-02 .717269 .473 Thai Pineapple/Philippines pineapple .396233E-03 .152006E-02 .260668 .794 Thai pineapple /ROW pineapple -.557884E-03 .154019E-02 -.362217 .717 Thai. pineapple /U.S grape .197651E-02 .356764E-02 .554011 .580 Thai pineapple /Argentinean grape -.135005E-02 .197110E-02 -.684921 .493 Thai pineapple /ROW grape .133803E-02 .401321E-02 .333407 .739 Thai pineapple /Israelis citrus .283899E-02 .214116E-02 1.32591 .185 Thai Pineapple /Italian citrus .269542E-02 .210120E-02 1.28280 .200 Thai Pineapple /ROW citrus -.874562E-03 .140846E-02 -.620935 .535 Philippines Pineapple /ROW pineapple .878964E-03 .150772E-02 .582977 .560 Philippines Pineapple /U.S grape -.232261E-02 .352548E-02 -.658807 .510 Philippines Pineapple /Argentinean grape .506606E-02 .192841E-02 2.62707 .009 Philippines Pineapple /ROW grape .694251E-02 .395334E-02 1.75611 .079 Philippines Pineapple /Israelis other .477432E-02 .212146E-02 2.25049 .024 Philippines Pineapple /Italian other .162857E-02 .207686E-02 .784150 .433 Philippines Pineapple /ROW other .990120E-03 .138554E-02 .714608 .475 ROW Pineapple /U.S grape .111495E-03 .410181E-02 .027182 .978 ROW Pineapple /Argentinean grape .551961E-02 .175497E-02 3.14513 .002 ROW Pineapple /ROW grape -.688996E-04 .423046E-02 -.016287 .987 ROW Pineapple /Israelis citrus .226408E-02 .217128E-02 1.04274 .297 ROW Pineapple /Italian citrus .172359E-02 .189352E-02 .910254 .363 ROW Pineapple /ROW citrus -.241507E-02 .192703E-02 -1.25326 .210 U.S grape /Argentinean grape .010813 .432607E-02 2.49947 .012 U.S grape /ROW grape .816719E-02 .988500E-02 .826220 .409 U.S grape /Israelis citrus .513264E-04 .497369E-02 .010320 .992 U.S grape /Italian citrus .600603E-02 .461376E-02 1.30177 .193 U.S grape /ROW citrus .857491E-02 .395721E-02 2.16691 .030 Argentinean grape /ROW grape .235407E-02 .506562E-02 .464715 .642 Argentinean grape /Israelis citrus -.334719E-03 .278251E-02 -.120294 .904 Argentinean grape /Italian citrus -.261527E-02 .296975E-02 -.880638 .379 Argentinean grape /ROW citrus -.864642E-03 .157661E-02 -.548419 .583 ROW grape /Israelis citrus .454343E-02 .563118E-02 .806836 .420 ROW grape /Italian citrus -.451451E-03 .549765E-02 -.082117 .935 ROW grape /ROW citrus .495812E-02 .392222E-02 1.26411 .206
121
Table A-1 Continued Products Estimates SE t-statistics p-value Israelis citrus /Italian citrus .131769E-02 .300401E-02 .438645 .661 Israelis citrus /ROW citrus .245213E-02 .197170E-02 1.24366 .214 Italian citrus/ROW citrus -.102125E-02 .169281E-02 -.603289 .546
Table A-2 Slutsky price coefficients of fruit juices in Japan Products Estimates SE t-statistics p-valueU.S orange/Brazilian orange .082282 .025949 3.17097 .002 U.S orange/ROW orange .454009E-02 .756354E-02 .600261 .548 U.S orange/U.S grapefruit -.149260E-02 .012198 -.122367 .903 U.S orange/Israelis grapefruit .517986E-03 .666639E-02 .077701 .938 U.S orange/ROW grapefruit -.865542E-02 .408132E-02 -2.12074 .034 U.S orange/U.S apple .033827 .012814 2.63988 .008 U.S orange/Chinese apple .596499E-02 .011921 .500388 .617 U.S orange/ROW apple .023483 .020525 1.14412 .253 U.S orange/Thai pineapple .660704E-02 .453142E-02 1.45805 .145 U.S orange/Philippines pineapple .885034E-02 .443495E-02 1.99559 .046 U.S orange/ROW pineapple -.441460E-02 .556065E-02 -.793900 .427 U.S orange/U.S grapes .013827 .011784 1.17337 .241 U.S orange/Argentinean grapes -.765712E-02 .524138E-02 -1.46090 .144 U.S orange/ROW grapes -.017687 .012464 -1.41903 .156 U.S orange/Israelis -.015551 .624978E-02 -2.48818 .013 U.S orange/Italian other citrus -.221666E-02 .558774E-02 -.396701 .692 U.S orange/ROW other citrus -.010372 .528722E-02 -1.96173 .050 Brazilian orange/ROW orange -.017968 .012793 -1.40455 .160 Brazilian orange/U.S grapefruit .029763 .018410 1.61672 .106 Brazilian orange/Israelis grapefruit .485829E-02 .012708 .382292 .702 Brazilian orange/row grapefruit -.129330E-02 .754557E-02 -.171399 .864 Brazilian orange/U.S apple .013972 .026524 .526754 .598 Brazilian orange/Chinese apple -.010119 .020733 -.488079 .625 Brazilian orange/ROW apple -.075466 .035468 -2.12770 .033 Brazilian orange/Thai pineapple .674517E-03 .581514E-02 .115993 .908 Brazilian orange/Philippines pineapple -.226599E-02 .558795E-02 -.405514 .685 Brazilian orange/ROW pineapple .021105 .850128E-02 2.48251 .013 Brazilian orange/U.S grape -.991197E-02 .018119 -.547062 .584 Brazilian orange/Argentinean grape -.221559E-02 .630536E-02 -.351382 .725 Brazilian orange/ROW grape .026257 .016793 1.56364 .118 Brazilian orange/Israelis other citrus .013457 .809779E-02 1.66179 .097 Brazilian orange/Italian other citrus .970311E-02 .705267E-02 1.37581 .169 Brazilian orange/ROW other citrus -.370368E-02 .911090E-02 -.406512 .684 ROW orange/U.S grapefruit .474842E-02 .569236E-02 .834174 .404 ROW orange/Israelis grapefruit .292449E-02 .351147E-02 .832838 .405 ROW orange/ROW grapefruit .257776E-02 .207234E-02 1.24389 .214 ROW orange/U.S Apple .013279 .699220E-02 1.89916 .058 ROW orange/Chinese Apple .150242E-02 .599242E-02 .250721 .802 ROW orange/ROW Apple .022175 .010243 2.16488 .030 ROW orange/Thai pineapple .157489E-02 .200135E-02 .786913 .431 ROW orange/Philippines pineapple -.807215E-03 .195074E-02 -.413799 .679 ROW orange/ROW pineapple -.182784E-02 .269182E-02 -.679033 .497 ROW orange/U.S grape .869056E-03 .551943E-02 .157454 .875 ROW orange/Argentinean grape .366619E-02 .225791E-02 1.62371 .104 ROW orange/ROW grape .290152E-02 .557597E-02 .520361 .603 ROW orange/Israelis other citrus .226608E-02 .280284E-02 .808496 .419
122
Table A-2 Continued Products Estimates SE t-statistics p-value ROW orange/Italian other citrus .241610E-02 .243604E-02 .991814 .321 ROW orange/ROW other citrus .211060E-02 .268820E-02 .785135 .432 U.S grapefruit/Israelis grapefruit .320838E-02 .497839E-02 .644463 .519 U.S grapefruit/ROW grapefruit .010633 .309481E-02 3.43581 .001 U.S grapefruit/U.S Apple .026777 .963178E-02 2.78009 .005 U.S grapefruit/Chinese Apple .154646E-02 .905958E-02 .170699 .864 U.S grapefruit/ROW Apple -.636382E-02 .016073 -.395942 .692 U.S grapefruit/Thai pineapple -.018514 .353372E-02 -5.23931 .000 U.S grapefruit/Philippines pineapple -.014434 .347973E-02 -4.14795 .000 U.S grapefruit/ROW pineapple -.309955E-02 .422474E-02 -.733665 .463 U.S grapefruit/U.S grape -.015526 .910662E-02 -1.70493 .088 U.S grapefruit/Argentinean grape .272285E-02 .419966E-02 .648351 .517 U.S grapefruit/ROW grape .020940 .957948E-02 2.18594 .029 U.S grapefruit/Israelis other citrus .336212E-02 .490233E-02 .685820 .493 U.S grapefruit/Italian other citrus -.540124E-02 .446068E-02 -1.21086 .226 U.S grapefruit/R other citrus .474331E-02 .400508E-02 1.18432 .236 Israelis grapefruit/ROW grapefruit .222278E-02 .191813E-02 1.15882 .247 Israelis grapefruit/U.S apple -.562606E-02 .662787E-02 -.848850 .396 Israelis grapefruit/Chinese apple .735612E-02 .544864E-02 1.35009 .177 Israelis grapefruit/ROW apple -.409241E-02 .885256E-02 -.462286 .644 Israelis grapefruit/Thai pineapple -.531611E-03 .172118E-02 -.308865 .757 Israelis grapefruit/Philipp pineapple .217403E-02 .164367E-02 1.32267 .186 Israelis grapefruit/ROW pineapple -.162539E-03 .240061E-02 -.067707 .946 Israelis grapefruit/U.S grape -.214827E-02 .491086E-02 -.437453 .662 Israelis grapefruit/Argentinean grape .166068E-02 .185574E-02 .894890 .371 Israelis grapefruit/ROW grape -.395191E-03 .472069E-02 -.083715 .933 Israelis grapefruit/Israelis other citrus -.162520E-02 .235185E-02 -.691032 .490 Israelis grapefruit/Italian other citrus .467061E-02 .201697E-02 2.31565 .021 Israelis grapefruit/ROW other citrus -.802191E-03 .245760E-02 -.326413 .744 ROW grapefruit/U.S apple .894266E-03 .380753E-02 .234868 .814 ROW grapefruit/Chinese apple -.158872E-03 .336522E-02 -.047210 .962 ROW grapefruit/ROW apple .140690E-02 .566142E-02 .248506 .804 ROW grapefruit/Thai pineapple -.580800E-03 .110019E-02 -.527908 .598 ROW grapefruit/Philippines pineapple .260812E-04 .106917E-02 .024394 .981 ROW grapefruit/ROW pineapple .990709E-03 .147715E-02 .670690 .502 ROW grapefruit/U.S grape .273978E-02 .301673E-02 .908194 .364 ROW grapefruit/Argentinean grape .502918E-03 .123608E-02 .406864 .684 ROW grapefruit/ROW grape .292486E-02 .304144E-02 .961670 .336 ROW grapefruit/Israelis other citrus -.210905E-03 .153769E-02 -.137157 .891 ROW grapefruit/Italian other citrus -.435647E-02 .134108E-02 -3.24847 .001 ROW grapefruit/ROW other citrus -.177760E-02 .146947E-02 -1.20969 .226 U.S Apple/Chinese apple -.544442E-02 .010787 -.504716 .614 U.S Apple/ROW apple -.037772 .017674 -2.13718 .033 U.S Apple/Thai pineapple .345748E-02 .326233E-02 1.05982 .289 U.S Apple/Philippines pineapple .669691E-02 .316462E-02 2.11618 .034 U.S Apple/ROW pineapple -.014945 .467049E-02 -3.19985 .001 U.S Apple/U.S grape .010037 .949319E-02 1.05732 .290 U.S Apple/Argentinean grape -.017637 .357247E-02 -4.93679 .000 U.S Apple/ROW grape .459419E-03 .916707E-02 .050116 .960 U.S Apple/Israelis other citrus -.012315 .453846E-02 -2.71349 .007 U.S Apple/Italian other citrus .815445E-03 .382526E-02 .213174 .831 U.S Apple/ROW other citrus .010345 .479600E-02 2.15703 .031 Chinese Apple/ROW apple .022531 .016566 1.36008 .174
123
Table A-2 Continued Products Estimates SE t-statistics p-value Chinese Apple/Thai pineapple .263183E-02 .319510E-02 .823708 .410 Chinese Apple/Philipp pineapple .401380E-02 .306819E-02 1.30820 .191 Chinese Apple/ROW pineapple -.724174E-02 .428482E-02 -1.69009 .091 Chinese Apple/U.S grape .021882 .870380E-02 2.51403 .012 Chinese Apple/Argentinean grape -.107513E-02 .351270E-02 -.306070 .760 Chinese Apple/ROW grape -.012265 .878017E-02 -1.39689 .162 Chinese Apple/Israelis other citrus .109458E-02 .444353E-02 .246332 .805 Chinese Apple/Italian other citrus .263951E-02 .376621E-02 .700842 .483 Chinese Apple/ROW other citrus .499918E-02 .422676E-02 1.18274 .237 ROW Apple/Thai pineapple .745366E-02 .563993E-02 1.32159 .186 ROW Apple/Philippines pineapple .371444E-03 .555838E-02 .066826 .947 ROW Apple/ROW pineapple .849048E-02 .720091E-02 1.17909 .238 ROW Apple/U.S grape -.338917E-02 .015068 -.224920 .822 ROW Apple/Argentinean grape .721368E-02 .652161E-02 1.10612 .269 ROW Apple/ROW grape -.011404 .015899 -.717273 .473 ROW Apple/Israelis other citrus .013391 .791921E-02 1.69099 .091 ROW Apple/Italian other citrus .256592E-02 .699084E-02 .367040 .714 ROW Apple/ROW other citrus .610075E-02 .708296E-02 .861327 .389 Thai pineapple/Philippines pineapple .416619E-03 .152016E-02 .274062 .784 Thai pineapple /ROW pineapple -.617166E-03 .153899E-02 -.401021 .688 Thai pineapple /U.S grape .204054E-02 .356986E-02 .571603 .568 Thai pineapple /Argentinean grape -.133573E-02 .197277E-02 -.677082 .498 Thai pineapple /ROW grape .149119E-02 .401579E-02 .371332 .710 Thai pineapple /Israelis other citrus .287536E-02 .214106E-02 1.34296 .179 Thai pineapple /Italian other citrus .271276E-02 .210182E-02 1.29067 .197 Thai pineapple /ROW citrus -.822354E-03 .141002E-02 -.583223 .560 Philippines pineapple /ROW Pineapple .846751E-03 .150464E-02 .562759 .574 Philippines pineapple /U.S grape -.228425E-02 .352466E-02 -.648076 .517 Philippines pineapple /Argentinean grape .507384E-02 .192828E-02 2.63128 .009 Philippines pineapple /ROW grape .702281E-02 .395287E-02 1.77663 .076 Philippines pineapple /Israelis other .479275E-02 .211923E-02 2.26155 .024 Philippines pineapple /Italian other .163808E-02 .207549E-02 .789249 .430 Philippines pineapple /ROW other .101901E-02 .138595E-02 .735243 .462 ROW pineapple /U.S grape .336520E-05 .409543E-02 .008216 .999 ROW pineapple /Argentinean grape .549660E-02 .175303E-02 3.13549 .002 ROW pineapple /ROW grape -.322326E-03 .422837E-02 -.076229 .939 ROW pineapple /Israelis citrus .220419E-02 .216736E-02 1.01700 .309 ROW pineapple /Italian citrus .169638E-02 .189107E-02 .897043 .370 ROW pineapple /ROW citrus -.250214E-02 .192251E-02 -1.30150 .193 U.S grape /Argentinean grape .010838 .432789E-02 2.50415 .012 U.S grape /ROW grape .844149E-02 .988916E-02 .853610 .393 U.S grape /Israelis citrus .113882E-03 .497135E-02 .022908 .982 U.S grape /Italian citrus .603460E-02 .461373E-02 1.30797 .191 U.S grape /ROW citrus .867090E-02 .395775E-02 2.19087 .028 Argentinean grape /ROW grape .241619E-02 .506416E-02 .477115 .633 Argentinean grape /Israelis citrus -.319937E-03 .278183E-02 -.115010 .908 Argentinean grape /Italian citrus -.260801E-02 .296938E-02 -.878300 .380 Argentinean grape /ROW citrus -.844436E-03 .157830E-02 -.535029 .593 ROW grape /Israelis other citrus .470066E-02 .562912E-02 .835060 .404 ROW grape /Italian other citrus -.384002E-03 .549607E-02 -.069868 .944 ROW grape /ROW other citrus .517972E-02 .392754E-02 1.31882 .187 Israelis other citrus /Italian citrus .133399E-02 .300169E-02 .444412 .657 Israelis other citrus /ROW citrus .250478E-02 .197167E-02 1.27039 .204 Italian other citrus/ROW other citrus -.997712E-03 .169376E-02 -.589052 .556
124
APPENDIX B PRICE ELASTICITES OF FRUIT JUICES IN JAPAN
Table B-1 Uncompensated price elasticities of fruit juices in Japan Products Estimates SE t-statistics P-value U.S orange /Brazilian orange 1.01730 .363449 2.79902 .005 U.S orange /ROW orange .047577 .104560 .455020 .649 U.S orange /U.S grapefruit -.058235 .168667 -.345268 .730 U.S orange /Israelis grapefruit -.004943 .092568 -.053402 .957 U.S orange /ROW grapefruit -.124661 .056044 -2.22433 .026 U.S orange /U.S apple .440432 .175644 2.50753 .012 U.S orange /Chinese apple .048443 .165292 .293073 .769 U.S orange /ROW apple .247186 .288515 .856754 .392 U.S orange /Thai pineapple .086096 .062549 1.37645 .169 U.S orange /Philippines pineapple .118619 .061281 1.93566 .053 U.S orange /ROW pineapple -.065094 .076698 -.848701 .396 U.S orange /U.S grape .161903 .162868 .994076 .320 U.S orange /Argentinean grape -.109951 .072340 -1.51992 .129 U.S orange /ROW grape -.274281 .171684 -1.59759 .110 U.S orange /Israelis citrus -.224905 .086521 -2.59945 .009 U.S orange /Italian citrus -.038637 .077298 -.499848 .617 U.S orange /ROW citrus -.154814 .073022 -2.12010 .034 Brazilian orange/U.S orange .124232 .102147 1.21621 .224 Brazilian orange /ROW orange -.159867 .050603 -3.15923 .002 Brazilian orange /U.S grapefruit -.105466 .073526 -1.43440 .151 Brazilian orange /Israelis grapefruit -.052393 .050359 -1.04039 .298 Brazilian orange /ROW grapefruit -.035843 .029621 -1.21003 .226 Brazilian orange /U.S apple -.101359 .104006 -.974551 .330 Brazilian orange /Chinese apple -.240162 .082360 -2.91601 .004 Brazilian orange /ROW apple -.751689 .143362 -5.24330 .000 Brazilian orange /Thai Pineapple -.027458 .022913 -1.19838 .231 Brazilian orange /Philippines pineapple -.029791 .022038 -1.35177 .176 Brazilian orange /ROW Pineapple .058385 .033422 1.74692 .081 Brazilian orange /U.S grape -.210059 .072015 -2.91687 .004 Brazilian orange /Argentinean grape -.033961 .024833 -1.36754 .171 Brazilian orange /ROW grape -.075079 .066525 -1.12859 .259 Brazilian orange /Israelis citrus -.007861 .032109 -.244832 .807 Brazilian orange /Italian citrus -.009394 .027917 -.336525 .736 Brazilian orange /ROW citrus -.083522 .036088 -2.31443 .021 ROW orange/U.S orange .132507 .232275 .570473 .568 ROW orange/Brazilian orange -.581065 .398317 -1.45880 .145 ROW orange /U.S grapefruit .138055 .176274 .783184 .434 ROW orange /Israelis grapefruit .087521 .108918 .803547 .422 ROW orange /ROW grapefruit .078372 .063671 1.23090 .218 ROW orange /U.S apple .403814 .214469 1.88285 .060 ROW orange /Chinese apple .038737 .185700 .208600 .835 ROW orange /ROW apple .666951 .321627 2.07368 .038 ROW orange /Thai pineapple .047451 .061743 .768529 .442 ROW orange /Philippines pineapple -.025703 .060271 -.426454 .670
125
Table B-1 Continued Products Estimates SE t-statistics P-value ROW orange /ROW pineapple -.057339 .082988 -.690934 .490 ROW orange /U.S grape .020307 .170785 .118905 .905 ROW orange /Argentinean grape .112168 .069678 1.60980 .107 ROW orange /ROW grape .082746 .171975 .481150 .630 ROW orange /Israelis citrus .067612 .086753 .779363 .436 ROW orange /Italian citrus .072744 .075386 .964951 .335 ROW orange /ROW citrus .062504 .083084 .752293 .452 U.S grapefruit/U.S orange -.058049 .149991 -.387013 .699 U.S grapefruit/Brazilian orange .229191 .229876 .997020 .319 U.S grapefruit/ROW orange .041020 .070458 .582192 .560 U.S grapefruit/Israelis grapefruit .025485 .061894 .411754 .681 U.S grapefruit /ROW grapefruit .125402 .038093 3.29204 .001 U.S grapefruit /U.S apple .300139 .118405 2.53486 .011 U.S grapefruit /Chinese apple -.020650 .112550 -.183474 .854 U.S grapefruit /ROW apple -.169006 .202082 -.836322 .403 U.S grapefruit /Thai Pineapple -.234957 .043684 -5.37853 .000 U.S grapefruit /Philippines Pineapple -.182657 .043078 -4.24016 .000 U.S grapefruit /ROW Pineapple -.043223 .052210 -.827867 .408 U.S grapefruit /U.S grape -.225985 .112734 -2.00459 .045 U.S grapefruit /Argentinean grape .028663 .051931 .551957 .581 U.S grapefruit /ROW grape .223572 .118393 1.88839 .059 U.S grapefruit /Israelis citrus .029513 .060780 .485568 .627 U.S grapefruit /Italian citrus -.076243 .055296 -1.37881 .168 U.S grapefruit /ROW citrus .044975 .049553 .907601 .364 Israelis grapefruit/U.S orange .024507 .255321 .095986 .924 Israelis grapefruit/Brazilian orange .203036 .492739 .412056 .680 Israelis grapefruit/ROW orange .114612 .135514 .845760 .398 Israelis grapefruit/U.S grapefruit .128595 .192431 .668265 .504 Israelis grapefruit/ROW grapefruit .086263 .073575 1.17246 .241 Israelis grapefruit /U.S apple -.212975 .253414 -.840422 .401 Israelis grapefruit /Chinese apple .287740 .210750 1.36532 .172 Israelis grapefruit /ROW apple -.147102 .347287 -.423574 .672 Israelis grapefruit /Thai Pineapple -.019773 .066299 -.298233 .766 Israelis grapefruit /Philippines Pineapple .084160 .063374 1.32800 .184 Israelis grapefruit /ROW Pineapple -.005692 .092335 -.061647 .951 Israelis grapefruit /U.S grape -.078771 .189903 -.414796 .678 Israelis grapefruit /Argentinean grape .064501 .071452 .902717 .367 Israelis grapefruit /ROW grape -.011125 .181725 -.061218 .951 Israelis grapefruit /Israelis citrus -.061164 .090942 -.672561 .501 Israelis grapefruit /Italian citrus .180869 .077938 2.32068 .020 Israelis grapefruit /ROW citrus -.029298 .094840 -.308917 .757 ROW grapefruit/U.S orange -.807905 .363400 -2.22318 .026 ROW grapefruit/Brazilian orange -.232761 .682333 -.341125 .733 ROW grapefruit/ROW orange .215761 .185835 1.16103 .246 ROW grapefruit/U.S grapefruit .914322 .278100 3.28774 .001 ROW grapefruit/Israelis grapefruit .186953 .172509 1.08373 .278 ROW grapefruit /U.S apple .053883 .338594 .159136 .874 ROW grapefruit /Chinese apple -.047727 .302162 -.157950 .874 ROW grapefruit /ROW apple .049827 .517778 .096232 .923 ROW grapefruit /Thai Pineapple -.057011 .098391 -.579430 .562 ROW grapefruit /Philippines Pineapple -.001157 .095792 -.012087 .990 ROW grapefruit /ROW Pineapple .084538 .132043 .640227 .522 ROW grapefruit /U.S grape .216565 .271057 .798963 .424
126
Table B-1 Continued Products Estimates SE t-statistics P-value ROW grapefruit /Argentinean grape .040783 .110599 .368743 .712 ROW grapefruit /ROW grape .231909 .271878 .852989 .394 ROW grapefruit /Israelis citrus -.029041 .137973 -.210481 .833 ROW grapefruit /Italian citrus -.397805 .120468 -3.30217 .001 ROW grapefruit /ROW citrus -.170606 .131778 -1.29464 .195 U.S apple/U.S orange .535800 .224661 2.38492 .017 U.S apple/Brazilian orange .036199 .472513 .076609 .939 U.S apple/ROW orange .207066 .123565 1.67576 .094 U.S apple/U.S grapefruit .404743 .170654 2.37172 .018 U.S apple/Israelis grapefruit -.120498 .117512 -1.02541 .305 U.S apple/ROW grapefruit .006523 .066747 .097738 .922 U.S apple /Chinese apple -.155936 .190625 -.818024 .413 U.S apple /ROW apple -.801436 .318350 -2.51747 .012 U.S apple /Thai Pineapple .051848 .057452 .902448 .367 U.S apple /Philippines Pineapple .111653 .055828 1.99996 .046 U.S apple /ROW Pineapple -.270519 .082156 -3.29276 .001 U.S apple /U.S grape .125434 .168230 .745608 .456 U.S apple /Argentinean grape -.318097 .062956 -5.05272 .000 U.S apple /ROW grape -.045390 .161440 -.281157 .779 U.S apple /Israelis citrus -.235060 .080338 -2.92590 .003 U.S apple /Italian citrus .000097 .067568 .001437 .999 U.S apple /ROW citrus .161472 .084730 1.90572 .057 Chinese apple/U.S orange .034808 .162927 .213642 .831 Chinese apple/Brazilian orange -.304383 .287227 -1.05973 .289 Chinese apple/ROW orange -.000442 .082433 -.005364 .996 Chinese apple/U.S grapefruit -.031352 .124832 -.251156 .802 Chinese apple/Israelis grapefruit .084149 .075261 1.11809 .264 Chinese apple/ROW grapefruit -.009451 .046039 -.205288 .837 Chinese apple /U.S apple -.111732 .147318 -.758443 .448 Chinese apple /ROW apple .201989 .231144 .873868 .382 Chinese apple /Thai Pineapple .029036 .043876 .661773 .508 Chinese apple /Philippines Pineapple .050202 .042202 1.18955 .234 Chinese apple /ROW Pineapple -.105297 .058808 -1.79052 .073 Chinese apple /U.S grape .260147 .119876 2.17012 .030 Chinese apple /Argentinean grape -.020735 .048254 -.429710 .667 Chinese apple /ROW grape -.210633 .120544 -1.74735 .081 Chinese apple /Israelis citrus .000667 .061216 .010911 .991 Chinese apple /Italian citrus .025017 .051889 .482131 .630 Chinese apple /ROW citrus .052375 .058165 .900448 .368 ROW apple/U.S orange .107005 .123498 .866451 .386 ROW apple/Brazilian orange -.579753 .216202 -2.68154 .007 ROW apple/ROW orange .118483 .061943 1.91276 .056 ROW apple/U.S grapefruit -.077664 .097293 -.798248 .425 ROW apple/Israelis grapefruit -.037344 .053830 -.693737 .488 ROW apple/ROW grapefruit .003101 .034075 .091010 .927 ROW apple /U.S apple -.256057 .106199 -2.41111 .016 ROW apple /Chinese apple .101073 .100362 1.00709 .314 ROW apple /Thai Pineapple .039802 .034110 1.16687 .243 ROW apple /Philippines Pineapple -.001426 .033667 -.042359 .966 ROW apple /ROW Pineapple .047041 .043524 1.08082 .280 ROW apple /U.S grape -.050609 .091436 -.553490 .580 ROW apple /Argentinean grape .039206 .039469 .993355 .321 ROW apple /ROW grape -.100389 .096014 -1.04557 .296
127
Table B-1 Continued Products Estimates SE t-statistics P-value ROW apple /Israelis citrus .070331 .048041 1.46397 .143 ROW apple /Italian citrus .007157 .042412 .168754 .866 ROW apple /ROW citrus .024781 .042907 .577562 .564 Thai pineapple/U.S orange .574566 .411579 1.39600 .163 Thai pineapple/Brazilian orange -.041274 .538182 -.076691 .939 Thai pineapple /ROW orange .130829 .182919 .715228 .474 Thai pineapple /U.S grapefruit -1.72498 .323097 -5.33889 .000 Thai pineapple /Israelis grapefruit -.059108 .158017 -.374063 .708 Thai pineapple /ROW grapefruit -.057611 .100026 -.575956 .565 Thai pineapple /U.S apple .293028 .296411 .988590 .323 Thai pineapple /Chinese apple .211083 .293336 .719596 .472 Thai pineapple /ROW apple .614374 .524229 1.17196 .241 Thai pineapple /Philippines pineapple .035009 .139039 .251793 .801 Thai pineapple /ROW Pineapple -.060033 .140571 -.427063 .669 Thai pineapple /U.S grape .161346 .326045 .494859 .621 Thai pineapple /Argentinean GR -.125802 .180211 -.698084 .485 Thai pineapple /ROW grape .110063 .367130 .299792 .764 Thai pineapple /Israelis citrus .253872 .196042 1.29499 .195 Thai pineapple /Italian citrus .240957 .192482 1.25184 .211 Thai pineapple /ROW citrus -.085308 .128739 -.662640 .508 Philippines pineapple/U.S orange 1.14343 .580643 1.96926 .049 Philippines pineapple/Brazilian orange -.380386 .745731 -.510085 .610 Philippines pineapple /ROW orange -.116823 .257193 -.454221 .650 Philippines pineapple /U.S grapefruit -1.92872 .458748 -4.20432 .000 Philippines pineapple /Israelis grapefruit .278250 .217713 1.27805 .201 Philippines pineapple /ROW grapefruit -.000151 .140211 -.001080 .999 Philippines pineapple /U.S apple .864587 .414549 2.08561 .037 Philippines pineapple /Chinese apple .505740 .406232 1.24495 .213 Philippines pineapple /ROW apple -.004122 .744671 -.005536 .996 Philippines pineapple /Thai pineapple .051407 .200281 .256675 .797 Philippines pineapple /ROW pineapple .108750 .198210 .548660 .583 Philippines pineapple /U.S grape -.321091 .464205 -.691702 .489 Philippines pineapple /Argentinean grape .665921 .254076 2.62095 .009 Philippines pineapple /ROW grape .904976 .520934 1.73722 .082 Philippines pineapple /Israelis other citrus .624716 .279880 2.23209 .026 Philippines pineapple /Italian citrus .210391 .274132 .767480 .443 Philippines pineapple /ROW other citrus .126285 .182412 .692308 .489 ROW pineapple/U.S orange -.433511 .618055 -.701413 .483 ROW pineapple/Brazilian orange 2.56872 .960501 2.67435 .007 ROW pineapple /ROW orange -.177503 .301080 -.589554 .555 ROW pineapple /U.S grapefruit -.279599 .473241 -.590818 .555 ROW pineapple /Israelis grapefruit .003300 .269649 .012240 .990 ROW pineapple /ROW grapefruit .119955 .164301 .730094 .465 ROW pineapple /U.S apple -1.62332 .518702 -3.12958 .002 ROW pineapple /Chinese apple -.749191 .480858 -1.55803 .119 ROW pineapple /ROW apple 1.08545 .819207 1.32501 .185 ROW pineapple /Thai pineapple -.059935 .171895 -.348671 .727 ROW pineapple /Philippines pineapple .100901 .168355 .599335 .549 ROW pineapple /U.S grape .051734 .458327 .112875 .910 ROW pineapple /Argentinean grape .621882 .195892 3.17462 .002 ROW pineapple /ROW grape .017522 .472157 .037111 .970 ROW pineapple /Israelis citrus .264592 .242764 1.08991 .276 ROW pineapple /Italian citrus .203866 .211865 .962244 .336
128
Table B-1 Continued Products Estimates SE t-statistics P-value ROW pineapple /ROW citrus -.258939 .215039 -1.20415 .229 U.S grape/U.S orange .213019 .188475 1.13022 .258 U.S grape/Brazilian orange -.192560 .294080 -.654789 .513 U.S grape/ROW orange .009763 .088865 .109864 .913 U.S grape/U.S grapefruit -.260314 .146804 -1.77321 .076 U.S grape/Israelis grapefruit -.037944 .079435 -.477669 .633 U.S grape/ROW grapefruit .042624 .048300 .882471 .378 U.S grape/U.S apple .154090 .151838 1.01483 .310 U.S grape/Chinese apple .342560 .140657 2.43542 .015 U.S grape/ROW apple -.076039 .246697 -.308227 .758 U.S grape/Thai Pineapple .031405 .057409 .547029 .584 U.S grape /Philippines Pineapple -.037737 .056761 -.664837 .506 U.S grape /ROW Pineapple -.001110 .065844 -.016867 .987 U.S grape /Argentinean grape .173165 .069613 2.48753 .013 U.S grape /ROW grape .127373 .159132 .800425 .423 U.S grape /Israelis citrus -.001043 .080177 -.013009 .990 U.S grape /Italian citrus .094837 .074390 1.27486 .202 U.S grape /ROW citrus .136238 .063673 2.13966 .032 Argentinean grape/U.S orange -.848684 .567265 -1.49610 .135 Argentinean grape/Brazilian orange -.290384 .697277 -.416454 .677 Argentinean grape/ROW orange .393454 .246106 1.59872 .110 Argentinean grape/U.S grapefruit .281305 .458102 .614067 .539 Argentinean grape/Israelis grapefruit .176053 .203177 .866502 .386 Argentinean grape/ROW grapefruit .052680 .133995 .393148 .694 Argentinean grape/U.S apple -1.93361 .387091 -4.99523 .000 Argentinean grape/Chinese apple -.131195 .384502 -.341207 .733 Argentinean grape/ROW apple .754669 .722820 1.04406 .296 Argentinean grape/Thai pineapple -.147720 .214881 -.687450 .492 Argentinean grape /Philippines pineapple. .551683 .210311 2.62317 .009 Argentinean grape /ROW Pineapple .597509 .190995 3.12840 .002 Argentinean grape /U.S grape 1.16956 .471310 2.48151 .013 Argentinean grape /ROW grape .250957 .552739 .454025 .650 Argentinean grape /Israelis other citrus -.039123 .303658 -.128837 .897 Argentinean grape /Italian citrus -.287640 .324357 -.886799 .375 Argentinean grape /ROW other CT -.096872 .171753 -.564020 .573 ROW grape/U.S orange -.294039 .191043 -1.53912 .124 ROW grape/Brazilian orange .331122 .262422 1.26179 .207 ROW grape/ROW orange .035333 .086060 .410568 .681 ROW grape/U.S grapefruit .299574 .147939 2.02499 .043 ROW grape/Israelis grapefruit -.013666 .073207 -.186672 .852 ROW grape/ROW grapefruit .041878 .046684 .897053 .370 ROW grape/U.S apple -.009454 .140601 -.067243 .946 ROW grape/Chinese apple -.210464 .136026 -1.54724 .122 ROW grape/ROW apple -.224117 .249520 -.898191 .369 ROW grape/Thai Pineapple .019824 .061918 .320159 .749 ROW grape /Philippines Pineapple .106159 .061047 1.73898 .082 ROW grape /ROW Pineapple -.007580 .065207 -.116248 .907 ROW grape /U.S grape .112155 .152514 .735374 .462 ROW grape /Argentinean grape .034612 .078101 .443174 .658 ROW grape /Israelis citrus .066102 .087016 .759649 .447 ROW grape /Italian citrus -.010959 .084977 -.128966 .897 ROW grape /ROW citrus .072630 .060573 1.19905 .231 Israelis citrus/U.S orange -.718991 .281024 -2.55847 .011
129
Table B-1 Continued Products Estimates SE t-statistics P-value Israelis citrus /Brazilian orange .556726 .371001 1.50060 .133 Israelis citrus /ROW orange .095900 .126939 .755481 .450 Israelis citrus /U.S grapefruit .135515 .222058 .610266 .542 Israelis citrus /Israelis grapefruit -.078944 .107006 -.737755 .461 Israelis citrus /ROW grapefruit -.011856 .069276 -.171146 .864 Israelis citrus /U.S apple -.569275 .204169 -2.78825 .005 Israelis citrus /Chinese apple .034520 .201885 .170988 .864 Israelis citrus /ROW apple .572135 .364275 1.57061 .116 Israelis citrus /Thai pineapple .127918 .096861 1.32064 .187 Israelis citrus /Philippines pineapple .215417 .096017 2.24353 .025 Israelis citrus /ROW Pineapple .097943 .098067 .998735 .318 Israelis citrus /U.S grape -.007681 .224844 -.034165 .973 Israelis citrus /Argentinean GR -.016379 .125873 -.130125 .896 Israelis citrus /ROW grape .199430 .254829 .782603 .434 Israelis citrus /Italian citrus .056825 .136156 .417351 .676 Israelis citrus /ROW citrus .108225 .089128 1.21427 .225 Italian citrus/U.S orange -.136630 .321111 -.425493 .670 Italian citrus /Brazilian orange .532182 .412199 1.29108 .197 Italian citrus /ROW orange .136078 .140906 .965732 .334 Italian citrus /U.S grapefruit -.321885 .258107 -1.24710 .212 Italian citrus /Israelis grapefruit .267284 .117198 2.28061 .023 Italian citrus /ROW grapefruit -.253390 .077159 -3.28401 .001 Italian citrus /U.S apple .040638 .220069 .184662 .853 Italian citrus /Chinese apple .144348 .218866 .659528 .510 Italian citrus /ROW apple .129425 .411250 .314712 .753 Italian citrus /Thai pineapple .155721 .121520 1.28145 .200 Italian citrus /Philippines pineapple .093918 .120168 .781555 .434 Italian citrus /ROW Pineapple .097134 .109392 .887950 .375 Italian citrus /U.S grape .342044 .266627 1.28285 .200 Italian citrus /Argentinean grape -.151979 .171695 -.885169 .376 Italian citrus /ROW grape -.029695 .318080 -.093358 .926 Italian citrus /Israelis citrus .074648 .173907 .429241 .668 Italian citrus /ROW citrus -.060625 .097837 -.619658 .535 ROW citrus/U.S orange -.432676 .210033 -2.06003 .039 ROW citrus /Brazilian orange -.213379 .366838 -.581672 .561 ROW citrus /ROW orange .075887 .107417 .706476 .480 ROW citrus /U.S grapefruit .168478 .160409 1.05030 .294 ROW citrus /Israelis grapefruit -.038717 .098598 -.392679 .695 ROW citrus /ROW grapefruit -.073833 .058395 -1.26437 .206 ROW citrus /U.S apple .398274 .190262 2.09330 .036 ROW citrus /Chinese apple .180769 .169474 1.06665 .286 ROW citrus /ROW apple .200896 .288067 .697392 .486 ROW citrus /Thai pineapple -.035644 .056262 -.633539 .526 ROW citrus /Philippines pineapple .038717 .055391 .698982 .485 ROW citrus /ROW Pineapple -.102178 .076658 -1.33290 .183 ROW citrus /U.S grape .330066 .158350 2.08441 .037 ROW citrus /Argentinean GR -.036070 .063003 -.572516 .567 ROW citrus /ROW grape .190036 .156688 1.21283 .225 ROW citrus /Israelis citrus .094281 .078957 1.19408 .232 ROW citrus /Italian citrus -.044278 .067806 -.653013 .514
130
Table B-2 Compensated price elasticities of fruit juices in Japan Products Estimates SE t-statistics P-value U.S orange /Brazilian orange 1.13564 .358135 3.17097 .002 U.S orange /ROW orange .062661 .104390 .600261 .548 U.S orange /U.S grapefruit -.020600 .168350 -.122367 .903 U.S orange /Israelis grapefruit .007149 .092007 .077701 .938 U.S orange /ROW grapefruit -.119459 .056329 -2.12074 .034 U.S orange /U.S apple .466868 .176852 2.63988 .008 U.S orange /Chinese apple .082327 .164526 .500388 .617 U.S orange /ROW apple .324111 .283284 1.14412 .253 U.S orange /Thai Pineapple .091188 .062541 1.45805 .145 U.S orange /Philippines Pineapple .122150 .061210 1.99559 .046 U.S orange /ROW Pineapple -.060929 .076746 -.793900 .427 U.S orange /U.S grape .190834 .162638 1.17337 .241 U.S orange /Argentinean grape -.105681 .072340 -1.46090 .144 U.S orange /ROW grape -.244117 .172030 -1.41903 .156 U.S orange /Israelis citrus -.214624 .086258 -2.48818 .013 U.S orange /Italian citrus -.030594 .077120 -.396701 .692 U.S orange /ROW citrus -.143153 .072973 -1.96173 .050 Brazilian orange/U.S orange .323646 .102065 3.17097 .002 Brazilian orange /ROW orange -.070674 .050318 -1.40455 .160 Brazilian orange /U.S grapefruit .117070 .072412 1.61672 .106 Brazilian orange /Israelis grapefruit .019109 .049986 .382292 .702 Brazilian orange /ROW grapefruit -.005087 .029679 -.171399 .864 Brazilian orange /U.S apple .054956 .104329 .526754 .598 Brazilian orange /Chinese apple -.039803 .081551 -.488079 .625 Brazilian orange /ROW apple -.296834 .139509 -2.12770 .033 Brazilian orange /Thai Pineapple .002653 .022873 .115993 .908 Brazilian orange /Philippines pineapple -.008912 .021979 -.405514 .685 Brazilian orange /ROW pineapple .083012 .033439 2.48251 .013 Brazilian orange /U.S grape -.038987 .071267 -.547062 .584 Brazilian orange /Argentinean grape -.008714 .024801 -.351382 .725 Brazilian orange /ROW grape .103280 .066051 1.56364 .118 Brazilian orange /Israelis citrus .052930 .031851 1.66179 .097 Brazilian orange /Italian citrus .038166 .027741 1.37581 .169 Brazilian orange /ROW citrus -.014568 .035836 -.406512 .684 ROW orange/U.S orange .140095 .233390 .600261 .548 ROW orange/Brazilian orange -.554440 .394745 -1.40455 .160 ROW orange /U.S grapefruit .146523 .175650 .834174 .404 ROW orange /Israelis grapefruit .090241 .108354 .832838 .405 ROW orange /ROW grapefruit .079543 .063947 1.24389 .214 ROW orange /U.S apple .409762 .215760 1.89916 .058 ROW orange /Chinese apple .046361 .184909 .250721 .802 ROW orange /ROW apple .684259 .316073 2.16488 .030 ROW orange /Thai Pineapple .048597 .061756 .786913 .431 ROW orange /Philippines pineapple -.024908 .060194 -.413799 .679 ROW orange /ROW pineapple -.056402 .083062 -.679033 .497 ROW orange /U.S grape .026817 .170314 .157454 .875 ROW orange /Argentinean grape .113128 .069673 1.62371 .104 ROW orange /ROW grape .089533 .172059 .520361 .603 ROW orange /Israelis citrus .069925 .086488 .808496 .419 ROW orange /Italian citrus .074554 .075169 .991814 .321 ROW orange /ROW citrus .065127 .082950 .785135 .432 U.S grapefruit/U.S orange -.018460 .150858 -.122367 .903 U.S grapefruit/Brazilian orange .368103 .227684 1.61672 .106
131
Table B-2 Continued Products Estimates SE t-statistics P-value U.S grapefruit/ROW orange .058727 .070401 .834174 .404 U.S grapefruit/Israelis grapefruit .039680 .061571 .644463 .519 U.S grapefruit /ROW grapefruit .131508 .038276 3.43581 .001 U.S grapefruit /U.S apple .331172 .119123 2.78009 .005 U.S grapefruit /Chinese apple .019126 .112046 .170699 .864 U.S grapefruit /ROW apple -.078706 .198781 -.395942 .692 U.S grapefruit /Thai Pineapple -.228979 .043704 -5.23931 .000 U.S grapefruit /Philippines Pineapple -.178512 .043036 -4.14795 .000 U.S grapefruit /ROW Pineapple -.038334 .052250 -.733665 .463 U.S grapefruit /U.S grape -.192023 .112628 -1.70493 .088 U.S grapefruit /Argentinean grape .033675 .051940 .648351 .517 U.S grapefruit /ROW grape .258981 .118476 2.18594 .029 U.S grapefruit /Israelis citrus .041582 .060631 .685820 .493 U.S grapefruit /Italian citrus -.066801 .055168 -1.21086 .226 U.S grapefruit /ROW citrus .058664 .049534 1.18432 .236 Israelis grapefruit/U.S orange .019938 .256600 .077701 .938 Israelis grapefruit/Brazilian orange .187004 .489164 .382292 .702 Israelis grapefruit/ROW orange .112568 .135162 .832838 .405 Israelis grapefruit/U.S grapefruit .123496 .191626 .644463 .519 Israelis grapefruit/ROW grapefruit .085558 .073832 1.15882 .247 Israelis grapefruit /U.S apple -.216557 .255118 -.848850 .396 Israelis grapefruit /Chinese apple .283150 .209727 1.35009 .177 Israelis grapefruit /ROW apple -.157524 .340750 -.462286 .644 Israelis grapefruit /Thai Pineapple -.020463 .066251 -.308865 .757 Israelis grapefruit /Philippines Pineapple .083682 .063267 1.32267 .186 Israelis grapefruit /ROW Pineapple -.006256 .092404 -.067707 .946 Israelis grapefruit /U.S grape -.082691 .189027 -.437453 .662 Israelis grapefruit /Argentinean grape .063922 .071431 .894890 .371 Israelis grapefruit /ROW grape -.015212 .181707 -.083715 .933 Israelis grapefruit /Israelis citrus -.062557 .090526 -.691032 .490 Israelis grapefruit /Italian citrus .179780 .077637 2.31565 .021 Israelis grapefruit /ROW citrus -.030878 .094597 -.326413 .744 ROW grapefruit/U.S orange -.774553 .365228 -2.12074 .034 ROW grapefruit/Brazilian orange -.115735 .675236 -.171399 .864 ROW grapefruit/ROW orange .230678 .185449 1.24389 .214 ROW grapefruit/U.S grapefruit .951541 .276948 3.43581 .001 ROW grapefruit/Israelis grapefruit .198911 .171649 1.15882 .247 ROW grapefruit /U.S apple .080026 .340727 .234868 .814 ROW grapefruit /Chinese apple -.014217 .301146 -.047210 .962 ROW grapefruit /ROW apple .125900 .506628 .248506 .804 ROW grapefruit /Thai Pineapple -.051974 .098454 -.527908 .598 ROW grapefruit /Philippines Pineapple .002333 .095678 .024394 .981 ROW grapefruit /ROW Pineapple .088656 .132187 .670690 .502 ROW grapefruit /U.S grape .245176 .269960 .908194 .364 ROW grapefruit /Argentinean grape .045005 .110614 .406864 .684 ROW grapefruit /ROW grape .261739 .272171 .961670 .336 ROW grapefruit /Israelis citrus -.018873 .137604 -.137157 .891 ROW grapefruit /Italian citrus -.389851 .120010 -3.24847 .001 ROW grapefruit /ROW citrus -.159073 .131499 -1.20969 .226 U.S apple/U.S orange .595592 .225613 2.63988 .008 U.S apple/Brazilian orange .246002 .467014 .526754 .598 U.S apple/ROW orange .233809 .123112 1.89916 .058 U.S apple/U.S grapefruit .471467 .169587 2.78009 .005
132
Table B-2 Continued Products Estimates SE t-statistics P-value U.S apple/Israelis grapefruit -.099058 .116697 -.848850 .396 U.S apple/ROW grapefruit .015745 .067039 .234868 .814 U.S apple /Chinese apple -.095860 .189929 -.504716 .614 U.S apple /ROW apple -.665053 .311183 -2.13718 .033 U.S apple /Thai Pineapple .060876 .057440 1.05982 .289 U.S apple /Philippines Pineapple .117913 .055720 2.11618 .034 U.S apple /ROW Pineapple -.263135 .082234 -3.19985 .001 U.S apple /U.S grape .176727 .167147 1.05732 .290 U.S apple /Argentinean grape -.310527 .062901 -4.93679 .000 U.S apple /ROW grape .008089 .161405 .050116 .960 U.S apple /Israelis citrus -.216832 .079909 -2.71349 .007 U.S apple /Italian citrus .014358 .067351 .213174 .831 U.S apple /ROW citrus .182147 .084443 2.15703 .031 Chinese apple/U.S orange .081939 .163750 .500388 .617 Chinese apple/Brazilian orange -.139007 .284804 -.488079 .625 Chinese apple/ROW orange .020638 .082315 .250721 .802 Chinese apple/U.S grapefruit .021243 .124448 .170699 .864 Chinese apple/Israelis grapefruit .101048 .074846 1.35009 .177 Chinese apple/ROW grapefruit -.002182 .046227 -.047210 .962 Chinese apple /U.S apple -.074788 .148178 -.504716 .614 Chinese apple /ROW apple .309493 .227555 1.36008 .174 Chinese apple /Thai Pineapple .036152 .043890 .823708 .410 Chinese apple /Philippines Pineapple .055136 .042147 1.30820 .191 Chinese apple /ROW Pineapple -.099477 .058859 -1.69009 .091 Chinese apple /U.S grape .300579 .119561 2.51403 .012 Chinese apple /Argentinean grape -.014769 .048253 -.306070 .760 Chinese apple /ROW grape -.168478 .120610 -1.39689 .162 Chinese apple /Israelis citrus .015036 .061039 .246332 .805 Chinese apple /Italian citrus .036258 .051735 .700842 .483 Chinese apple /ROW citrus .068672 .058061 1.18274 .237 ROW apple/U.S orange .142094 .124195 1.14412 .253 ROW apple/Brazilian orange -.456630 .214612 -2.12770 .033 ROW apple/ROW orange .134177 .061979 2.16488 .030 ROW apple/U.S grapefruit -.038506 .097252 -.395942 .692 ROW apple/Israelis grapefruit -.024762 .053565 -.462286 .644 ROW apple/ROW grapefruit .008512 .034256 .248506 .804 ROW apple /U.S apple -.228551 .106941 -2.13718 .033 ROW apple /Chinese apple .136328 .100236 1.36008 .174 ROW apple /Thai Pineapple .045101 .034126 1.32159 .186 ROW apple /Philippines Pineapple .002247 .033633 .066826 .947 ROW apple /ROW Pineapple .051374 .043571 1.17909 .238 ROW apple /U.S grape -.020507 .091176 -.224920 .822 ROW apple /Argentinean grape .043649 .039461 1.10612 .269 ROW apple /ROW grape -.069005 .096204 -.717273 .473 ROW apple /Israelis citrus .081028 .047918 1.69099 .091 ROW apple /Italian citrus .015526 .042300 .367040 .714 ROW apple /ROW citrus .036915 .042858 .861327 .389 Thai Pineapple/U.S orange .603899 .414183 1.45805 .145 Thai Pineapple/Brazilian orange .061652 .531517 .115993 .908 Thai Pineapple /ROW orange .143949 .182928 .786913 .431 Thai Pineapple /U.S grapefruit -1.69225 .322991 -5.23931 .000 Thai Pineapple /Israelis grapefruit -.048590 .157320 -.308865 .757 Thai Pineapple /ROW grapefruit -.053086 .100560 -.527908 .598
133
Table B-2 Continued Products Estimates SE t-statistics P-value Thai Pineapple /U.S apple .316022 .298184 1.05982 .289 Thai Pineapple /Chinese apple .240555 .292040 .823708 .410 Thai Pineapple /ROW apple .681282 .515502 1.32159 .186 Thai Pineapple /Philippines pineapple .038080 .138946 .274062 .784 Thai Pineapple /ROW Pineapple -.056410 .140667 -.401021 .688 Thai Pineapple /U.S grape .186510 .326293 .571603 .568 Thai pineapple /Argentinean grape -.122089 .180316 -.677082 .498 Thai pineapple /ROW grape .136298 .367053 .371332 .710 Thai pineapple /Israelis citrus .262815 .195698 1.34296 .179 Thai pineapple /Italian citrus .247953 .192111 1.29067 .197 Thai pineapple /ROW citrus -.075165 .128879 -.583223 .560 Philippines pineapple/U.S orange 1.16671 .584644 1.99559 .046 Philippines pineapple/Brazilian orange -.298718 .736640 -.405514 .685 Philippines pineapple /ROW orange -.106412 .257159 -.413799 .679 Philippines pineapple /U.S grapefruit -1.90275 .458721 -4.14795 .000 Philippines pineapple /Israelis grapefruit .286595 .216679 1.32267 .186 Philippines pineapple /ROW grapefruit .003438 .140945 .024394 .981 Philippines pineapple /U.S apple .882831 .417181 2.11618 .034 Philippines pineapple /Chinese apple .529125 .404469 1.30820 .191 Philippines pineapple /ROW apple .048966 .732742 .066826 .947 Philippines pineapple /Thai pineapple .054921 .200398 .274062 .784 Philippines pineapple /ROW pineapple .111624 .198352 .562759 .574 Philippines pineapple /U.S grape -.301125 .464644 -.648076 .517 Philippines pineapple /Argentinean grape .668868 .254198 2.63128 .009 Philippines pineapple /ROW grape .925793 .521094 1.77663 .076 Philippines pineapple /Israelis other citrus .631812 .279371 2.26155 .024 Philippines pineapple /Italian citrus .215942 .273604 .789249 .430 Philippines pineapple /ROW other citrus .134333 .182706 .735243 .462 ROW pineapple/U.S orange -.493378 .621461 -.793900 .427 ROW pineapple/Brazilian orange 2.35865 .950108 2.48251 .013 ROW pineapple /ROW orange -.204280 .300840 -.679033 .497 ROW pineapple /U.S grapefruit -.346407 .472159 -.733665 .463 ROW pineapple /Israelis grapefruit -.018165 .268294 -.067707 .946 ROW pineapple /ROW grapefruit .110722 .165087 .670690 .502 ROW pineapple /U.S apple -1.67025 .521976 -3.19985 .001 ROW pineapple /Chinese apple -.809341 .478874 -1.69009 .091 ROW pineapple /ROW apple .948901 .804777 1.17909 .238 ROW pineapple /Thai pineapple -.068975 .171998 -.401021 .688 ROW pineapple /Philippines pineapple .094633 .168160 .562759 .574 ROW pineapple /U.S grape .000376 .457708 .000821 .999 ROW pineapple /Argentinean grape .614303 .195919 3.13549 .002 ROW pineapple /ROW grape -.036023 .472565 -.076229 .939 ROW pineapple /Israelis citrus .246342 .242225 1.01700 .309 ROW pineapple /Italian citrus .189588 .211347 .897043 .370 ROW pineapple /ROW citrus -.279640 .214860 -1.30150 .193 U.S grape/U.S orange .222450 .189582 1.17337 .241 U.S grape/Brazilian orange -.159466 .291496 -.547062 .584 U.S grape/ROW orange .013982 .088798 .157454 .875 U.S grape/U.S grapefruit -.249789 .146509 -1.70493 .088 U.S grape/Israelis grapefruit -.034562 .079007 -.437453 .662 U.S grape/ROW grapefruit .044078 .048534 .908194 .364 U.S grape/U.S apple .161483 .152729 1.05732 .290 U.S grape/Chinese apple .352037 .140029 2.51403 .012
134
Table B-2 Continued Products Estimates SE t-statistics P-value U.S grape/ROW apple -.054526 .242424 -.224920 .822 U.S grape/Thai Pineapple .032829 .057433 .571603 .568 U.S grape /Philippines Pineapple -.036750 .056706 -.648076 .517 U.S grape /ROW Pineapple .005414 .065888 .000821 .999 U.S grape /Argentinean grape .174359 .069628 2.50415 .012 U.S grape /ROW grape .135809 .159099 .853610 .393 U.S grape /Israelis citrus .001832 .079980 .022908 .982 U.S grape /Italian citrus .097086 .074227 1.30797 .191 U.S grape /ROW citrus .139500 .063673 2.19087 .028 Argentinean grape/U.S orange -.834763 .571405 -1.46090 .144 Argentinean grape/Brazilian orange -.241539 .687398 -.351382 .725 Argentinean grape/ROW orange .399680 .246153 1.62371 .104 Argentinean grape/U.S grapefruit .296840 .457838 .648351 .517 Argentinean grape/Israelis grapefruit .181044 .202309 .894890 .371 Argentinean grape/ROW grapefruit .054827 .134755 .406864 .684 Argentinean grape/U.S apple -1.92270 .389463 -4.93679 .000 Argentinean grape/Chinese apple -.117209 .382947 -.306070 .760 Argentinean grape/ROW apple .786420 .710972 1.10612 .269 Argentinean grape/Thai pineapple -.145618 .215067 -.677082 .498 Argentinean grape /Philippines pineapple. .553140 .210217 2.63128 .009 Argentinean grape /ROW Pineapple .599228 .191111 3.13549 .002 Argentinean grape /U.S grape 1.18150 .471817 2.50415 .012 Argentinean grape /ROW grape .263408 .552084 .477115 .633 Argentinean grape /Israelis other citrus -.034879 .303269 -.115010 .908 Argentinean grape /Italian citrus -.284320 .323716 -.878300 .380 Argentinean grape /ROW other citrus -.092059 .172063 -.535029 .593 ROW grape/U.S orange -.272934 .192338 -1.41903 .156 ROW grape/Brazilian orange .405177 .259125 1.56364 .118 ROW grape/ROW orange .044773 .086042 .520361 .603 ROW grape/U.S grapefruit .323126 .147820 2.18594 .029 ROW grape/Israelis grapefruit -.006098 .072845 -.083715 .933 ROW grape/ROW grapefruit .045133 .046932 .961670 .336 ROW grape/U.S apple .007089 .141456 .050116 .960 ROW grape/Chinese apple -.189259 .135486 -1.39689 .162 ROW grape/ROW apple -.175977 .245342 -.717273 .473 ROW grape/Thai pineapple .023010 .061967 .371332 .710 ROW grape /Philippines pineapple .108369 .060997 1.77663 .076 ROW grape /ROW pineapple -.004973 .065248 -.076229 .939 ROW grape /U.S grape .130260 .152599 .853610 .393 ROW grape /Argentinean grape .037284 .078145 .477115 .633 ROW grape /Israelis citrus .072536 .086863 .835060 .404 ROW grape /Italian citrus -.005925 .084810 -.069868 .944 ROW grape /ROW citrus .079928 .060606 1.31882 .187 Israelis citrus/U.S orange -.704027 .282948 -2.48818 .013 Israelis citrus /Brazilian orange .609234 .366614 1.66179 .097 Israelis citrus /ROW orange .102593 .126894 .808496 .419 Israelis citrus /U.S grapefruit .152214 .221945 .685820 .493 Israelis citrus /Israelis grapefruit -.073578 .106476 -.691032 .490 Israelis citrus /ROW grapefruit -.009548 .069616 -.137157 .891 Israelis citrus /U.S apple -.557545 .205471 -2.71349 .007 Israelis citrus /Chinese apple .049555 .201173 .246332 .805 Israelis citrus /ROW apple .606269 .358529 1.69099 .091 Israelis citrus /Thai pineapple .130177 .096933 1.34296 .179
135
Table B-2 Continued Products Estimates SE t-statistics P-value Israelis citrus /Philippines pineapple .216984 .095945 2.26155 .024 Israelis citrus /ROW Pineapple .099791 .098123 1.01700 .309 Israelis citrus /U.S grape .005155 .225070 .022908 .982 Israelis citrus /Argentinean grape -.014485 .125943 -.115010 .908 Israelis citrus /ROW grape .212814 .254849 .835060 .404 Israelis citrus /Italian citrus .060394 .135897 .444412 .657 Israelis citrus /ROW citrus .113400 .089264 1.27039 .204 Italian citrus/U.S orange -.128274 .323353 -.396701 .692 Italian citrus /Brazilian orange .561502 .408126 1.37581 .169 Italian citrus /ROW orange .139815 .140969 .991814 .321 Italian citrus /U.S grapefruit -.312561 .258132 -1.21086 .226 Italian citrus /Israelis grapefruit .270280 .116719 2.31565 .021 Italian citrus /ROW grapefruit -.252101 .077606 -3.24847 .001 Italian citrus /U.S apple .047188 .221361 .213174 .831 Italian citrus /Chinese apple .152744 .217944 .700842 .483 Italian citrus /ROW apple .148485 .404548 .367040 .714 Italian citrus /Thai pineapple .156983 .121628 1.29067 .197 Italian citrus /Philippines pineapple .094793 .120105 .789249 .430 Italian citrus /ROW pineapple .098166 .109433 .897043 .370 Italian citrus /U.S grape .349212 .266988 1.30797 .191 Italian citrus /Argentinean grape -.150921 .171833 -.878300 .380 Italian citrus /ROW grape -.022222 .318048 -.069868 .944 Italian citrus /Israelis citrus .077196 .173703 .444412 .657 Italian citrus /ROW citrus -.057736 .098015 -.589052 .556 ROW citrus/U.S orange -.413995 .211035 -1.96173 .050 ROW citrus /Brazilian orange -.147830 .363654 -.406512 .684 ROW citrus /ROW orange .084243 .107297 .785135 .432 ROW citrus /U.S grapefruit .189325 .159859 1.18432 .236 ROW citrus /Israelis grapefruit -.032019 .098093 -.326413 .744 ROW citrus /ROW grapefruit -.070952 .058653 -1.20969 .226 ROW citrus /U.S apple .412918 .191429 2.15703 .031 ROW citrus /Chinese apple .199538 .168708 1.18274 .237 ROW citrus /ROW apple .243507 .282711 .861327 .389 ROW citrus /Thai pineapple -.032824 .056280 -.583223 .560 ROW citrus /Philippines pineapple .040673 .055319 .735243 .462 ROW citrus /ROW Pineapple -.099871 .076735 -1.30150 .193 ROW citrus /U.S grape .346092 .157971 2.19087 .028 ROW citrus /Argentinean grape -.033705 .062997 -.535029 .593 ROW citrus /ROW grape .206745 .156765 1.31882 .187 ROW citrus /Israelis citrus .099976 .078698 1.27039 .204 ROW citrus /Italian citrus -.039823 .067605 -.589052 .556
136
APPENDIX C PARAMETER ESTIMATES OF ROTTERDAM MODEL UNDER DIFFERENT
SEPARABILITY ASSUMPTIONS
Table C-1. Marginal value shares of fruit juices in a block independent Rotterdam model Product Estimates ( )iθ SE t-statistics P-value
U.S. orange .027961 .913834E-02 3.05978 0.002 Brazilian orange .745750 .034906 21.3644 0.000 ROW orange .010748 .558347E-02 1.92496 0.054 U.S. grapefruit .029762 .580410E-02 5.12771 0.000 Israelis grapefruit .236792E-02 .298566E-02 .793096 0.428 ROW grapefruit .517695E-03 .169452E-02 .305512 0.760 U.S. apple .040278 .944699E-02 4.26357 0.000 Chinese apple .036968 .773508E-02 4.77923 0.000 ROW apple .047043 .012962 3.62937 0.000 Thai pineapple .460872E-02 .148641E-02 3.10056 0.002 Philippines pineapple .786131E-02 .156078E-02 5.03680 0.000 ROW pineapple .853774E-03 .178778E-02 .477560 0.633 U.S. grapes .880780E-02 .497279E-02 1.77120 0.077 Argentinean grapes -.846428E-04 .185156E-02 -.045714 0.964 ROW grapes .011431 .471108E-02 2.42647 0.015 Israelis other citrus .715117E-02 .224853E-02 3.18038 0.001 Italian other citrus .894720E-02 .227572E-02 3.93159 0.000 ROW other citrus .902818E-02 .217923E-02 4.14283 0.000 Coefficient of income flexibility ( )φ
-1.81278 .279739 -6.48024 0.000
Table C-2. Relative price coefficients of fruit juices in a block independent Rotterdam
model Products Estimates ( )ijν SE t-statistics P-value
U.S. orange -.114167 .020919 -5.45760 0.000 U.S. orange/Brazilian orange .057930 .026494 2.18651 0.029 U.S. orange/ROW orange .554920E-02 .682128E-02 .813514 0.416 Brazilian orange -1.42189 .247776 -5.73861 0.000 Brazilian orange/ROW orange .012078 .012434 .971341 0.331 ROW orange -.037110 .486140E-02 -7.63370 0.000 U.S. grapefruit -.067859 .010130 -6.69859 0.000 U.S. grapefruit/Israelis grapefruit .810086E-02 .440660E-02 1.83835 0.066 U.S. grapefruit/ROW grapefruit .580657E-02 .257851E-02 2.25191 0.024 Israelis grapefruit -.015038 .395186E-02 -3.80540 0.000 Israelis. grapefruit/ROW grapefruit .264504E-02 .160837E-02 1.64455 0.100 ROW grapefruit -.939007E-02 .130989E-02 -7.16862 0.000 U.S. apple -.023643 .013639 -1.73356 0.083 U.S. apple/Chinese apple -.998880E-02 .897542E-02 -1.11291 0.266 U.S. apple/ROW apple -.039383 .014233 -2.76692 0.006 Chinese apple -.041321 .012269 -3.36804 0.001
137
Table C-2. Continued Products Estimates ( )ijν SE t-statistics P-value
Chinese apple/ROW apple -.015704 .013935 -1.12695 0.260 ROW apple -.030191 .026148 -1.15461 0.248 Thai. pineapple -.010102 .186886E-02 -5.40530 0.000 Thai. pineapple/Philippines pineapple .229312E-02 .142989E-02 1.60371 0.109 Thai. pineapple/ROW pineapple -.545936E-03 .143880E-02 -.379438 0.704 Philippines pineapple -.018299 .189011E-02 -9.68162 0.000 Philippines pineapple/ROW apple .175536E-02 .145313E-02 1.20799 0.227 ROW pineapple -.275713E-02 .248249E-02 -1.11063 0.267 U.S. grapes -.038038 .010333 -3.68119 0.000 U.S. grapes/Argentinean grapes .714373E-02 .428626E-02 1.66666 0.096 U.S. grapes/ROW grapes .014928 .871201E-02 1.71350 0.087 Argentinean grapes -.849552E-02 .393029E-02 -2.16155 0.031 Argentinean grapes/ROW grapes .150522E-02 .469637E-02 .320507 0.749 ROW grapes -.037156 .011279 -3.29425 0.001 Israelis citrus -.020374 .361853E-02 -5.63055 0.000 Israelis citrus/Italian citrus .392133E-02 .258285E-02 1.51822 0.129 Israelis citrus/ROW citrus .348951E-02 .186858E-02 1.86746 0.062 Italian citrus -.019783 .372399E-02 -5.31230 0.000 Italian citrus/ROW citrus -.357662E-03 .160896E-02 -.222294 0.824 ROW citrus -.019498 .337577E-02 -5.77585 0.000 Table C-3. Marginal value shares of fruit juices in a block independent uniform-substitute
Rotterdam model Product Estimates ( )iθ SE t-statistics P-value
U.S. orange .027220 .792396E-02 3.43511 0.001 Brazilian orange .907700E-02 .306674E-02 2.95983 0.003 ROW orange .743746 .035095 21.1922 0.000 U.S. grapefruit .028811 .566702E-02 5.08406 0.000 Israelis grapefruit .311609E-02 .199038E-02 1.56558 0.117 ROW grapefruit .189446E-02 .110629E-02 1.71245 0.087 U.S. apple .035644 .891810E-02 3.99682 0.000 Chinese apple .038398 .761162E-02 5.04471 0.000 ROW apple .051404 .013209 3.89161 0.000 Thai pineapple .415274E-02 .136728E-02 3.03723 0.002 Philippine pineapple .817306E-02 .156327E-02 5.22817 0.000 ROW pineapple .102950E-02 .921286E-03 1.11746 0.264 U.S. grape .907845E-02 .462842E-02 1.96146 0.050 Argentinean grape .152513E-02 .968323E-03 1.57503 0.115 ROW grape .010845 .458813E-02 2.36372 0.018 Israelis other citrus .873475E-02 .204385E-02 4.27369 0.000 Italian other citrus .877093E-02 .216511E-02 4.05103 0.000 ROW other citrus .837951E-02 .191502E-02 4.37568 0.000 φ -1.79007 .273683 -6.54066 0.000 K1 .733130 .177836 4.12250 0.000 K2 19.5729 8.29988 2.35822 0.018 K3 -18.1472 14.8485 -1.22215 0.222 K4 27.0649 16.8737 1.60397 0.109 K5 31.4922 16.8636 1.86747 0.062 K6 9.85689 6.17161 1.59713 0.110
138
Table C-4. Marginal value shares of fruit juices in a block-wise dependent Rotterdam model
Products Estimates ( )iθ SE t-statistics P-value
U.S. orange .034515 .973502E-02 3.54549 0.000 Brazilian orange .712897 .035439 20.1161 0.000 ROW orange .404939E-02 .594273E-02 .681402 0.496 U.S. grapefruit .036395 .736038E-02 4.94476 0.000 Israelis grapefruit -.200246E-02 .250301E-02 -.800022 0.424 ROW grapefruit .312040E-02 .163735E-02 1.90576 0.057 U.S. apple .052199 .010814 4.82675 0.000 Chinese apple .049507 .801679E-02 6.17537 0.000 ROW apple .066990 .014307 4.68229 0.000 Thai pineapple .965923E-03 .123731E-02 .780663 0.435 Philippine pineapple .150696E-02 .192426E-02 .783138 0.434 ROW pineapple .130863E-03 .253873E-03 .515466 0.606 U.S. grape .775069E-02 .434665E-02 1.78314 0.075 Argentinean grape .633438E-03 .137377E-02 .461094 0.645 ROW grape .013108 .565344E-02 2.31853 0.020 Israelis other citrus .530260E-02 .198499E-02 2.67135 0.008 Italian other citrus .629891E-02 .219617E-02 2.86813 0.004 ROW other citrus .663242E-02 .228522E-02 2.90231 0.004 φ -1.81259
Table C-5. Constant of proportionality of fruit juice groups in a in block-wise dependent
Rotterdam model Products Estimates ( )ghΦ SE t-statistics P-value
Orange/grapefruit .347771 .318666 1.09133 0.275 Orange/apple .991497 .180586 5.49044 0.000 Orange/pineapple -3.79227 6.36190 -.596090 0.551 Orange/grapes .633182 .679618 .931673 0.352 Orange/other citrus .702609 .482882 1.45503 0.146 Grapefruit/apple -.715711 1.53489 -.466294 0.641 Grapefruit/pineapple 174.662 226.483 .771191 0.441 Grapefruit/grape -10.4734 9.74829 -1.07439 0.283 Grapefruit/other citrus 7.54488 6.33844 1.19034 0.234 Apple/pineapple -36.9035 50.4271 -.731819 0.464 Apple/grapes 1.76609 3.16376 .558226 0.577 Apple/other citrus -3.48605 2.88642 -1.20774 0.227 Pineapple/grape -105.564 156.725 -.673559 0.501 Pineapple/other citrus -111.149 160.841 -.691045 0.490 Grapes/other citrus -38.1447 25.1767 -1.51508 0.130
Table C-6. Within-group relative price coefficients of fruit juices in a block-wise
dependent Rotterdam Products Estimates SE t-statistics P-value U.S. Orange -.106408 .020796 -5.11685 0.000 U.S. orange/Brazilian orange .048202 .025105 1.92002 0.055 U.S. orange/ROW orange .795979E-02 .679637E-02 1.17118 0.242 Brazilian orange -1.10707 .108089 -10.2423 0.000 Brazilian orange/ROW orange .021053 .011660 1.80558 0.071 ROW orange -.034908 .486964E-02 -7.16855 0.000 U.S. grapefruit -.049379 .012263 -4.02655 0.000
139
Table C-6. Continued Products Estimates SE t-statistics P-value U.S. grapefruit/Israelis grapefruit .012179 .354942E-02 3.43113 0.001 U.S. grapefruit/ROW grapefruit .473403E-02 .216328E-02 2.18835 0.029 Israelis grapefruit -.012886 .351582E-02 -3.66509 0.000 Israelis grapefruit/ROW grapefruit .249360E-02 .147894E-02 1.68608 0.092 ROW grapefruit -.010011 .127196E-02 -7.87072 0.000 U.S. apple -.012925 .012130 -1.06555 0.287 U.S. apple/Chinese apple -.145159E-03 .871601E-02 -.016654 0.987 U.S. apple/ROW apple -.025104 .013887 -1.80777 0.071 Chinese apple -.034340 .011599 -2.96050 0.003 Chinese apple/ROW apple -.172017E-02 .013678 -.125766 0.900 ROW apple -.022167 .022440 -.987838 0.323 Thai. pineapple -.012133 .209892E-02 -5.78078 0.000 Thai. pineapple/Philippines pineapple -.677701E-03 .147053E-02 -.460856 0.645 Thai. pineapple/ROW P. apple -.877698E-03 .135307E-02 -.648669 0.517 Philipp. pineapple -.021271 .208471E-02 -10.2034 0.000 Philip. pineapple/ROW apple .592741E-03 .142519E-02 .415902 0.677 ROW pineapple -.156960E-02 .226334E-02 -.693489 0.488 U.S. grape -.039620 .010588 -3.74213 0.000 U.S. grapes/Argentinean grapes .669583E-02 .424842E-02 1.57608 0.115 U.S. grapes/ROW grape .010593 .951660E-02 1.11313 0.266 Argentinean grape -.940781E-02 .389373E-02 -2.41614 0.016 Argentinean grape/ROW grape .886930E-03 .487254E-02 .182026 0.856 ROW grape -.049246 .013584 -3.62531 0.000 Israelis citrus -.021556 .502676E-02 -4.28828 0.000 Israelis citrus/Italian citrus .221872E-02 .276928E-02 .801189 0.423 Israelis citrus/ROW citrus .240213E-02 .187447E-02 1.28150 0.200 Italian citrus -.020363 .561152E-02 -3.62880 0.000 Italian citrus/ROW citrus -.197293E-02 .162961E-02 -1.21068 0.226 ROW citrus -.021612 .511057E-02 -4.22881 0.000 Table C-7. Marginal value shares of fruit juices in a block-wise dependent uniform-
substitute Rotterdam model Products Estimates ( )iθ SE t-statistics P-value
U.S. orange .023935 .807320E-02 2.96476 0.003 Brazilian orange .781733 .026914 29.0453 0.000 ROW orange .744781E-02 .288702E-02 2.57975 0.010 U.S. grapefruit .030032 .553653E-02 5.42426 0.000 Israelis grapefruit .243101E-02 .142535E-02 1.70555 0.088 ROW grapefruit .154478E-02 .809084E-03 1.90929 0.056 U.S. apple .039106 .837319E-02 4.67034 0.000 Chinese apple .818502E-02 .265280E-02 3.08543 0.002 ROW apple .043093 .011718 3.67749 0.000 Thai pineapple .677959E-02 .111513E-02 6.07963 0.000 Philippine pineapple .614985E-02 .734483E-03 8.37304 0.000 ROW pineapple .362882E-03 .913188E-03 .397379 0.691 U.S. grapes .920016E-02 .442613E-02 2.07860 0.038 Argentinean grape .197440E-02 .107270E-02 1.84059 0.066 ROW grape .012908 .467776E-02 2.75942 0.006 Israelis citrus .969673E-02 .194521E-02 4.98492 0.000 Italian citrus .950458E-02 .210530E-02 4.51459 0.000 ROW citrus .591649E-02 .146101E-02 4.04957 0.000 φ -1.97888 .311601 -6.35068 0.000
140
Table C-7 Continued Products Estimates ( )iθ SE t-statistics P-value
K1 .783859 .164457 4.76634 0.000 K2 21.2047 7.18763 2.95016 0.003 K3 841.401 19338.8 .043508 0.965 K4 4.94019 17.2078 .287090 0.774 K5 27.6275 13.9415 1.98167 0.048 K6 11.4152 7.02358 1.62527 0.104 Table C-8. Constant of proportionality of fruit juice groups in a block-wise dependent
uniform-substitute-Rotterdam model Products Estimates ( )ghΦ SE t-statistics P-value
Orange/grapefruit .251369 .398478 .630825 0.528 Orange/apple .791079 .291463 2.71416 0.007 Orange/pineapple .341223 .482032 .707885 0.479 Orange/grapes 1.34229 .674993 1.98861 0.047 Orange/other citrus -.094332 .435253 -.216729 0.828 Grapefruit/apple -7.41443 3.23381 -2.29279 0.022 Grapefruit/pineapple 50.6194 13.4650 3.75932 0.000 Grapefruit/grapes -8.11878 9.42408 -.861493 0.389 Grapefruit/other citrus 6.87398 5.71871 1.20202 0.229 Apple/pineapple -15.7554 5.15323 -3.05738 0.002 Apple/grape 1.85589 5.43242 .341633 0.733 Apple/other citrus -1.71658 3.12146 -.549928 0.582 Pineapple/grape -11.7357 13.9605 -.840635 0.401 Pineapple/other citrus -20.4015 9.37285 -2.17666 0.030 Grapes/other citrus -13.0223 10.3033 -1.26389 0.206 Table C-9 Within-group relative price coefficients of block-wise dependent uniform
substitute Rotterdam model Products Estimates SE t-statistics P-value U.S. Orange -.108070 .019726 -5.47866 0.000 U.S. orange/Brazilian orange .067488 .023219 2.90661 0.004 U.S. orange/ROW orange .642978E-03 .150383E-03 4.27559 0.000 Brazilian orange -1.39291 .059985 -23.2209 0.000 Brazilian orange/ROW orange .021000 .576683E-02 3.64151 0.000 ROW orange -.034071 .481407E-02 -7.07730 0.000 U.S. grapefruit -.065260 .897959E-02 -7.26760 0.000 U.S. grapefruit/Israelis grapefruit .926256E-02 .304626E-02 3.04064 0.002 U.S. grapefruit/ROW grapefruit .588587E-02 .156089E-02 3.77084 0.000 Israelis grapefruit -.013795 .363264E-02 -3.79765 0.000 Israelis grapefruit/ROW grapefruit .476452E-03 .198748E-03 2.39727 0.017 ROW grapefruit -.893998E-02 .130118E-02 -6.87068 0.000 U.S. apple -.027739 .012240 -2.26633 0.023 U.S. apple/Chinese apple -.598793E-02 .317106E-02 -1.88831 0.059 U.S. apple/ROW apple -.031526 .011917 -2.64553 0.008 Chinese apple -.107132E-02 .379707E-02 -.282144 0.778 Chinese apple/ROW apple -.659852E-02 .296054E-02 -2.22882 0.026 ROW apple -.033782 .019677 -1.71680 0.086 Thai. pineapple -.011702 .159136E-02 -7.35359 0.000 Thai. pineapple/Philippines pineapple .367849E-03 .130420E-02 .282051 0.778 Thai. pineapple/ROW pineapple .217055E-04 .908518E-04 .238911 0.811
141
Table C-9 Continued Products Estimates SE t-statistics P-value Philipp. pineapple -.010649 .105798E-02 -10.0657 0.000 Philip. pineapple/ROW apple .196894E-04 .824165E-04 .238901 0.811 ROW pineapple -.646910E-03 .161833E-02 -.399739 0.689 U.S. grape -.034212 .961848E-02 -3.55695 0.000 U.S. grape/Argentinean grape .250221E-02 .129180E-02 1.93699 0.053 U.S. grape/ROW grape .016359 .742953E-02 2.20183 0.028 Argentinean grape -.930737E-02 .368287E-02 -2.52721 0.011 Argentinean grape/ROW grape .351062E-02 .196294E-02 1.78845 0.074 ROW grape -.041408 .010598 -3.90724 0.000 Israelis citrus -.020174 .331409E-02 -6.08720 0.000 Israelis citrus/Italian citrus .246119E-02 .148253E-02 1.66013 0.097 Israelis citrus/ROW citrus .153206E-02 .863379E-03 1.77449 0.076 Italian citrus -.019823 .348924E-02 -5.68105 0.000 Italian citrus/ROW citrus .150170E-02 .824028E-03 1.82239 0.068 ROW citrus -.012906 .221224E-02 -5.83400 0.000
142
APPENDIX D PRICE ELASTICITIES OF FRUIT JUICES IN JAPAN IN DIFFERENT MARKET
STRUCTURES
Table D-1 Compensated price elasticities in the block independent non-uniform substitute Rotterdam model
Products Estimates SE t-statistic p-value U.S orange /Brazilian orange 1.32124 .289587 4.56250 0.000 U.S orange /ROW orange .084107 .094148 .893351 0.372 U.S orange /U.S grapefruit .020821 .788673E-02 2.63995 0.008 U.S orange /Israelis grapefruit .165653E-02 .213166E-02 .777110 0.437 U.S orange /ROW grapefruit .362166E-03 .119832E-02 .302228 0.762 U.S orange /U.S apple .028177 .011604 2.42832 0.015 U.S orange /Chinese apple .025862 .010026 2.57943 0.010 U.S orange /ROW apple .032910 .014574 2.25809 0.024 U.S orange /Thai pineapple .322414E-02 .141774E-02 2.27414 0.023 U.S orange /Philippines pineapple .549957E-02 .197710E-02 2.78164 0.005 U.S orange /ROW pineapple .597278E-03 .126514E-02 .472102 0.637 U.S orange /U.S grape .616171E-02 .406805E-02 1.51466 0.130 U.S orange /Argentinean grape -.592139E-04 .129673E-02 -.045664 0.964 U.S orange /ROW grape .799705E-02 .441252E-02 1.81235 0.070 U.S orange /Israelis citrus .500277E-02 .221493E-02 2.25865 0.024 U.S orange /Italian citrus .625923E-02 .263442E-02 2.37594 0.018 U.S orange /ROW citrus .631588E-02 .208804E-02 3.02478 0.002 Brazilian orange/U.S orange .376541 .082530 4.56250 0.000 Brazilian orange /ROW orange .104657 .031357 3.33755 0.001 Brazilian orange /U.S grapefruit .158256 .029750 5.31954 0.000 Brazilian orange /Israelis grapefruit .012591 .015786 .797640 0.425 Brazilian orange /ROW grapefruit .275280E-02 .899244E-02 .306124 0.760 Brazilian orange /U.S apple .214174 .052592 4.07236 0.000 Brazilian orange /Chinese apple .196573 .043131 4.55758 0.000 Brazilian orange /ROW apple .250147 .065800 3.80161 0.000 Brazilian orange /Thai Pineapple .024506 .796387E-02 3.07721 0.002 Brazilian orange /Philippines pineapple .041802 .805902E-02 5.18698 0.000 Brazilian orange /ROW Pineapple .453987E-02 .957600E-02 .474089 0.635 Brazilian orange /U.S grape .046835 .025360 1.84680 0.065 Brazilian orange /Argentinean grape -.450081E-03 .984323E-02 -.045725 0.964 Brazilian orange /ROW grape .060785 .025038 2.42773 0.015 Brazilian orange /Israelis citrus .038026 .011626 3.27073 0.001 Brazilian orange /Italian citrus .047576 .011853 4.01387 0.000 Brazilian orange /ROW citrus .048007 .883406E-02 5.43427 0.000 ROW orange/U.S orange .188043 .210492 .893351 0.372 ROW orange/Brazilian orange .821035 .245999 3.33755 0.001 ROW orange /U.S grapefruit .017893 .010040 1.78216 0.075 ROW orange /Israelis grapefruit .142361E-02 .194436E-02 .732175 0.464 ROW orange /ROW grapefruit .311243E-03 .103301E-02 .301297 0.763 ROW orange /U.S apple .024215 .012856 1.88366 0.060 ROW orange /Chinese apple .022225 .012497 1.77840 0.075 ROW orange /ROW apple .028283 .016161 1.75001 0.080
143
Table D-1 Continued Products Estimates SE t-statistic p-value ROW orange /Thai pineapple .277080E-02 .171742E-02 1.61335 0.107 ROW orange /Philippines pineapple .472630E-02 .264276E-02 1.78839 0.074 ROW orange /ROW pineapple .513297E-03 .111734E-02 .459390 0.646 ROW orange /U.S grape .529533E-02 .397220E-02 1.33310 0.182 ROW orange /Argentinean grape -.508881E-04 .111433E-02 -.045667 0.964 ROW orange /ROW grape .687261E-02 .448822E-02 1.53125 0.126 ROW orange /Israelis citrus .429935E-02 .258938E-02 1.66038 0.097 ROW orange /Italian citrus .537914E-02 .306400E-02 1.75560 0.079 ROW orange /ROW citrus .542783E-02 .281229E-02 1.93004 0.054 U.S grapefruit/U.S orange .018657 .706730E-02 2.63995 0.008 U.S grapefruit/Brazilian orange .497606 .093543 5.31954 0.000 U.S grapefruit/ROW orange .717161E-02 .402412E-02 1.78216 0.075 U.S grapefruit/Israelis grapefruit .101769 .053267 1.91054 0.056 U.S grapefruit /ROW grapefruit .072159 .031081 2.32165 0.020 U.S grapefruit /U.S apple .026876 .784770E-02 3.42465 0.001 U.S grapefruit /Chinese apple .024667 .670010E-02 3.68157 0.000 U.S grapefruit /ROW apple .031390 .010538 2.97884 0.003 U.S grapefruit /Thai Pineapple .307519E-02 .110592E-02 2.78066 0.005 U.S grapefruit /Philippines Pineapple .524550E-02 .128740E-02 4.07450 0.000 U.S grapefruit /ROW Pineapple .569685E-03 .119376E-02 .477218 0.633 U.S grapefruit /U.S grape .587705E-02 .354771E-02 1.65658 0.098 U.S grapefruit /Argentinean grape -.564783E-04 .123661E-02 -.045672 0.964 U.S grapefruit /ROW grape .762760E-02 .346979E-02 2.19829 0.028 U.S grapefruit /Israelis citrus .477165E-02 .170892E-02 2.79221 0.005 U.S grapefruit /Italian citrus .597006E-02 .171974E-02 3.47148 0.001 U.S grapefruit /ROW citrus .602410E-02 .112057E-02 5.37593 0.000 Israelis grapefruit/U.S orange .461992E-02 .594500E-02 .777110 0.437 Israelis grapefruit/Brazilian orange .123217 .154477 .797640 0.425 Israelis grapefruit/ROW orange .177583E-02 .242542E-02 .732175 0.464 Israelis grapefruit/U.S grapefruit .316733 .165782 1.91054 0.056 Israelis grapefruit/ROW grapefruit .101898 .061756 1.65001 0.099 Israelis grapefruit /U.S apple .665494E-02 .838059E-02 .794089 0.427 Israelis grapefruit /Chinese apple .610801E-02 .780126E-02 .782951 0.434 Israelis grapefruit /ROW apple .777268E-02 .992639E-02 .783032 0.434 Israelis grapefruit /Thai Pineapple .761478E-03 .999705E-03 .761703 0.446 Israelis grapefruit /Philippines Pineapple .129889E-02 .165398E-02 .785311 0.432 Israelis grapefruit /ROW Pineapple .141065E-03 .349551E-03 .403561 0.687 Israelis grapefruit /U.S grape .145527E-02 .208258E-02 .698783 0.485 Israelis grapefruit /Argentinean grape -.139852E-04 .306963E-03 -.045560 0.964 Israelis grapefruit /ROW grape .188874E-02 .252413E-02 .748275 0.454 Israelis grapefruit /Israelis citrus .118156E-02 .156192E-02 .756475 0.449 Israelis grapefruit /Italian citrus .147831E-02 .192165E-02 .769292 0.442 Israelis grapefruit /ROW citrus .149169E-02 .187534E-02 .795422 0.426 ROW grapefruit/U.S orange .234822E-02 .776969E-02 .302228 0.762 ROW grapefruit/Brazilian orange .062629 .204587 .306124 0.760 ROW grapefruit/ROW orange .902623E-03 .299579E-02 .301297 0.763 ROW grapefruit/U.S grapefruit .522116 .224890 2.32165 0.020 ROW grapefruit/Israelis grapefruit .236898 .143574 1.65001 0.099 ROW grapefruit /U.S apple .338258E-02 .011218 .301536 0.763 ROW grapefruit /Chinese apple .310459E-02 .010345 .300091 0.764 ROW grapefruit /ROW apple .395071E-02 .013065 .302394 0.762 ROW grapefruit /Thai Pineapple .387045E-03 .127665E-02 .303173 0.762 ROW grapefruit /Philippines Pineapple .660202E-03 .216847E-02 .304455 0.761 ROW grapefruit /ROW Pineapple .717009E-04 .283864E-03 .252589 0.801
144
Table D-1 Continued Products Estimates SE t-statistic p-value ROW grapefruit /U.S grape .739689E-03 .246901E-02 .299590 0.764 ROW grapefruit /Argentinean grape -.710840E-05 .156141E-03 -.045526 0.964 ROW grapefruit /ROW grape .960014E-03 .319410E-02 .300559 0.764 ROW grapefruit /Israelis citrus .600563E-03 .197850E-02 .303545 0.761 ROW grapefruit /Italian citrus .751396E-03 .246251E-02 .305134 0.760 ROW grapefruit /ROW citrus .758197E-03 .248439E-02 .305184 0.760 U.S apple/U.S orange .035946 .014803 2.42832 0.015 U.S apple/Brazilian orange .958718 .235421 4.07236 0.000 U.S apple/ROW orange .013817 .733532E-02 1.88366 0.060 U.S apple/U.S grapefruit .038261 .011172 3.42465 0.001 U.S apple/Israelis grapefruit .304413E-02 .383349E-02 .794089 0.427 U.S apple/ROW grapefruit .665536E-03 .220715E-02 .301536 0.763 U.S apple /Chinese apple -.128349 .151178 -.848993 0.396 U.S apple /ROW apple -.632934 .241559 -2.62021 0.009 U.S apple /Thai Pineapple .592485E-02 .235090E-02 2.52025 0.012 U.S apple /Philippines Pineapple .010106 .288415E-02 3.50409 0.000 U.S apple /ROW Pineapple .109759E-02 .231411E-02 .474303 0.635 U.S apple /U.S grape .011323 .699435E-02 1.61889 0.105 U.S apple /Argentinean grape -.108815E-03 .238045E-02 -.045712 0.964 U.S apple /ROW grape .014696 .712247E-02 2.06330 0.039 U.S apple /Israelis citrus .919336E-02 .351033E-02 2.61895 0.009 U.S apple /Italian citrus .011502 .408548E-02 2.81541 0.005 U.S apple /ROW citrus .011606 .289072E-02 4.01506 0.000 Chinese apple/U.S orange .025740 .997880E-02 2.57943 0.010 Chinese apple/Brazilian orange .686498 .150628 4.55758 0.000 Chinese apple/ROW orange .989397E-02 .556340E-02 1.77840 0.075 Chinese apple/U.S grapefruit .027397 .744170E-02 3.68157 0.000 Chinese apple/Israelis grapefruit .217978E-02 .278405E-02 .782951 0.434 Chinese apple/ROW grapefruit .476562E-03 .158806E-02 .300091 0.764 Chinese apple /U.S apple -.100134 .117945 -.848993 0.396 Chinese apple /ROW apple -.172418 .185903 -.927462 0.354 Chinese apple /Thai Pineapple .424254E-02 .151665E-02 2.79730 0.005 Chinese apple /Philippines Pineapple .723671E-02 .192800E-02 3.75347 0.000 Chinese apple /ROW Pineapple .785939E-03 .165091E-02 .476065 0.634 Chinese apple /U.S grape .810799E-02 .492428E-02 1.64653 0.100 Chinese apple /Argentinean grape -.779176E-04 .170353E-02 -.045739 0.964 Chinese apple /ROW grape .010523 .497631E-02 2.11463 0.034 Chinese apple /Israelis citrus .658298E-02 .229570E-02 2.86753 0.004 Chinese apple /Italian citrus .823631E-02 .264109E-02 3.11853 0.002 Chinese apple /ROW citrus .831086E-02 .183976E-02 4.51735 0.000 ROW apple/U.S orange .014428 .638953E-02 2.25809 0.024 ROW apple/Brazilian orange .384809 .101223 3.80161 0.000 ROW apple/ROW orange .554597E-02 .316911E-02 1.75001 0.080 ROW apple/U.S grapefruit .015357 .515542E-02 2.97884 0.003 ROW apple/Israelis grapefruit .122185E-02 .156041E-02 .783032 0.434 ROW apple/ROW grapefruit .267132E-03 .883391E-03 .302394 0.762 ROW apple /U.S apple -.217513 .083014 -2.62021 0.009 ROW apple /Chinese apple -.075948 .081888 -.927462 0.354 ROW apple /Thai Pineapple .237811E-02 .970736E-03 2.44980 0.014 ROW apple /Philippines Pineapple .405646E-02 .127597E-02 3.17911 0.001 ROW apple /ROW Pineapple .440550E-03 .923691E-03 .476945 0.633 ROW apple /U.S grape .454485E-02 .302379E-02 1.50303 0.133 ROW apple /Argentinean grape -.436759E-04 .953259E-03 -.045818 0.963 ROW apple /ROW grape .589859E-02 .285285E-02 2.06761 0.039
145
Table D-1 Continued Products Estimates SE t-statistic p-value ROW apple /Israelis citrus .369002E-02 .157714E-02 2.33969 0.019 ROW apple /Italian citrus .461678E-02 .169382E-02 2.72567 0.006 ROW apple /ROW citrus .465857E-02 .128729E-02 3.61890 0.000 Thai pineapple/U.S orange .021352 .938905E-02 2.27414 0.023 Thai pineapple/Brazilian orange .569475 .185062 3.07721 0.002 Thai pineapple /ROW orange .820742E-02 .508719E-02 1.61335 0.107 Thai pineapple /U.S grapefruit .022727 .817320E-02 2.78066 0.005 Thai pineapple /Israelis grapefruit .180821E-02 .237390E-02 .761703 0.446 Thai pineapple /ROW grapefruit .395326E-03 .130396E-02 .303173 0.762 Thai pineapple /U.S apple .030757 .012204 2.52025 0.012 Thai pineapple /Chinese apple .028230 .010092 2.79730 0.005 Thai pineapple /ROW apple .035923 .014664 2.44980 0.014 Thai pineapple /Philippines pineapple .215600 .129259 1.66797 0.095 Thai pineapple /ROW Pineapple -.049248 .130583 -.377139 0.706 Thai pineapple /U.S grape .672588E-02 .438852E-02 1.53261 0.125 Thai pineapple /Argentinean GR -.646356E-04 .141318E-02 -.045738 0.964 Thai pineapple /ROW grape .872926E-02 .458452E-02 1.90407 0.057 Thai pineapple /Israelis citrus .546083E-02 .242554E-02 2.25138 0.024 Thai pineapple /Italian citrus .683233E-02 .273757E-02 2.49576 0.013 Thai pineapple /ROW citrus .689417E-02 .220942E-02 3.12035 0.002 Philippines pineapple/U.S orange .052529 .018884 2.78164 0.005 Philippines pineapple/Brazilian orange 1.40099 .270098 5.18698 0.000 Philippines pineapple /ROW orange .020191 .011290 1.78839 0.074 Philippines pineapple /U.S grapefruit .055912 .013722 4.07450 0.000 Philippines pineapple /Israelis grapefruit .444845E-02 .566457E-02 .785311 0.432 Philippines pineapple /ROW grapefruit .972560E-03 .319443E-02 .304455 0.761 Philippines pineapple /U.S apple .075667 .021594 3.50409 0.000 Philippines pineapple /Chinese apple .069449 .018503 3.75347 0.000 Philippines pineapple /ROW apple .088376 .027799 3.17911 0.001 Philippines pineapple /Thai pineapple .310952 .186425 1.66797 0.095 Philippines pineapple /ROW pineapple .233007 .190120 1.22558 0.220 Philippines pineapple /U.S grape .016547 .986842E-02 1.67673 0.094 Philippines pineapple /Argentinean grape -.159013E-03 .347860E-02 -.045712 0.964 Philippines pineapple /ROW grape .021475 .940750E-02 2.28278 0.022 Philippines pineapple /Israelis other citrus .013434 .481960E-02 2.78746 0.005 Philippines pineapple /Italian citrus .016809 .520368E-02 3.23012 0.001 Philippines pineapple /ROW other citrus .016961 .309996E-02 5.47125 0.000 ROW pineapple/U.S orange .483651E-02 .010245 .472102 0.637 ROW pineapple/Brazilian orange .128994 .272088 .474089 0.635 ROW pineapple /ROW orange .185909E-02 .404686E-02 .459390 0.646 ROW pineapple /U.S grapefruit .514795E-02 .010787 .477218 0.633 ROW pineapple /Israelis grapefruit .409583E-03 .101492E-02 .403561 0.687 ROW pineapple /ROW grapefruit .895467E-04 .354515E-03 .252589 0.801 ROW pineapple /U.S apple .696694E-02 .014689 .474303 0.635 ROW pineapple /Chinese apple .639437E-02 .013432 .476065 0.634 ROW pineapple /ROW apple .813709E-02 .017061 .476945 0.633 ROW pineapple /Thai pineapple -.060217 .159668 -.377139 0.706 ROW pineapple /Philippines pineapple .197540 .161181 1.22558 0.220 ROW pineapple /U.S grape .152350E-02 .333130E-02 .457329 0.647 ROW pineapple /Argentinean grape -.146408E-04 .322087E-03 -.045456 0.964 ROW pineapple /ROW grape .197730E-02 .420240E-02 .470516 0.638 ROW pineapple /Israelis citrus .123695E-02 .265972E-02 .465068 0.642 ROW pineapple /Italian citrus .154761E-02 .328073E-02 .471728 0.637 ROW pineapple /ROW citrus .156162E-02 .329046E-02 .474591 0.635
146
Table D-1 Continued Products Estimates SE t-statistic p-value U.S grape/U.S orange .718253E-02 .474201E-02 1.51466 0.130 U.S grape/Brazilian orange .191564 .103727 1.84680 0.065 U.S grape/ROW orange .276087E-02 .207101E-02 1.33310 0.182 U.S grape/U.S grapefruit .764503E-02 .461496E-02 1.65658 0.098 U.S grape/Israelis grapefruit .608256E-03 .870451E-03 .698783 0.485 U.S grape/ROW grapefruit .132982E-03 .443882E-03 .299590 0.764 U.S grape/U.S apple .010346 .639101E-02 1.61889 0.105 U.S grape/Chinese apple .949604E-02 .576730E-02 1.64653 0.100 U.S grape/ROW apple .012084 .803981E-02 1.50303 0.133 U.S grape/Thai Pineapple .118386E-02 .772449E-03 1.53261 0.125 U.S grape /Philippines Pineapple .201937E-02 .120435E-02 1.67673 0.094 U.S grape /ROW Pineapple .219312E-03 .479550E-03 .457329 0.647 U.S grape /Argentinean grape .114908 .068875 1.66836 0.095 U.S grape /ROW grape .243103 .139356 1.74447 0.081 U.S grape /Israelis citrus .183695E-02 .121649E-02 1.51004 0.131 U.S grape /Italian citrus .229830E-02 .144642E-02 1.58896 0.112 U.S grape /ROW citrus .231911E-02 .128542E-02 1.80416 0.071 Argentinean grape/U.S orange -.467724E-03 .010243 -.045664 0.964 Argentinean grape/Brazilian orange -.012475 .272817 -.045725 0.964 Argentinean grape/ROW orange -.179787E-03 .393691E-02 -.045667 0.964 Argentinean grape/U.S grapefruit -.497842E-03 .010900 -.045672 0.964 Argentinean grape/Israelis grapefruit -.396094E-04 .869395E-03 -.045560 0.964 Argentinean grape/ROW grapefruit -.865977E-05 .190218E-03 -.045526 0.964 Argentinean grape/U.S apple -.673750E-03 .014739 -.045712 0.964 Argentinean grape/Chinese apple -.618379E-03 .013520 -.045739 0.964 Argentinean grape/ROW apple -.786912E-03 .017175 -.045818 0.963 Argentinean grape/Thai pineapple -.770926E-04 .168554E-02 -.045738 0.964 Argentinean grape /Philippines pineapple. -.131501E-03 .287673E-02 -.045712 0.964 Argentinean grape /ROW Pineapple -.142815E-04 .314184E-03 -.045456 0.964 Argentinean grape /U.S grape .778648 .466716 1.66836 0.095 Argentinean grape /ROW grape .163905 .510347 .321164 0.748 Argentinean grape /Israelis other citrus -.119622E-03 .261243E-02 -.045789 0.963 Argentinean grape /Italian citrus -.149665E-03 .327287E-02 -.045729 0.964 Argentinean grape /ROW other CT -.151019E-03 .330236E-02 -.045731 0.964 ROW grape/U.S orange .894107E-02 .493340E-02 1.81235 0.070 ROW grape/Brazilian orange .238466 .098226 2.42773 0.015 ROW grape/ROW orange .343683E-02 .224445E-02 1.53125 0.126 ROW grape/U.S grapefruit .951681E-02 .432919E-02 2.19829 0.028 ROW grape/Israelis grapefruit .757180E-03 .101190E-02 .748275 0.454 ROW grape/ROW grapefruit .165541E-03 .550779E-03 .300559 0.764 ROW grape/U.S apple .012880 .624218E-02 2.06330 0.039 ROW grape/Chinese apple .011821 .559012E-02 2.11463 0.034 ROW grape/ROW apple .015043 .727540E-02 2.06761 0.039 ROW grape/Thai Pineapple .147371E-02 .773979E-03 1.90407 0.057 ROW grape /Philippines Pineapple .251378E-02 .110119E-02 2.28278 0.022 ROW grape /ROW Pineapple .273008E-03 .580232E-03 .470516 0.638 ROW grape /U.S grape .233171 .133663 1.74447 0.081 ROW grape /Argentinean grape .023200 .072237 .321164 0.748 ROW grape /Israelis citrus .228670E-02 .116738E-02 1.95883 0.050 ROW grape /Italian citrus .286101E-02 .130515E-02 2.19210 0.028 ROW grape /ROW citrus .288691E-02 .119962E-02 2.40651 0.016 Israelis citrus/U.S orange .016410 .726559E-02 2.25865 0.024 Israelis citrus /Brazilian orange .437681 .133817 3.27073 0.001 Israelis citrus /ROW orange .630796E-02 .379911E-02 1.66038 0.097
147
Table D-1 Continued Products Estimates SE t-statistic p-value Israelis citrus /U.S grapefruit .017467 .625569E-02 2.79221 0.005 Israelis citrus /Israelis grapefruit .138973E-02 .183711E-02 .756475 0.449 Israelis citrus /ROW grapefruit .303835E-03 .100096E-02 .303545 0.761 Israelis citrus /U.S apple .023639 .902618E-02 2.61895 0.009 Israelis citrus /Chinese apple .021696 .756620E-02 2.86753 0.004 Israelis citrus /ROW apple .027609 .011800 2.33969 0.019 Israelis citrus /Thai pineapple .270485E-02 .120142E-02 2.25138 0.024 Israelis citrus /Philippines pineapple .461380E-02 .165520E-02 2.78746 0.005 Israelis citrus /ROW Pineapple .501080E-03 .107743E-02 .465068 0.642 Israelis citrus /U.S grape .516929E-02 .342329E-02 1.51004 0.131 Israelis citrus /Argentinean GR -.496768E-04 .108490E-02 -.045789 0.963 Israelis citrus /ROW grape .670903E-02 .342501E-02 1.95883 0.050 Israelis citrus /Italian citrus .182783 .115721 1.57952 0.114 Israelis citrus /ROW citrus .163280 .084438 1.93374 0.053 Italian citrus/U.S orange .026244 .011046 2.37594 0.018 Italian citrus /Brazilian orange .699947 .174382 4.01387 0.000 Italian citrus /ROW orange .010088 .574609E-02 1.75560 0.079 Italian citrus /U.S grapefruit .027934 .804666E-02 3.47148 0.001 Italian citrus /Israelis grapefruit .222248E-02 .288900E-02 .769292 0.442 Italian citrus /ROW grapefruit .485899E-03 .159241E-02 .305134 0.760 Italian citrus /U.S apple .037804 .013428 2.81541 0.005 Italian citrus /Chinese apple .034697 .011126 3.11853 0.002 Italian citrus /ROW apple .044154 .016199 2.72567 0.006 Italian citrus /Thai pineapple .432565E-02 .173320E-02 2.49576 0.013 Italian citrus /Philippines pineapple .737848E-02 .228427E-02 3.23012 0.001 Italian citrus /ROW Pineapple .801336E-03 .169872E-02 .471728 0.637 Italian citrus /U.S grape .826683E-02 .520268E-02 1.58896 0.112 Italian citrus /Argentinean grape -.794441E-04 .173728E-02 -.045729 0.964 Italian citrus /ROW grape .010729 .489450E-02 2.19210 0.028 Italian citrus /Israelis citrus .233632 .147914 1.57952 0.114 Italian citrus /ROW citrus -.012224 .092850 -.131649 0.895 ROW citrus/U.S orange .018265 .603858E-02 3.02478 0.002 ROW citrus /Brazilian orange .487153 .089645 5.43427 0.000 ROW citrus /ROW orange .702097E-02 .363774E-02 1.93004 0.054 ROW citrus /U.S grapefruit .019442 .361641E-02 5.37593 0.000 ROW citrus /Israelis grapefruit .154682E-02 .194465E-02 .795422 0.426 ROW citrus /ROW grapefruit .338179E-03 .110811E-02 .305184 0.760 ROW citrus /U.S apple .026311 .655310E-02 4.01506 0.000 ROW citrus /Chinese apple .024149 .534578E-02 4.51735 0.000 ROW citrus /ROW apple .030730 .849159E-02 3.61890 0.000 ROW citrus /Thai pineapple .301060E-02 .964826E-03 3.12035 0.002 ROW citrus /Philippines pineapple .513532E-02 .938601E-03 5.47125 0.000 ROW citrus /ROW Pineapple .557719E-03 .117516E-02 .474591 0.635 ROW citrus /U.S grape .575360E-02 .318908E-02 1.80416 0.071 ROW citrus /Argentinean GR -.552920E-04 .120908E-02 -.045731 0.964 ROW citrus /ROW grape .746738E-02 .310299E-02 2.40651 0.016 ROW citrus /Israelis citrus .143952 .074443 1.93374 0.053 ROW citrus /Italian citrus -.843113E-02 .064043 -.131649 0.895
148
Table D-2 Compensated price elasticities of fruit juices in the block independent uniform substitute Rotterdam model
Products Estimates SE t-statistics p-value U.S orange /Brazilian orange 1.35664 .140729 9.64010 0.000 U.S orange /ROW orange .016557 .347986E-02 4.75796 0.000 U.S orange /U.S grapefruit .019375 .683231E-02 2.83585 0.005 U.S orange /Israelis grapefruit .209554E-02 .147707E-02 1.41872 0.156 U.S orange /ROW grapefruit .127401E-02 .836867E-03 1.52235 0.128 U.S orange /U.S apple .023970 .904463E-02 2.65021 0.008 U.S orange /Chinese apple .025823 .910398E-02 2.83640 0.005 U.S orange /ROW apple .034569 .013784 2.50788 0.012 U.S orange /Thai pineapple .279267E-02 .119275E-02 2.34137 0.019 U.S orange /Philippines pineapple .549629E-02 .180020E-02 3.05316 0.002 U.S orange /ROW pineapple .692329E-03 .645454E-03 1.07262 0.283 U.S orange /U.S grape .610516E-02 .358166E-02 1.70456 0.088 U.S orange /Argentinean grape .102563E-02 .709685E-03 1.44520 0.148 U.S orange /ROW grape .729317E-02 .386648E-02 1.88625 0.059 U.S orange /Israelis citrus .587402E-02 .212249E-02 2.76751 0.006 U.S orange /Italian citrus .589835E-02 .227377E-02 2.59408 0.009 U.S orange /ROW citrus .563513E-02 .163536E-02 3.44581 0.001 Brazilian orange/U.S orange .386631 .040106 9.64010 0.000 Brazilian orange /ROW orange .128930 .015603 8.26342 0.000 Brazilian orange /U.S grapefruit .150877 .028799 5.23897 0.000 Brazilian orange /Israelis grapefruit .016318 .010331 1.57954 0.114 Brazilian orange /ROW grapefruit .992072E-02 .570828E-02 1.73795 0.082 Brazilian orange /U.S apple .186657 .049195 3.79425 0.000 Brazilian orange /Chinese apple .201081 .042108 4.77532 0.000 Brazilian orange /ROW apple .269186 .066068 4.07439 0.000 Brazilian orange /Thai Pineapple .021747 .721234E-02 3.01519 0.003 Brazilian orange /Philippines pineapple .042800 .782044E-02 5.47281 0.000 Brazilian orange /ROW Pineapple .539119E-02 .481788E-02 1.11900 0.263 Brazilian orange /U.S grape .047541 .023125 2.05586 0.040 Brazilian orange /Argentinean grape .798666E-02 .503245E-02 1.58703 0.113 Brazilian orange /ROW grape .056792 .023989 2.36740 0.018 Brazilian orange /Israelis citrus .045741 .983527E-02 4.65073 0.000 Brazilian orange /Italian citrus .045931 .011038 4.16096 0.000 Brazilian orange /ROW citrus .043881 .803815E-02 5.45908 0.000 ROW orange/U.S orange .037018 .778012E-02 4.75796 0.000 ROW orange/Brazilian orange 1.01146 .122402 8.26342 0.000 ROW orange /U.S grapefruit .014446 .566488E-02 2.55002 0.011 ROW orange /Israelis grapefruit .156235E-02 .113281E-02 1.37918 0.168 ROW orange /ROW grapefruit .949849E-03 .642005E-03 1.47950 0.139 ROW orange /U.S apple .017871 .717880E-02 2.48945 0.013 ROW orange /Chinese apple .019252 .755050E-02 2.54980 0.011 ROW orange /ROW apple .025773 .010890 2.36675 0.018 ROW orange /Thai pineapple .208211E-02 .966564E-03 2.15413 0.031 ROW orange /Philippines pineapple .409782E-02 .153002E-02 2.67828 0.007 ROW orange /ROW pineapple .516173E-03 .489660E-03 1.05415 0.292 ROW orange /U.S grape .455177E-02 .274755E-02 1.65667 0.098 ROW orange /Argentinean grape .764674E-03 .538824E-03 1.41915 0.156 ROW orange /ROW grape .543750E-02 .298340E-02 1.82259 0.068 ROW orange /Israelis citrus .437944E-02 .173810E-02 2.51967 0.012 ROW orange /Italian citrus .439758E-02 .183113E-02 2.40157 0.016 ROW orange /ROW citrus .420133E-02 .140617E-02 2.98779 0.003 U.S grapefruit/U.S orange .017362 .612243E-02 2.83585 0.005
149
Table D-2 Continued Products Estimates SE t-statistics p-value U.S grapefruit/Brazilian orange .474404 .090553 5.23897 0.000 U.S grapefruit/ROW orange .578984E-02 .227051E-02 2.55002 0.011 U.S grapefruit/Israelis grapefruit .117086 .136823E-02 85.5749 0.000 U.S grapefruit /ROW grapefruit .071184 .779499E-03 91.3199 0.000 U.S grapefruit /U.S apple .022736 .698958E-02 3.25281 0.001 U.S grapefruit /Chinese apple .024493 .650614E-02 3.76456 0.000 U.S grapefruit /ROW apple .032788 .010617 3.08834 0.002 U.S grapefruit /Thai Pineapple .264886E-02 .975307E-03 2.71592 0.007 U.S grapefruit /Philippines Pineapple .521325E-02 .125832E-02 4.14302 0.000 U.S grapefruit /ROW Pineapple .656676E-03 .592520E-03 1.10828 0.268 U.S grapefruit /U.S grape .579076E-02 .318094E-02 1.82046 0.069 U.S grapefruit /Argentinean grape .972818E-03 .634359E-03 1.53355 0.125 U.S grapefruit /ROW grape .691759E-02 .320989E-02 2.15509 0.031 U.S grapefruit /Israelis citrus .557153E-02 .160619E-02 3.46878 0.001 U.S grapefruit /Italian citrus .559461E-02 .159199E-02 3.51422 0.000 U.S grapefruit /ROW citrus .534494E-02 .101128E-02 5.28533 0.000 Israelis grapefruit/U.S orange .584426E-02 .411941E-02 1.41872 0.156 Israelis grapefruit/Brazilian orange .159688 .101097 1.57954 0.114 Israelis grapefruit/ROW orange .194890E-02 .141308E-02 1.37918 0.168 Israelis grapefruit/U.S grapefruit .364404 .425831E-02 85.5749 0.000 Israelis grapefruit/ROW grapefruit .023961 .479156E-03 50.0066 0.000 Israelis grapefruit /U.S apple .765302E-02 .529825E-02 1.44444 0.149 Israelis grapefruit /Chinese apple .824442E-02 .569723E-02 1.44709 0.148 Israelis grapefruit /ROW apple .011037 .766012E-02 1.44081 0.150 Israelis grapefruit /Thai Pineapple .891622E-03 .643206E-03 1.38622 0.166 Israelis grapefruit /Philippines Pineapple .175481E-02 .115787E-02 1.51555 0.130 Israelis grapefruit /ROW Pineapple .221042E-03 .244049E-03 .905727 0.365 Israelis grapefruit /U.S grape .194921E-02 .163711E-02 1.19064 0.234 Israelis grapefruit /Argentinean grape .327457E-03 .300618E-03 1.08928 0.276 Israelis grapefruit /ROW grape .232851E-02 .181310E-02 1.28427 0.199 Israelis grapefruit /Israelis citrus .187541E-02 .127453E-02 1.47145 0.141 Israelis grapefruit /Italian citrus .188318E-02 .127557E-02 1.47635 0.140 Israelis grapefruit /ROW citrus .179914E-02 .114897E-02 1.56587 0.117 ROW grapefruit/U.S orange .826042E-02 .542609E-02 1.52235 0.128 ROW grapefruit/Brazilian orange .225706 .129869 1.73795 0.082 ROW grapefruit/ROW orange .275462E-02 .186185E-02 1.47950 0.139 ROW grapefruit/U.S grapefruit .515058 .564015E-02 91.3199 0.000 ROW grapefruit/Israelis grapefruit .055706 .111397E-02 50.0066 0.000 ROW grapefruit /U.S apple .010817 .698299E-02 1.54904 0.121 ROW grapefruit /Chinese apple .011653 .749386E-02 1.55499 0.120 ROW grapefruit /ROW apple .015600 .010045 1.55304 0.120 ROW grapefruit /Thai Pineapple .126024E-02 .846561E-03 1.48866 0.137 ROW grapefruit /Philippines Pineapple .248029E-02 .150239E-02 1.65090 0.099 ROW grapefruit /ROW Pineapple .312425E-03 .335513E-03 .931186 0.352 ROW grapefruit /U.S grape .275506E-02 .219401E-02 1.25572 0.209 ROW grapefruit /Argentinean grape .462835E-03 .407403E-03 1.13606 0.256 ROW grapefruit /ROW grape .329117E-02 .241929E-02 1.36039 0.174 ROW grapefruit /Israelis citrus .265075E-02 .166666E-02 1.59046 0.112 ROW grapefruit /Italian citrus .266173E-02 .165783E-02 1.60555 0.108 ROW grapefruit /ROW citrus .254295E-02 .148028E-02 1.71789 0.086 U.S apple/U.S orange .030579 .011538 2.65021 0.008 U.S apple/Brazilian orange .835539 .220212 3.79425 0.000 U.S apple/ROW orange .010197 .409621E-02 2.48945 0.013 U.S apple/U.S grapefruit .032367 .995060E-02 3.25281 0.001
150
Table D-2 Continued Products Estimates SE t-statistics p-value U.S apple/Israelis grapefruit .350068E-02 .242355E-02 1.44444 0.149 U.S apple/ROW grapefruit .212828E-02 .137393E-02 1.54904 0.121 U.S apple /Chinese apple -.195784 .015172 -12.9039 0.000 U.S apple /ROW apple -.262096 .021506 -12.1873 0.000 U.S apple /Thai Pineapple .466527E-02 .194148E-02 2.40294 0.016 U.S apple /Philippines Pineapple .918178E-02 .270749E-02 3.39126 0.001 U.S apple /ROW Pineapple .115656E-02 .106526E-02 1.08571 0.278 U.S apple /U.S grape .010199 .582505E-02 1.75087 0.080 U.S apple /Argentinean grape .171337E-02 .116862E-02 1.46614 0.143 U.S apple /ROW grape .012184 .610349E-02 1.99616 0.046 U.S apple /Israelis citrus .981280E-02 .322691E-02 3.04093 0.002 U.S apple /Italian citrus .985345E-02 .354608E-02 2.77869 0.005 U.S apple /ROW citrus .941372E-02 .251109E-02 3.74885 0.000 Chinese apple/U.S orange .025701 .906104E-02 2.83640 0.005 Chinese apple/Brazilian orange .702242 .147057 4.77532 0.000 Chinese apple/ROW orange .857047E-02 .336123E-02 2.54980 0.011 Chinese apple/U.S grapefruit .027204 .722627E-02 3.76456 0.000 Chinese apple/Israelis grapefruit .294220E-02 .203318E-02 1.44709 0.148 Chinese apple/ROW grapefruit .178874E-02 .115033E-02 1.55499 0.120 Chinese apple /U.S apple -.152746 .011837 -12.9039 0.000 Chinese apple /ROW apple -.220282 .016359 -13.4652 0.000 Chinese apple /Thai Pineapple .392100E-02 .141692E-02 2.76726 0.006 Chinese apple /Philippines Pineapple .771697E-02 .193201E-02 3.99426 0.000 Chinese apple /ROW Pineapple .972052E-03 .879485E-03 1.10525 0.269 Chinese apple /U.S grape .857184E-02 .476039E-02 1.80066 0.072 Chinese apple /Argentinean grape .144002E-02 .970316E-03 1.48408 0.138 Chinese apple /ROW grape .010240 .489075E-02 2.09372 0.036 Chinese apple /Israelis citrus .824731E-02 .228754E-02 3.60532 0.000 Chinese apple /Italian citrus .828148E-02 .254828E-02 3.24983 0.001 Chinese apple /ROW citrus .791190E-02 .166761E-02 4.74444 0.000 ROW apple/U.S orange .015155 .604306E-02 2.50788 0.012 ROW apple/Brazilian orange .414099 .101635 4.07439 0.000 ROW apple/ROW orange .505385E-02 .213536E-02 2.36675 0.018 ROW apple/U.S grapefruit .016042 .519421E-02 3.08834 0.002 ROW apple/Israelis grapefruit .173496E-02 .120416E-02 1.44081 0.150 ROW apple/ROW grapefruit .105479E-02 .679178E-03 1.55304 0.120 ROW apple /U.S apple -.090071 .739062E-02 -12.1873 0.000 ROW apple /Chinese apple -.097032 .720611E-02 -13.4652 0.000 ROW apple /Thai Pineapple .231214E-02 .933327E-03 2.47731 0.013 ROW apple /Philippines Pineapple .455055E-02 .133857E-02 3.39955 0.001 ROW apple /ROW Pineapple .573201E-03 .524794E-03 1.09224 0.275 ROW apple /U.S grape .505466E-02 .309022E-02 1.63569 0.102 ROW apple /Argentinean grape .849156E-03 .608877E-03 1.39463 0.163 ROW apple /ROW grape .603824E-02 .293661E-02 2.05620 0.040 ROW apple /Israelis citrus .486329E-02 .169898E-02 2.86248 0.004 ROW apple /Italian citrus .488344E-02 .169020E-02 2.88926 0.004 ROW apple /ROW citrus .466550E-02 .120581E-02 3.86918 0.000 Thai pineapple/U.S orange .018495 .789903E-02 2.34137 0.019 Thai pineapple/Brazilian orange .505342 .167599 3.01519 0.003 Thai pineapple /ROW orange .616742E-02 .286307E-02 2.15413 0.031 Thai pineapple /U.S grapefruit .019576 .720791E-02 2.71592 0.007 Thai pineapple /Israelis grapefruit .211725E-02 .152736E-02 1.38622 0.166 Thai pineapple /ROW grapefruit .128720E-02 .864674E-03 1.48866 0.137 Thai pineapple /U.S apple .024219 .010079 2.40294 0.016
151
Table D-2 Continued Products Estimates SE t-statistics p-value Thai pineapple /Chinese apple .026090 .942810E-02 2.76726 0.006 Thai pineapple /ROW apple .034927 .014099 2.47731 0.013 Thai pineapple /Philippines pineapple .240930 .238113E-02 101.183 0.000 Thai pineapple /ROW Pineapple .030348 .704226E-03 43.0944 0.000 Thai pineapple /U.S grape .616840E-02 .374974E-02 1.64502 0.100 Thai pineapple /Argentinean GR .103626E-02 .745014E-03 1.39093 0.164 Thai pineapple /ROW grape .736872E-02 .398682E-02 1.84827 0.065 Thai pineapple /Israelis citrus .593487E-02 .228441E-02 2.59798 0.009 Thai pineapple /Italian citrus .595946E-02 .237175E-02 2.51268 0.012 Thai pineapple /ROW citrus .569351E-02 .186220E-02 3.05741 0.002 Philippines pineapple/U.S orange .052498 .017195 3.05316 0.002 Philippines pineapple/Brazilian orange 1.43444 .262102 5.47281 0.000 Philippines pineapple /ROW orange .017506 .653647E-02 2.67828 0.007 Philippines pineapple /U.S grapefruit .055568 .013412 4.14302 0.000 Philippines pineapple /Israelis grapefruit .600990E-02 .396549E-02 1.51555 0.130 Philippines pineapple /ROW grapefruit .365378E-02 .221321E-02 1.65090 0.099 Philippines pineapple /U.S apple .068745 .020271 3.39126 0.001 Philippines pineapple /Chinese apple .074058 .018541 3.99426 0.000 Philippines pineapple /ROW apple .099141 .029163 3.39955 0.001 Philippines pineapple /Thai pineapple .347485 .343423E-02 101.183 0.000 Philippines pineapple /ROW pineapple .086145 .183038E-02 47.0638 0.000 Philippines pineapple /U.S grape .017509 .944625E-02 1.85357 0.064 Philippines pineapple /Argentinean grape .294147E-02 .192807E-02 1.52560 0.127 Philippines pineapple /ROW grape .020916 .933541E-02 2.24055 0.025 Philippines pineapple /Israelis other citrus .016846 .465127E-02 3.62189 0.000 Philippines pineapple /Italian citrus .016916 .498566E-02 3.39297 0.001 Philippines pineapple /ROW other citrus .016161 .278002E-02 5.81338 0.000 ROW pineapple/U.S orange .560620E-02 .522663E-02 1.07262 0.283 ROW pineapple/Brazilian orange .153183 .136893 1.11900 0.263 ROW pineapple /ROW orange .186951E-02 .177348E-02 1.05415 0.292 ROW pineapple /U.S grapefruit .593405E-02 .535430E-02 1.10828 0.268 ROW pineapple /Israelis grapefruit .641794E-03 .708595E-03 .905727 0.365 ROW pineapple /ROW grapefruit .390186E-03 .419020E-03 .931186 0.352 ROW pineapple /U.S apple .734127E-02 .676171E-02 1.08571 0.278 ROW pineapple /Chinese apple .790858E-02 .715546E-02 1.10525 0.269 ROW pineapple /ROW apple .010587 .969311E-02 1.09224 0.275 ROW pineapple /Thai pineapple .037108 .861080E-03 43.0944 0.000 ROW pineapple /Philippines pineapple .073032 .155177E-02 47.0638 0.000 ROW pineapple /U.S grape .186981E-02 .192404E-02 .971815 0.331 ROW pineapple /Argentinean grape .314118E-03 .342824E-03 .916265 0.360 ROW pineapple /ROW grape .223366E-02 .218496E-02 1.02229 0.307 ROW pineapple /Israelis citrus .179902E-02 .166480E-02 1.08062 0.280 ROW pineapple /Italian citrus .180647E-02 .166462E-02 1.08521 0.278 ROW pineapple /ROW citrus .172585E-02 .153880E-02 1.12156 0.262 U.S grape/U.S orange .711661E-02 .417504E-02 1.70456 0.088 U.S grape/Brazilian orange .194453 .094585 2.05586 0.040 U.S grape/ROW orange .237319E-02 .143251E-02 1.65667 0.098 U.S grape/U.S grapefruit .753279E-02 .413786E-02 1.82046 0.069 U.S grape/Israelis grapefruit .814705E-03 .684259E-03 1.19064 0.234 U.S grape/ROW grapefruit .495309E-03 .394443E-03 1.25572 0.209 U.S grape/U.S apple .931915E-02 .532257E-02 1.75087 0.080 U.S grape/Chinese apple .010039 .557535E-02 1.80066 0.072 U.S grape/ROW apple .013440 .821645E-02 1.63569 0.102 U.S grape/Thai Pineapple .108574E-02 .660014E-03 1.64502 0.100
152
Table D-2 Continued Products Estimates SE t-statistics p-value U.S grape /Philippines Pineapple .213685E-02 .115283E-02 1.85357 0.064 U.S grape /ROW Pineapple .269164E-03 .276971E-03 .971815 0.331 U.S grape /Argentinean grape .039092 .378769E-03 103.209 0.000 U.S grape /ROW grape .277981 .201337E-02 138.067 0.000 U.S grape /Israelis citrus .228371E-02 .128703E-02 1.77439 0.076 U.S grape /Italian citrus .229317E-02 .129962E-02 1.76448 0.078 U.S grape /ROW citrus .219083E-02 .109370E-02 2.00314 0.045 Argentinean grape/U.S orange .810137E-02 .560572E-02 1.44520 0.148 Argentinean grape/Brazilian orange .221360 .139481 1.58703 0.113 Argentinean grape/ROW orange .270158E-02 .190366E-02 1.41915 0.156 Argentinean grape/U.S grapefruit .857513E-02 .559170E-02 1.53355 0.125 Argentinean grape/Israelis grapefruit .927439E-03 .851425E-03 1.08928 0.276 Argentinean grape/ROW grapefruit .563847E-03 .496317E-03 1.13606 0.256 Argentinean grape/U.S apple .010609 .723580E-02 1.46614 0.143 Argentinean grape/Chinese apple .011428 .770073E-02 1.48408 0.138 Argentinean grape/ROW apple .015299 .010970 1.39463 0.163 Argentinean grape/Thai pineapple .123597E-02 .888598E-03 1.39093 0.164 Argentinean grape /Philippines pineapple. .243254E-02 .159448E-02 1.52560 0.127 Argentinean grape /ROW Pineapple .306410E-03 .334412E-03 .916265 0.360 Argentinean grape /U.S grape .264899 .256663E-02 103.209 0.000 Argentinean grape /ROW grape .316446 .274614E-02 115.233 0.000 Argentinean grape /Israelis other citrus .259971E-02 .177406E-02 1.46541 0.143 Argentinean grape /Italian citrus .261048E-02 .178099E-02 1.46575 0.143 Argentinean grape /ROW other CT .249399E-02 .158467E-02 1.57382 0.116 ROW grape/U.S orange .815410E-02 .432291E-02 1.88625 0.059 ROW grape/Brazilian orange .222801 .094112 2.36740 0.018 ROW grape/ROW orange .271916E-02 .149192E-02 1.82259 0.068 ROW grape/U.S grapefruit .863095E-02 .400492E-02 2.15509 0.031 ROW grape/Israelis grapefruit .933476E-03 .726853E-03 1.28427 0.199 ROW grape/ROW grapefruit .567517E-03 .417173E-03 1.36039 0.174 ROW grape/U.S apple .010678 .534914E-02 1.99616 0.046 ROW grape/Chinese apple .011503 .549400E-02 2.09372 0.036 ROW grape/ROW apple .015399 .748901E-02 2.05620 0.040 ROW grape/Thai Pineapple .124402E-02 .673073E-03 1.84827 0.065 ROW grape /Philippines Pineapple .244837E-02 .109275E-02 2.24055 0.025 ROW grape /ROW Pineapple .308404E-03 .301680E-03 1.02229 0.307 ROW grape /U.S grape .266623 .193111E-02 138.067 0.000 ROW grape /Argentinean grape .044791 .388703E-03 115.233 0.000 ROW grape /Israelis citrus .261664E-02 .124001E-02 2.11018 0.035 ROW grape /Italian citrus .262747E-02 .121925E-02 2.15498 0.031 ROW grape /ROW citrus .251022E-02 .106985E-02 2.34632 0.019 Israelis citrus/U.S orange .019268 .696237E-02 2.76751 0.006 Israelis citrus /Brazilian orange .526486 .113205 4.65073 0.000 Israelis citrus /ROW orange .642547E-02 .255013E-02 2.51967 0.012 Israelis citrus /U.S grapefruit .020395 .587965E-02 3.46878 0.001 Israelis citrus /Israelis grapefruit .220583E-02 .149909E-02 1.47145 0.141 Israelis citrus /ROW grapefruit .134106E-02 .843192E-03 1.59046 0.112 Israelis citrus /U.S apple .025232 .829740E-02 3.04093 0.002 Israelis citrus /Chinese apple .027182 .753933E-02 3.60532 0.000 Israelis citrus /ROW apple .036388 .012712 2.86248 0.004 Israelis citrus /Thai pineapple .293966E-02 .113151E-02 2.59798 0.009 Israelis citrus /Philippines pineapple .578558E-02 .159739E-02 3.62189 0.000 Israelis citrus /ROW Pineapple .728769E-03 .674397E-03 1.08062 0.280 Israelis citrus /U.S grape .642650E-02 .362180E-02 1.77439 0.076
153
Table D-2 Continued Products Estimates SE t-statistics p-value Israelis citrus /Argentinean GR .107962E-02 .736736E-03 1.46541 0.143 Israelis citrus /ROW grape .767703E-02 .363810E-02 2.11018 0.035 Israelis citrus /Italian citrus .088372 .209589E-02 42.1645 0.000 Israelis citrus /ROW citrus .084428 .126626E-02 66.6754 0.000 Italian citrus/U.S orange .024731 .953356E-02 2.59408 0.009 Italian citrus /Brazilian orange .675741 .162400 4.16096 0.000 Italian citrus /ROW orange .824704E-02 .343402E-02 2.40157 0.016 Italian citrus /U.S grapefruit .026177 .744891E-02 3.51422 0.000 Italian citrus /Israelis grapefruit .283117E-02 .191769E-02 1.47635 0.140 Italian citrus /ROW grapefruit .172124E-02 .107205E-02 1.60555 0.108 Italian citrus /U.S apple .032385 .011655 2.77869 0.005 Italian citrus /Chinese apple .034887 .010735 3.24983 0.001 Italian citrus /ROW apple .046704 .016165 2.88926 0.004 Italian citrus /Thai pineapple .377303E-02 .150159E-02 2.51268 0.012 Italian citrus /Philippines pineapple .742574E-02 .218857E-02 3.39297 0.001 Italian citrus /ROW Pineapple .935369E-03 .861923E-03 1.08521 0.278 Italian citrus /U.S grape .824836E-02 .467466E-02 1.76448 0.078 Italian citrus /Argentinean grape .138568E-02 .945376E-03 1.46575 0.143 Italian citrus /ROW grape .985341E-02 .457238E-02 2.15498 0.031 Italian citrus /Israelis citrus .112957 .267896E-02 42.1645 0.000 Italian citrus /ROW citrus .108363 .180921E-02 59.8952 0.000 ROW citrus/U.S orange .016297 .472942E-02 3.44581 0.001 ROW citrus /Brazilian orange .445287 .081568 5.45908 0.000 ROW citrus /ROW orange .543448E-02 .181890E-02 2.98779 0.003 ROW citrus /U.S grapefruit .017250 .326370E-02 5.28533 0.000 ROW citrus /Israelis grapefruit .186563E-02 .119143E-02 1.56587 0.117 ROW citrus /ROW grapefruit .113423E-02 .660248E-03 1.71789 0.086 ROW citrus /U.S apple .021340 .569251E-02 3.74885 0.000 ROW citrus /Chinese apple .022989 .484556E-02 4.74444 0.000 ROW citrus /ROW apple .030776 .795412E-02 3.86918 0.000 ROW citrus /Thai pineapple .248628E-02 .813200E-03 3.05741 0.002 ROW citrus /Philippines pineapple .489328E-02 .841728E-03 5.81338 0.000 ROW citrus /ROW Pineapple .616372E-03 .549568E-03 1.12156 0.262 ROW citrus /U.S grape .543535E-02 .271342E-02 2.00314 0.045 ROW citrus /Argentinean GR .913111E-03 .580188E-03 1.57382 0.116 ROW citrus /ROW grape .649302E-02 .276732E-02 2.34632 0.019 ROW citrus /Israelis citrus .074434 .111637E-02 66.6754 0.000 ROW citrus /Italian citrus .074743 .124789E-02 59.8952 0.000 Table D-3 Compensated price elasticities of fruit juices in the block-wise dependent non-
uniform substitute Rotterdam model Products Estimates SD t-statistic p-value U.S orange /Brazilian orange 1.28083 .282607 4.53218 0.000 U.S orange /ROW orange .113355 .093998 1.20594 0.228 U.S orange /U.S grapefruit .011616 .017615 .659440 0.510 U.S orange /Israelis grapefruit -.000639 .001230 -.519525 0.603 U.S orange /ROW grapefruit .000995 .001601 .622045 0.534 U.S orange /U.S apple -.035930 .019872 -1.80811 0.071 U.S orange /Chinese apple -.034077 .018055 -1.88738 0.059 U.S orange /ROW apple -.046112 .026299 -1.75339 0.080 U.S orange /Thai pineapple .006567 .004082 1.60853 0.108 U.S orange /Philippines pineapple .010245 .006032 1.69847 0.089 U.S orange /ROW pineapple .000889 .001460 .609119 0.542
154
Table D-3 Continued Products Estimates SD t-statistic p-value U.S orange /U.S grape -.000988 .008365 -.118156 0.906 U.S orange /Argentinean grape -.000080 .000693 -.116462 0.907 U.S orange /ROW grape -.001671 .014148 -.118160 0.906 U.S orange /Israelis citrus -.001252 .004129 -.303321 0.762 U.S orange /Italian citrus -.001487 .004859 -.306142 0.759 U.S orange /ROW citrus -.001566 .005051 -.310128 0.756 Brazilian orange/U.S orange .365024 .080540 4.53218 0.000 Brazilian orange /ROW orange .103392 .030470 3.39330 0.001 Brazilian orange /U.S grapefruit .068377 .103344 .661641 0.508 Brazilian orange /Israelis grapefruit -.003762 .007228 -.520477 0.603 Brazilian orange /ROW grapefruit .005862 .009350 .626958 0.531 Brazilian orange /U.S apple -.211498 .094338 -2.24193 0.025 Brazilian orange /Chinese apple -.200590 .086278 -2.32491 0.020 Brazilian orange /ROW apple -.271428 .120918 -2.24472 0.025 Brazilian orange /Thai Pineapple .038656 .020728 1.86492 0.062 Brazilian orange /Philippines pineapple .060308 .030041 2.00754 0.045 Brazilian orange /ROW Pineapple .005237 .008429 .621271 0.534 Brazilian orange /U.S grape -.005818 .049042 -.118643 0.906 Brazilian orange /Argentinean grape -.00475 .004070 -.116820 0.907 Brazilian orange /ROW grape -.009839 .082944 -.118634 0.906 Brazilian orange /Israelis citrus -.007372 .024064 -.306361 0.759 Brazilian orange /Italian citrus -.008757 .028312 -.309320 0.757 Brazilian orange /ROW citrus -.009221 .029489 -.312697 0.755 ROW orange/U.S orange .253434 .210156 1.20594 0.228 ROW orange/Brazilian orange .811115 .239034 3.39330 0.001 ROW orange /U.S grapefruit .003046 .006228 .489167 0.625 ROW orange /Israelis grapefruit -.001673 .000396 -.423284 0.672 ROW orange /ROW grapefruit .000261 .000549 .475047 0.635 ROW orange /U.S apple -.009424 .013975 -.674378 0.500 ROW orange /Chinese apple -.008938 .013722 -.651417 0.515 ROW orange /ROW apple -.012095 .018411 -.656940 0.511 ROW orange /Thai pineapple .001722 .002676 .643703 0.520 ROW orange /Philippines pineapple .002687 .004137 .649538 0.516 ROW orange /ROW pineapple .002333 .000506 .461089 0.645 ROW orange /U.S grape -.000259 .002216 -.116958 0.907 ROW orange /Argentinean grape -.000021 .000183 -.115402 0.908 ROW orange /ROW grape -.000043 .003749 -.116954 0.907 ROW orange /Israelis citrus -.000328 .001178 -.278836 0.780 ROW orange /Italian citrus -.000390 .001387 -.281230 0.779 ROW orange /ROW citrus -.000410 .001447 -.283820 0.777 U.S grapefruit/U.S orange .010409 .015785 .659440 0.510 U.S grapefruit/Brazilian orange .214997 .324945 .661641 0.508 U.S grapefruit/ROW orange .001221 .002496 .489167 0.625 U.S grapefruit/Israelis grapefruit .148987 .043520 3.42342 0.001 U.S grapefruit /ROW grapefruit .061095 .026486 2.30672 0.021 U.S grapefruit /U.S apple .097839 .113102 .865052 0.387 U.S grapefruit /Chinese apple .092793 .107689 .861671 0.389 U.S grapefruit /ROW apple .125562 .146899 .854753 0.393 U.S grapefruit /Thai Pineapple -.248714 .056484 -4.40328 0.000 U.S grapefruit /Philippines Pineapple -.388024 .068007 -5.70563 0.000 U.S grapefruit /ROW Pineapple -.033696 .049471 -.681124 0.496 U.S grapefruit /U.S grape .126374 .100765 1.25415 0.210 U.S grapefruit /Argentinean grape .010328 .024192 .426929 0.669
155
Table D-3 Continued Products Estimates SD t-statistic p-value U.S grapefruit /ROW grape .213719 .157773 1.35460 0.176 U.S grapefruit /Israelis citrus -.054840 .047911 -1.14461 0.252 U.S grapefruit /Italian citrus -.065144 .054519 -1.19489 0.232 U.S grapefruit /ROW citrus -.068593 .061144 -1.12182 0.262 Israelis grapefruit/U.S orange -.001782 .003430 -.519525 0.603 Israelis grapefruit/Brazilian orange -.036815 .070734 -.520477 0.603 Israelis grapefruit/ROW orange -.000209 .000494 -.423284 0.672 Israelis grapefruit/U.S grapefruit .463688 .135446 3.42342 0.001 Israelis grapefruit/ROW grapefruit .095547 .056916 1.67873 0.093 Israelis grapefruit /U.S apple -.016754 .028293 -.592153 0.554 Israelis grapefruit /Chinese apple -.015889 .026757 -.593844 0.553 Israelis grapefruit /ROW apple -.021501 .036321 -.591964 0.554 Israelis grapefruit /Thai Pineapple .042589 .052997 .803605 0.422 Israelis grapefruit /Philippines Pineapple .066444 .083515 .795594 0.426 Israelis grapefruit /ROW Pineapple .005769 .010845 .532046 0.595 Israelis grapefruit /U.S grape -.021640 .031283 -.691754 0.489 Israelis grapefruit /Argentinean grape -.001768 .004719 -.374713 0.708 Israelis grapefruit /ROW grape -.036597 .052753 -.693737 0.488 Israelis grapefruit /Israelis citrus .009390 .014119 .665110 0.506 Israelis grapefruit /Italian citrus .011155 .016583 .672662 0.501 Israelis grapefruit /ROW citrus .011746 .018159 .646836 0.518 ROW grapefruit/U.S orange .006457 .010381 .622045 0.534 ROW grapefruit/Brazilian orange .133374 .212733 .626958 0.531 ROW grapefruit/ROW orange .000757 .001594 .475047 0.635 ROW grapefruit/U.S grapefruit .442059 .191639 2.30672 0.021 ROW grapefruit/Israelis grapefruit .222133 .132322 1.67873 0.093 ROW grapefruit /U.S apple .060695 .075145 .807705 0.419 ROW grapefruit /Chinese apple .057564 .071635 .803577 0.422 ROW grapefruit /ROW apple .077893 .097274 .800758 0.423 ROW grapefruit /Thai Pineapple -.154291 .079717 -1.93548 0.053 ROW grapefruit /Philippines Pineapple -.240712 .118528 -2.03086 0.042 ROW grapefruit /ROW Pineapple -.020903 .032333 -.646503 0.518 ROW grapefruit /U.S grape .078397 .070842 1.10664 0.268 ROW grapefruit /Argentinean grape .006407 .015265 .419726 0.675 ROW grapefruit /ROW grape .132582 .112827 1.17509 0.240 ROW grapefruit /Israelis citrus -.034020 .033901 -1.00352 0.316 ROW grapefruit /Italian citrus -.040412 .039002 -1.03617 0.300 ROW grapefruit /ROW citrus -.042552 .043290 -.982962 0.326 U.S apple/U.S orange -.045837 .025351 -1.80811 0.071 U.S apple/Brazilian orange -.946737 .422287 -2.24193 0.025 U.S apple/ROW orange -.005377 .007974 -.674378 0.500 U.S apple/U.S grapefruit .139287 .161015 .865052 0.387 U.S apple/Israelis grapefruit -.007663 .012942 -.592153 0.554 U.S apple/ROW grapefruit .011942 .014785 .807705 0.419 U.S apple /Chinese apple .079917 .151156 .528707 0.597 U.S apple /ROW apple -.330413 .239359 -1.38041 0.167 U.S apple /Thai Pineapple .109245 .039794 2.74523 0.006 U.S apple /Philippines Pineapple .170435 .056741 3.00377 0.003 U.S apple /ROW Pineapple .014800 .022252 .665123 0.506 U.S apple /U.S grape -.028422 .074801 -.379961 0.704 U.S apple /Argentinean grape -.002322 .007904 -.293840 0.769 U.S apple /ROW grape -.048065 .126125 -.381094 0.703 U.S apple /Israelis citrus .064651 .041515 1.55729 0.119
156
Table D-3 Continued Products Estimates SD t-statistic p-value U.S apple /Italian citrus .076798 .046045 1.66791 0.095 U.S apple /ROW citrus .080864 .056066 1.44230 0.149 Chinese apple/U.S orange -.033917 .017970 -1.88738 0.059 Chinese apple/Brazilian orange -.700527 .301313 -2.32491 0.020 Chinese apple/ROW orange -.003979 .006108 -.651417 0.515 Chinese apple/U.S grapefruit .103063 .119609 .861671 0.389 Chinese apple/Israelis grapefruit -.005670 .009548 -.593844 0.553 Chinese apple/ROW grapefruit .008836 .010996 .803577 0.422 Chinese apple /U.S apple .062349 .117928 .528707 0.597 Chinese apple /ROW apple .058946 .184747 .319066 0.750 Chinese apple /Thai Pineapple .080834 .028601 2.82632 0.005 Chinese apple /Philippines Pineapple .126112 .040593 3.10671 0.002 Chinese apple /ROW Pineapple .010951 .016427 .666664 0.505 Chinese apple /U.S grape -.021030 .055556 -.378537 0.705 Chinese apple /Argentinean grape -.001718 .005877 -.292416 0.770 Chinese apple /ROW grape -.035565 .093662 -.379720 0.704 Chinese apple /Israelis citrus .047838 .030630 1.56179 0.118 Chinese apple /Italian citrus .056826 .033945 1.67407 0.094 Chinese apple /ROW citrus .059835 .042297 1.41462 0.157 ROW apple/U.S orange -.020216 .011530 -1.75339 0.080 ROW apple/Brazilian orange -.417547 .186013 -2.24472 0.025 ROW apple/ROW orange -.002371 .003610 -.656940 0.511 ROW apple/U.S grapefruit .061431 .071869 .854753 0.393 ROW apple/Israelis grapefruit -.003379 .005709 -.591964 0.554 ROW apple/ROW grapefruit .005266 .006577 .800758 0.423 ROW apple /U.S apple -.113549 .082258 -1.38041 0.167 ROW apple /Chinese apple .025965 .081379 .319066 0.750 ROW apple /Thai Pineapple .048181 .017491 2.75466 0.006 ROW apple /Philippines Pineapple .075168 .024951 3.01265 0.003 ROW apple /ROW Pineapple .006527 .009820 .664696 0.506 ROW apple /U.S grape -.012535 .033141 -.378236 0.705 ROW apple /Argentinean grape -.001024 .003517 -.291212 0.771 ROW apple /ROW grape -.021199 .055731 -.380373 0.704 ROW apple /Israelis citrus .028513 .018413 1.54854 0.121 ROW apple /Italian citrus .033871 .020149 1.68106 0.093 ROW apple /ROW citrus .035664 .024413 1.46085 0.144 Thai pineapple/U.S orange .043491 .027038 1.60853 0.108 Thai pineapple/Brazilian orange .898281 .481673 1.86492 0.062 Thai pineapple /ROW orange .005102 .007926 .643703 0.520 Thai pineapple /U.S grapefruit -1.83810 .417438 -4.40328 0.000 Thai pineapple /Israelis grapefruit .101132 .125848 .803605 0.422 Thai pineapple /ROW grapefruit -.157592 .081423 -1.93548 0.053 Thai pineapple /U.S apple .567116 .206582 2.74523 0.006 Thai pineapple /Chinese apple .537867 .190306 2.82632 0.005 Thai pineapple /ROW apple .727813 .264211 2.75466 0.006 Thai pineapple /Philippines pineapple -.061702 .134429 -.458997 0.646 Thai pineapple /ROW Pineapple -.080203 .123655 -.648599 0.517 Thai pineapple /U.S grape .238571 .175750 1.35745 0.175 Thai pineapple /Argentinean GR .019498 .042904 .454451 0.650 Thai pineapple /ROW grape .403463 .273516 1.47510 0.140 Thai pineapple /Israelis citrus .171808 .086405 1.98839 0.047 Thai pineapple /Italian citrus .204089 .100443 2.03188 0.042 Thai pineapple /ROW citrus .214894 .120744 1.77975 0.075
157
Table D-3 Continued Products Estimates SD t-statistic p-value Philippines pineapple/U.S orange .097859 .057616 1.69847 0.089 Philippines pineapple/Brazilian orange 2.02123 1.00682 2.00754 0.045 Philippines pineapple /ROW orange .011481 .017676 .649538 0.516 Philippines pineapple /U.S grapefruit -4.13593 .724885 -5.70563 0.000 Philippines pineapple /Israelis grapefruit .227558 .286022 .795594 0.426 Philippines pineapple /ROW grapefruit -.354599 .174606 -2.03086 0.042 Philippines pineapple /U.S apple 1.27607 .424825 3.00377 0.003 Philippines pineapple /Chinese apple 1.21026 .389563 3.10671 0.002 Philippines pineapple /ROW apple 1.63766 .543595 3.01265 0.003 Philippines pineapple /Thai pineapple -.088991 .193882 -.458997 0.646 Philippines pineapple /ROW pineapple .078186 .187860 .416193 0.677 Philippines pineapple /U.S grape .536813 .391140 1.37243 0.170 Philippines pineapple /Argentinean grape .043872 .096498 .454640 0.649 Philippines pineapple /ROW grape .907838 .607558 1.49424 0.135 Philippines pineapple /Israelis other citrus .386587 .185970 2.07876 0.038 Philippines pineapple /Italian citrus .459223 .216054 2.12550 0.034 Philippines pineapple /ROW other citrus .483537 .263230 1.83694 0.066 ROW pineapple/U.S orange .007204 .011828 .609119 0.542 ROW pineapple/Brazilian orange .148805 .239517 .621271 0.534 ROW pineapple /ROW orange .000845 .001833 .461089 0.645 ROW pineapple /U.S grapefruit -.304491 .447042 -.681124 0.496 ROW pineapple /Israelis grapefruit .016753 .031488 .532046 0.595 ROW pineapple /ROW grapefruit -.026106 .040380 -.646503 0.518 ROW pineapple /U.S apple .093946 .141246 .665123 0.506 ROW pineapple /Chinese apple .089101 .133651 .666664 0.505 ROW pineapple /ROW apple .120566 .181386 .664696 0.506 ROW pineapple /Thai pineapple -.098066 .151197 -.648599 0.517 ROW pineapple /Philippines pineapple .066285 .159265 .416193 0.677 ROW pineapple /U.S grape .039521 .065703 .601504 0.548 ROW pineapple /Argentinean grape .003229 .008547 .377864 0.706 ROW pineapple /ROW grape .066836 .109084 .612702 0.540 ROW pineapple /Israelis citrus .028461 .043343 .656647 0.511 ROW pineapple /Italian citrus .033808 .051000 .662904 0.507 ROW pineapple /ROW citrus .035598 .054427 .654055 0.513 U.S grape/U.S orange -.001152 .009751 -.118156 0.906 U.S grape/Brazilian orange -.023799 .200591 -.118643 0.906 U.S grape/ROW orange -.000135 .001155 -.116958 0.907 U.S grape/U.S grapefruit .164391 .131078 1.25415 0.210 U.S grape/Israelis grapefruit -.009044 .013075 -.691754 0.489 U.S grape/ROW grapefruit .014094 .012736 1.10664 0.268 U.S grape/U.S apple -.025970 .068349 -.379961 0.704 U.S grape/Chinese apple -.024630 .065068 -.378537 0.705 U.S grape/ROW apple -.033329 .088116 -.378236 0.705 U.S grape/Thai Pineapple .041992 .030935 1.35745 0.175 U.S grape /Philippines Pineapple .065513 .047735 1.37243 0.170 U.S grape /ROW Pineapple .005689 .009458 .601504 0.548 U.S grape /Argentinean grape .107867 .068279 1.57981 0.114 U.S grape /ROW grape .173388 .153024 1.13308 0.257 U.S grape /Israelis citrus .084064 .043558 1.92991 0.054 U.S grape /Italian citrus .099858 .049998 1.99725 0.046 U.S grape /ROW citrus .105145 .057393 1.83201 0.067 Argentinean grape/U.S orange -.000633 .005479 -.116462 0.907 Argentinean grape/Brazilian orange -.013180 .112821 -.116820 0.907 Argentinean grape/ROW orange -.000074 .000648 -.115402 0.908
158
Table D-3 Continued Products Estimates SD t-statistic p-value Argentinean grape/U.S grapefruit .091040 .213243 .426929 0.669 Argentinean grape/Israelis grapefruit -.005008 .013367 -.374713 0.708 Argentinean grape/ROW grapefruit .007805 .018596 .419726 0.675 Argentinean grape/U.S apple -.014382 .048945 -.293840 0.769 Argentinean grape/Chinese apple -.013640 .046647 -.292416 0.770 Argentinean grape/ROW apple -.018457 .063381 -.291212 0.771 Argentinean grape/Thai pineapple .023255 .051172 .454451 0.650 Argentinean grape /Philippines pineapple. .036281 .079802 .454640 0.649 Argentinean grape /ROW Pineapple .003150 .008338 .377864 0.706 Argentinean grape /U.S grape .730936 .462673 1.57981 0.114 Argentinean grape /ROW grape .098332 .529644 .185657 0.853 Argentinean grape /Israelis other citrus .046554 .100989 .460984 0.645 Argentinean grape /Italian citrus .055302 .119983 .460912 0.645 Argentinean grape /ROW other CT .058230 .125553 .463786 0.643 ROW grape/U.S orange -.001869 .015818 -.118160 0.906 ROW grape/Brazilian orange -.038603 .325396 -.118634 0.906 ROW grape/ROW orange -.000219 .001874 -.116954 0.907 ROW grape/U.S grapefruit .266654 .196850 1.35460 0.176 ROW grape/Israelis grapefruit -.014671 .021148 -.693737 0.488 ROW grape/ROW grapefruit .022862 .019455 1.17509 0.240 ROW grape/U.S apple -.042125 .110537 -.381094 0.703 ROW grape/Chinese apple -.039952 .105215 -.379720 0.704 ROW grape/ROW apple -.054061 .142127 -.380373 0.704 ROW grape/Thai Pineapple .068114 .046176 1.47510 0.140 ROW grape /Philippines Pineapple .106267 .071118 1.49424 0.135 ROW grape /ROW Pineapple .009228 .015061 .612702 0.540 ROW grape /U.S grape .166304 .146772 1.13308 0.257 ROW grape /Argentinean grape .013918 .074968 .185657 0.853 ROW grape /Israelis citrus .136357 .059527 2.29066 0.022 ROW grape /Italian citrus .161977 .069118 2.34348 0.019 ROW grape /ROW citrus .170553 .085054 2.00523 0.045 Israelis citrus/U.S orange -.004108 .013545 -.303321 0.762 Israelis citrus /Brazilian orange -.084856 .276981 -.306361 0.759 Israelis citrus /ROW orange -.000481 .001728 -.278836 0.780 Israelis citrus /U.S grapefruit -.200748 .175385 -1.14461 0.252 Israelis citrus /Israelis grapefruit .011045 .016606 .665110 0.506 Israelis citrus /ROW grapefruit -.017211 .017151 -1.00352 0.316 Israelis citrus /U.S apple .166238 .106748 1.55729 0.119 Israelis citrus /Chinese apple .157664 .100951 1.56179 0.118 Israelis citrus /ROW apple .213343 .137771 1.54854 0.121 Israelis citrus /Thai pineapple .085100 .042798 1.98839 0.047 Israelis citrus /Philippines pineapple .132766 .063868 2.07876 0.038 Israelis citrus /ROW Pineapple .011529 .017558 .656647 0.511 Israelis citrus /U.S grape .236560 .122576 1.92991 0.054 Israelis citrus /Argentinean GR .019333 .041939 .460984 0.645 Israelis citrus /ROW grape .400062 .174649 2.29066 0.022 Israelis citrus /Italian citrus .103190 .124971 .825712 0.409 Israelis citrus /ROW citrus .111639 .084809 1.31636 0.188 Italian citrus/U.S orange -.006237 .020376 -.306142 0.759 Italian citrus /Brazilian orange -.128842 .416532 -.309320 0.757 Italian citrus /ROW orange -.000731 .002602 -.281230 0.779 Italian citrus /U.S grapefruit -.304807 .255092 -1.19489 0.232 Italian citrus /Israelis grapefruit .016770 .024931 .672662 0.501 Italian citrus /ROW grapefruit -.026133 .025221 -1.03617 0.300
159
Table D-3 Continued Products Estimates SD t-statistic p-value Italian citrus /U.S apple .252409 .151332 1.66791 0.095 Italian citrus /Chinese apple .239391 .142999 1.67407 0.094 Italian citrus /ROW apple .323931 .192694 1.68106 0.093 Italian citrus /Thai pineapple .129212 .063592 2.03188 0.042 Italian citrus /Philippines pineapple .201586 .094842 2.12550 0.034 Italian citrus /ROW Pineapple .017506 .026407 .662904 0.507 Italian citrus /U.S grape .359183 .179839 1.99725 0.046 Italian citrus /Argentinean grape .029355 .063689 .460912 0.645 Italian citrus /ROW grape .607437 .259203 2.34348 0.019 Italian citrus /Israelis citrus .131897 .159737 .825712 0.409 Italian citrus /ROW citrus -.109788 .094188 -1.16563 0.244 ROW citrus/U.S orange -.004530 .014608 -.310128 0.756 ROW citrus /Brazilian orange -.093573 .299245 -.312697 0.755 ROW citrus /ROW orange -.000531 .001872 -.283820 0.777 ROW citrus /U.S grapefruit -.221370 .197331 -1.12182 0.262 ROW citrus /Israelis grapefruit .012180 .018830 .646836 0.518 ROW citrus /ROW grapefruit -.018979 .019308 -.982962 0.326 ROW citrus /U.S apple .183315 .127099 1.44230 0.149 ROW citrus /Chinese apple .173861 .122903 1.41462 0.157 ROW citrus /ROW apple .235259 .161043 1.46085 0.144 ROW citrus /Thai pineapple .093842 .052727 1.77975 0.075 ROW citrus /Philippines pineapple .146404 .079700 1.83694 0.066 ROW citrus /ROW Pineapple .012714 .019438 .654055 0.513 ROW citrus /U.S grape .260861 .142391 1.83201 0.067 ROW citrus /Argentinean GR .021319 .045968 .463786 0.643 ROW citrus /ROW grape .441159 .220004 2.00523 0.045 ROW citrus /Israelis citrus .098424 .074770 1.31636 0.188 ROW citrus /Italian citrus -.075726 .064965 -1.16563 0.244 Table D-4 Compensated price elasticities of fruit juices in the block-wise dependent
uniform substitute Rotterdam model Products Estimates SD t-statistic p-value U.S orange /Brazilian orange 1.36236 .137491 9.90867 0.000 U.S orange /ROW orange .012980 .282909E-02 4.58790 0.000 U.S orange /U.S grapefruit .961063E-02 .011016 .872392 0.383 U.S orange /Israelis grapefruit .777966E-03 .100635E-02 .773058 0.439 U.S orange /ROW grapefruit .494356E-03 .627476E-03 .787848 0.431 U.S orange /U.S apple -.689830E-02 .010577 -.652216 0.514 U.S orange /Chinese apple -.144385E-02 .225587E-02 -.640041 0.522 U.S orange /ROW apple -.760172E-02 .011820 -.643096 0.520 U.S orange /Thai pineapple .160928E-02 .304587E-02 .528347 0.597 U.S orange /Philippines pineapple .145979E-02 .274833E-02 .531158 0.595 U.S orange /ROW pineapple .861375E-04 .268313E-03 .321033 0.748 U.S orange /U.S grape -.628742E-02 .624819E-02 -1.00628 0.314 U.S orange /Argentinean grape -.134931E-02 .138383E-02 -.975055 0.330 U.S orange /ROW grape -.882131E-02 .849967E-02 -1.03784 0.299 U.S orange /Israelis citrus .618641E-02 .409829E-02 1.50951 0.131 U.S orange /Italian citrus .606382E-02 .422129E-02 1.43648 0.151 U.S orange /ROW citrus .377466E-02 .260870E-02 1.44695 0.148 Brazilian orange/U.S orange .388260 .039184 9.90867 0.000 Brazilian orange /ROW orange .120813 .014278 8.46124 0.000 Brazilian orange /U.S grapefruit .089455 .099507 .898986 0.369 Brazilian orange /Israelis grapefruit .724126E-02 .911753E-02 .794213 0.427
160
Table D-4 Continued Products Estimates SD t-statistic p-value Brazilian orange /ROW grapefruit .460144E-02 .567576E-02 .810717 0.418 Brazilian orange /U.S apple -.064209 .095645 -.671324 0.502 Brazilian orange /Chinese apple -.013439 .020426 -.657936 0.511 Brazilian orange /ROW apple -.070756 .106205 -.666225 0.505 Brazilian orange /Thai Pineapple .014979 .027921 .536485 0.592 Brazilian orange /Philippines pineapple .013588 .025194 .539332 0.590 Brazilian orange /ROW Pineapple .801763E-03 .248616E-02 .322491 0.747 Brazilian orange /U.S grape -.058523 .054171 -1.08035 0.280 Brazilian orange /Argentinean grape -.012559 .012192 -1.03017 0.303 Brazilian orange /ROW grape -.082108 .072746 -1.12870 0.259 Brazilian orange /Israelis citrus .057583 .033956 1.69580 0.090 Brazilian orange /Italian citrus .056442 .034976 1.61371 0.107 Brazilian orange /ROW citrus .035134 .022236 1.58006 0.114 ROW orange/U.S orange .029019 .632515E-02 4.58790 0.000 ROW orange/Brazilian orange .947781 .112014 8.46124 0.000 ROW orange /U.S grapefruit .668603E-02 .771332E-02 .866816 0.386 ROW orange /Israelis grapefruit .541224E-03 .703277E-03 .769574 0.442 ROW orange /ROW grapefruit .343919E-03 .438182E-03 .784878 0.433 ROW orange /U.S apple -.479909E-02 .749026E-02 -.640710 0.522 ROW orange /Chinese apple -.100447E-02 .159920E-02 -.628109 0.530 ROW orange /ROW apple -.528845E-02 .835891E-02 -.632672 0.527 ROW orange /Thai pineapple .111956E-02 .213331E-02 .524800 0.600 ROW orange /Philippines pineapple .101557E-02 .192494E-02 .527584 0.598 ROW orange /ROW pineapple .599251E-04 .187108E-03 .320271 0.749 ROW orange /U.S grape -.437411E-02 .445997E-02 -.980749 0.327 ROW orange /Argentinean grape -.938705E-03 .985018E-03 -.952982 0.341 ROW orange /ROW grape -.613691E-02 .608998E-02 -1.00771 0.314 ROW orange /Israelis citrus .430384E-02 .295876E-02 1.45461 0.146 ROW orange /Italian citrus .421855E-02 .303273E-02 1.39101 0.164 ROW orange /ROW citrus .262600E-02 .188159E-02 1.39563 0.163 U.S grapefruit/U.S orange .861208E-02 .987181E-02 .872392 0.383 U.S grapefruit/Brazilian orange .281275 .312880 .898986 0.369 U.S grapefruit/ROW orange .267979E-02 .309153E-02 .866816 0.386 U.S grapefruit/Israelis grapefruit .116063 .989322E-03 117.316 0.000 U.S grapefruit /ROW grapefruit .073752 .583666E-03 126.360 0.000 U.S grapefruit /U.S apple .324087 .124901 2.59474 0.009 U.S grapefruit /Chinese apple .067833 .033093 2.04976 0.040 U.S grapefruit /ROW apple .357134 .150466 2.37352 0.018 U.S grapefruit /Thai Pineapple -.350699 .058706 -5.97383 0.000 U.S grapefruit /Philippines Pineapple -.318124 .052397 -6.07142 0.000 U.S grapefruit /ROW Pineapple -.018771 .047162 -.398023 0.691 U.S grapefruit /U.S grape .082947 .091900 .902588 0.367 U.S grapefruit /Argentinean grape .017801 .020461 .870002 0.384 U.S grapefruit /ROW grape .116376 .124390 .935576 0.349 U.S grapefruit /Israelis citrus -.062922 .055330 -1.13723 0.255 U.S grapefruit /Italian citrus -.061675 .053788 -1.14664 0.252 U.S grapefruit /ROW citrus -.038392 .034046 -1.12765 0.259 Israelis grapefruit/U.S orange .216968E-02 .280662E-02 .773058 0.439 Israelis grapefruit/Brazilian orange .070863 .089224 .794213 0.427 Israelis grapefruit/ROW orange .675130E-03 .877277E-03 .769574 0.442 Israelis grapefruit/U.S grapefruit .361221 .307904E-02 117.316 0.000 Israelis grapefruit/ROW grapefruit .018581 .257048E-03 72.2846 0.000 Israelis grapefruit /U.S apple .081648 .054871 1.48800 0.137 Israelis grapefruit /Chinese apple .017089 .012568 1.35972 0.174
161
Table D-4 Continued Products Estimates SD t-statistic p-value Israelis grapefruit /ROW apple .089974 .063396 1.41925 0.156 Israelis grapefruit /Thai Pineapple -.088353 .051088 -1.72942 0.084 Israelis grapefruit /Philippines Pineapple -.080146 .045179 -1.77399 0.076 Israelis grapefruit /ROW Pineapple -.472915E-02 .012279 -.385153 0.700 Israelis grapefruit /U.S grape .020897 .025319 .825351 0.409 Israelis grapefruit /Argentinean grape .448466E-02 .558437E-02 .803074 0.422 Israelis grapefruit /ROW grape .029319 .034275 .855416 0.392 Israelis grapefruit /Israelis citrus -.015852 .016834 -.941693 0.346 Israelis grapefruit /Italian citrus -.015538 .016405 -.947153 0.344 Israelis grapefruit /ROW citrus -.967231E-02 .010293 -.939713 0.347 ROW grapefruit/U.S orange .320532E-02 .406844E-02 .787848 0.431 ROW grapefruit/Brazilian orange .104687 .129129 .810717 0.418 ROW grapefruit/ROW orange .997386E-03 .127075E-02 .784878 0.433 ROW grapefruit/U.S grapefruit .533640 .422317E-02 126.360 0.000 ROW grapefruit/Israelis grapefruit .043197 .597601E-03 72.2846 0.000 ROW grapefruit /U.S apple .120621 .073853 1.63326 0.102 ROW grapefruit /Chinese apple .025247 .017216 1.46651 0.143 ROW grapefruit /ROW apple .132921 .085289 1.55848 0.119 ROW grapefruit /Thai Pineapple -.130526 .066405 -1.96560 0.049 ROW grapefruit /Philippines Pineapple -.118402 .058498 -2.02403 0.043 ROW grapefruit /ROW Pineapple -.698649E-02 .018076 -.386516 0.699 ROW grapefruit /U.S grape .030872 .036724 .840655 0.401 ROW grapefruit /Argentinean grape .662529E-02 .812521E-02 .815400 0.415 ROW grapefruit /ROW grape .043314 .049741 .870781 0.384 ROW grapefruit /Israelis citrus -.023419 .024059 -.973391 0.330 ROW grapefruit /Italian citrus -.022955 .023399 -.981037 0.327 ROW grapefruit /ROW citrus -.014289 .014709 -.971482 0.331 U.S apple/U.S orange -.880028E-02 .013493 -.652216 0.514 U.S apple/Brazilian orange -.287421 .428141 -.671324 0.502 U.S apple/ROW orange -.273835E-02 .427393E-02 -.640710 0.522 U.S apple/U.S grapefruit .461381 .177814 2.59474 0.009 U.S apple/Israelis grapefruit .037348 .025100 1.48800 0.137 U.S apple/ROW grapefruit .023733 .014531 1.63326 0.102 U.S apple /Chinese apple -.096026 .400697E-02 -23.9647 0.000 U.S apple /ROW apple -.505566 .019164 -26.3808 0.000 U.S apple /Thai Pineapple .212565 .061409 3.46149 0.001 U.S apple /Philippines Pineapple .192821 .055655 3.46456 0.001 U.S apple /ROW Pineapple .011378 .028765 .395543 0.692 U.S apple /U.S grape -.022163 .093699 -.236538 0.813 U.S apple /Argentinean grape -.475639E-02 .020223 -.235199 0.814 U.S apple /ROW grape -.031096 .131501 -.236467 0.813 U.S apple /Israelis citrus .043051 .058527 .735584 0.462 U.S apple /Italian citrus .042198 .057112 .738868 0.460 U.S apple /ROW citrus .026268 .035172 .746834 0.455 Chinese apple/U.S orange -.143704E-02 .224523E-02 -.640041 0.522 Chinese apple/Brazilian orange -.046934 .071336 -.657936 0.511 Chinese apple/ROW orange -.447159E-03 .711912E-03 -.628109 0.530 Chinese apple/U.S grapefruit .075341 .036756 2.04976 0.040 Chinese apple/Israelis grapefruit .609875E-02 .448531E-02 1.35972 0.174 Chinese apple/ROW grapefruit .387543E-02 .264263E-02 1.46651 0.143 Chinese apple /U.S apple -.074917 .312614E-02 -23.9647 0.000 Chinese apple /ROW apple -.082556 .346618E-02 -23.8177 0.000 Chinese apple /Thai Pineapple .034711 .014817 2.34266 0.019 Chinese apple /Philippines Pineapple .031487 .013561 2.32177 0.020
162
Table D-4 Continued Products Estimates SD t-statistic p-value Chinese apple /ROW Pineapple .185792E-02 .473434E-02 .392435 0.695 Chinese apple /U.S grape -.361918E-02 .015433 -.234507 0.815 Chinese apple /Argentinean grape -.776694E-03 .333149E-02 -.233137 0.816 Chinese apple /ROW grape -.507774E-02 .021665 -.234376 0.815 Chinese apple /Israelis citrus .703006E-02 .991531E-02 .709010 0.478 Chinese apple /Italian citrus .689075E-02 .968113E-02 .711771 0.477 Chinese apple /ROW citrus .428941E-02 .597903E-02 .717409 0.473 ROW apple/U.S orange -.333268E-02 .518224E-02 -.643096 0.520 ROW apple/Brazilian orange -.108847 .163379 -.666225 0.505 ROW apple/ROW orange -.103702E-02 .163911E-02 -.632672 0.527 ROW apple/U.S grapefruit .174726 .073614 2.37352 0.018 ROW apple/Israelis grapefruit .014144 .996566E-02 1.41925 0.156 ROW apple/ROW grapefruit .898762E-02 .576692E-02 1.55848 0.119 ROW apple /U.S apple -.173742 .658595E-02 -26.3808 0.000 ROW apple /Chinese apple -.036365 .152681E-02 -23.8177 0.000 ROW apple /Thai Pineapple .080499 .025738 3.12769 0.002 ROW apple /Philippines Pineapple .073022 .023563 3.09904 0.002 ROW apple /ROW Pineapple .430875E-02 .010908 .395022 0.693 ROW apple /U.S grape -.839335E-02 .035585 -.235865 0.814 ROW apple /Argentinean grape -.180125E-02 .767938E-02 -.234557 0.815 ROW apple /ROW grape -.011776 .049872 -.236122 0.813 ROW apple /Israelis citrus .016304 .022502 .724545 0.469 ROW apple /Italian citrus .015981 .021877 .730470 0.465 ROW apple /ROW citrus .994769E-02 .013486 .737639 0.461 Thai pineapple/U.S orange .010657 .020171 .528347 0.597 Thai pineapple/Brazilian orange .348079 .648814 .536485 0.592 Thai pineapple /ROW orange .331625E-02 .631908E-02 .524800 0.600 Thai pineapple /U.S grapefruit -2.59181 .433861 -5.97383 0.000 Thai pineapple /Israelis grapefruit -.209803 .121314 -1.72942 0.084 Thai pineapple /ROW grapefruit -.133319 .067826 -1.96560 0.049 Thai pineapple /U.S apple 1.10348 .318786 3.46149 0.001 Thai pineapple /Chinese apple .230963 .098590 2.34266 0.019 Thai pineapple /ROW apple 1.21600 .388785 3.12769 0.002 Thai pineapple /Philippines pineapple .039981 .157894E-02 25.3216 0.000 Thai pineapple /ROW Pineapple .235915E-02 .954555E-03 2.47147 0.013 Thai pineapple /U.S grape .195801 .222746 .879035 0.379 Thai pineapple /Argentinean GR .042020 .049519 .848560 0.396 Thai pineapple /ROW grape .274710 .309032 .888938 0.374 Thai pineapple /Israelis citrus .351352 .148721 2.36250 0.018 Thai pineapple /Italian citrus .344390 .146229 2.35514 0.019 Thai pineapple /ROW citrus .214379 .091838 2.33431 0.020 Philippines pineapple/U.S orange .013943 .026251 .531158 0.595 Philippines pineapple/Brazilian orange .455392 .844363 .539332 0.590 Philippines pineapple /ROW orange .433865E-02 .822363E-02 .527584 0.598 Philippines pineapple /U.S grapefruit -3.39087 .558497 -6.07142 0.000 Philippines pineapple /Israelis grapefruit -.274485 .154728 -1.77399 0.076 Philippines pineapple /ROW grapefruit -.174421 .086175 -2.02403 0.043 Philippines pineapple /U.S apple 1.44368 .416699 3.46456 0.001 Philippines pineapple /Chinese apple .302169 .130146 2.32177 0.020 Philippines pineapple /ROW apple 1.59089 .513348 3.09904 0.002 Philippines pineapple /Thai pineapple .057664 .227725E-02 25.3216 0.000 Philippines pineapple /ROW pineapple .308648E-02 .123898E-02 2.49116 0.013 Philippines pineapple /U.S grape .256167 .291250 .879542 0.379 Philippines pineapple /Argentinean grape .054975 .064703 .849640 0.396
163
Table D-4 Continued Products Estimates SD t-statistic p-value Philippines pineapple /ROW grape .359404 .403773 .890114 0.373 Philippines pineapple /Israelis other citrus .459674 .193120 2.38025 0.017 Philippines pineapple /Italian citrus .450565 .190809 2.36134 0.018 Philippines pineapple /ROW other citrus .280471 .119762 2.34190 0.019 ROW pineapple/U.S orange .697506E-03 .217269E-02 .321033 0.748 ROW pineapple/Brazilian orange .022781 .070641 .322491 0.747 ROW pineapple /ROW orange .217040E-03 .677679E-03 .320271 0.749 ROW pineapple /U.S grapefruit -.169628 .426175 -.398023 0.691 ROW pineapple /Israelis grapefruit -.013731 .035651 -.385153 0.700 ROW pineapple /ROW grapefruit -.872538E-02 .022574 -.386516 0.699 ROW pineapple /U.S apple .072220 .182584 .395543 0.692 ROW pineapple /Chinese apple .015116 .038518 .392435 0.695 ROW pineapple /ROW apple .079584 .201467 .395022 0.693 ROW pineapple /Thai pineapple .288461E-02 .116716E-02 2.47147 0.013 ROW pineapple /Philippines pineapple .261667E-02 .105038E-02 2.49116 0.013 ROW pineapple /U.S grape .012815 .035370 .362302 0.717 ROW pineapple /Argentinean grape .275009E-02 .763540E-02 .360177 0.719 ROW pineapple /ROW grape .017979 .049381 .364088 0.716 ROW pineapple /Israelis citrus .022995 .058650 .392071 0.695 ROW pineapple /Italian citrus .022539 .057264 .393604 0.694 ROW pineapple /ROW citrus .014031 .035664 .393410 0.694 U.S grape/U.S orange -.732907E-02 .728333E-02 -1.00628 0.314 U.S grape/Brazilian orange -.239371 .221569 -1.08035 0.280 U.S grape/ROW orange -.228056E-02 .232533E-02 -.980749 0.327 U.S grape/U.S grapefruit .107900 .119546 .902588 0.367 U.S grape/Israelis grapefruit .873437E-02 .010583 .825351 0.409 U.S grape/ROW grapefruit .555023E-02 .660227E-02 .840655 0.401 U.S grape/U.S apple -.020252 .085617 -.236538 0.813 U.S grape/Chinese apple -.423877E-02 .018075 -.234507 0.815 U.S grape/ROW apple -.022317 .094616 -.235865 0.814 U.S grape/Thai Pineapple .034464 .039207 .879035 0.379 U.S grape /Philippines Pineapple .031263 .035544 .879542 0.379 U.S grape /ROW Pineapple .184471E-02 .509164E-02 .362302 0.717 U.S grape /Argentinean grape .040744 .423866E-03 96.1243 0.000 U.S grape /ROW grape .266368 .214706E-02 124.062 0.000 U.S grape /Israelis citrus .054435 .042863 1.26998 0.204 U.S grape /Italian citrus .053356 .042038 1.26925 0.204 U.S grape /ROW citrus .033214 .025146 1.32083 0.187 Argentinean grape/U.S orange -.010658 .010931 -.975055 0.330 Argentinean grape/Brazilian orange -.348097 .337904 -1.03017 0.303 Argentinean grape/ROW orange -.331643E-02 .348005E-02 -.952982 0.341 Argentinean grape/U.S grapefruit .156910 .180356 .870002 0.384 Argentinean grape/Israelis grapefruit .012702 .015816 .803074 0.422 Argentinean grape/ROW grapefruit .807123E-02 .989850E-02 .815400 0.415 Argentinean grape/U.S apple -.029450 .125214 -.235199 0.814 Argentinean grape/Chinese apple -.616409E-02 .026440 -.233137 0.816 Argentinean grape/ROW apple -.032453 .138360 -.234557 0.815 Argentinean grape/Thai pineapple .050118 .059063 .848560 0.396 Argentinean grape /Philippines pineapple. .045463 .053508 .849640 0.396 Argentinean grape /ROW Pineapple .268261E-02 .744803E-02 .360177 0.719 Argentinean grape /U.S grape .276090 .287222E-02 96.1243 0.000 Argentinean grape /ROW grape .387357 .342908E-02 112.962 0.000 Argentinean grape /Israelis other citrus .079160 .067539 1.17206 0.241 Argentinean grape /Italian citrus .077591 .066198 1.17212 0.241
164
Table D-4 Continued Products Estimates SD t-statistic p-value Argentinean grape /ROW other CT .048300 .039802 1.21350 0.225 ROW grape/U.S orange -.986263E-02 .950303E-02 -1.03784 0.299 ROW grape/Brazilian orange -.322119 .285388 -1.12870 0.259 ROW grape/ROW orange -.306892E-02 .304545E-02 -1.00771 0.314 ROW grape/U.S grapefruit .145200 .155199 .935576 0.349 ROW grape/Israelis grapefruit .011754 .013740 .855416 0.392 ROW grape/ROW grapefruit .746887E-02 .857721E-02 .870781 0.384 ROW grape/U.S apple -.027252 .115248 -.236467 0.813 ROW grape/Chinese apple -.570406E-02 .024337 -.234376 0.815 ROW grape/ROW apple -.030031 .127185 -.236122 0.813 ROW grape/Thai Pineapple .046378 .052172 .888938 0.374 ROW grape /Philippines Pineapple .042070 .047264 .890114 0.373 ROW grape /ROW Pineapple .248240E-02 .681814E-02 .364088 0.716 ROW grape /U.S grape .255485 .205934E-02 124.062 0.000 ROW grape /Argentinean grape .054828 .485369E-03 112.962 0.000 ROW grape /Israelis citrus .073252 .056263 1.30197 0.193 ROW grape /Italian citrus .071801 .054847 1.30911 0.190 ROW grape /ROW citrus .044695 .033202 1.34617 0.178 Israelis citrus/U.S orange .020293 .013444 1.50951 0.131 Israelis citrus /Brazilian orange .662784 .390838 1.69580 0.090 Israelis citrus /ROW orange .631454E-02 .434106E-02 1.45461 0.146 Israelis citrus /U.S grapefruit -.230334 .202540 -1.13723 0.255 Israelis citrus /Israelis grapefruit -.018645 .019800 -.941693 0.346 Israelis citrus /ROW grapefruit -.011848 .012172 -.973391 0.330 Israelis citrus /U.S apple .110699 .150491 .735584 0.462 Israelis citrus /Chinese apple .023170 .032679 .709010 0.478 Israelis citrus /ROW apple .121986 .168363 .724545 0.469 Israelis citrus /Thai pineapple .174032 .073664 2.36250 0.018 Israelis citrus /Philippines pineapple .157866 .066323 2.38025 0.017 Israelis citrus /ROW Pineapple .931515E-02 .023759 .392071 0.695 Israelis citrus /U.S grape .153183 .120618 1.26998 0.204 Israelis citrus /Argentinean grape .032874 .028048 1.17206 0.241 Israelis citrus /ROW grape .214917 .165071 1.30197 0.193 Israelis citrus /Italian citrus .118389 .002331 50.7888 0.000 Israelis citrus /ROW citrus .073696 .000869 84.7633 0.000 Italian citrus/U.S orange .025425 .017699 1.43648 0.151 Italian citrus /Brazilian orange .830381 .514580 1.61371 0.107 Italian citrus /ROW orange .007911 .005687 1.39101 0.164 Italian citrus /U.S grapefruit -.288579 .251672 -1.14664 0.252 Italian citrus /Israelis grapefruit -.023360 .024663 -.947153 0.344 Italian citrus /ROW grapefruit -.014844 .015131 -.981037 0.327 Italian citrus /U.S apple .138691 .187707 .738868 0.460 Italian citrus /Chinese apple .029029 .040784 .711771 0.477 Italian citrus /ROW apple .152833 .209226 .730470 0.465 Italian citrus /Thai pineapple .218039 .092580 2.35514 0.019 Italian citrus /Philippines pineapple .197786 .083760 2.36134 0.018 Italian citrus /ROW Pineapple .011671 .029651 .393604 0.694 Italian citrus /U.S grape .191918 .151206 1.26925 0.204 Italian citrus /Argentinean grape .041187 .035139 1.17212 0.241 Italian citrus /ROW grape .269263 .205683 1.30911 0.190 Italian citrus /Israelis citrus .151324 .002979 50.7888 0.000 Italian citrus /ROW citrus .092331 .001202 76.7658 0.000 ROW citrus/U.S orange .010916 .007544 1.44695 0.148 ROW citrus /Brazilian orange .356530 .225643 1.58006 0.114
165
Table D-4 Continued Products Estimates SD t-statistic p-value ROW citrus /ROW orange .003396 .002433 1.39563 0.163 ROW citrus /U.S grapefruit -.123903 .109877 -1.12765 0.259 ROW citrus /Israelis grapefruit -.010030 .010673 -.939713 0.347 ROW citrus /ROW grapefruit -.006373 .006560 -.971482 0.331 ROW citrus /U.S apple .059548 .079734 .746834 0.455 ROW citrus /Chinese apple .012464 .017373 .717409 0.473 ROW citrus /ROW apple .065620 .088959 .737639 0.461 ROW citrus /Thai pineapple .093616 .040104 2.33431 0.020 ROW citrus /Philippines pineapple .084921 .036261 2.34190 0.019 ROW citrus /ROW Pineapple .005010 .012737 .393410 0.694 ROW citrus /U.S grape .082401 .062386 1.32083 0.187 ROW citrus /Argentinean GR .017684 .014573 1.21350 0.225 ROW citrus /ROW grape .115610 .085880 1.34617 0.178 ROW citrus /Israelis citrus .064972 .000766 84.7633 0.000 ROW citrus /Italian citrus .063685 .000829 76.7658 0.000
166
APPENDIX E TWO-STAGE ROTTERDAM MODEL
Theoretical Models
Parameter estimates of a system of consumer demand can be obtained in two
ways. One is direct estimation by including all the products of interest in one system of
equations. The other is estimating a multi-stage demand system consistent with a multi-
stage budget allocation. Both ways provide the same estimates. However, under
situations where there are a limited number of observations, it is advantageous to apply a
multi-stage system. In this study, a two-stage Rotterdam demand system is applied
consistent with a two stage budget allocation. In the first stage, total expenditure is
allocated over broader groups of goods. In the second stage, group expenditures are
allocated over individual goods. In this framework, the first stage involves the allocation
of total expenditure over groups of imported fruit juice including orange, grapefruit, other
citrus, apple, pineapple and grape juices, and the second stage involves allocation of
group expenditure over supplying countries of a certain product group.
The optimal allocation of expenditure on imported fruit juices will be carried out
by specifying a first stage (group) demand equation for the fruit juice groups and a
second stage (conditional) demand equation for individual country-specific products.
Since the system of the first and second stage demand systems can be estimated
simultaneously, it is possible to show how a change in trade policy applied to a product
group translates into an impact on the export of individual countries, and conversely how
167
a change in trade policy applied to a product from a certain country is translated into an
impact on the total demand for imported fruit juices.
The study estimates four two-stage demand systems of fruit juices using the
relative price version of the Rotterdam model. These are
(1) Two-stage block independent Rotterdam model (2) Two-stage block-wise dependent Rotterdam model (3) Two-stage block independent uniform substitute-Rotterdam model (4) Two-stage block-wise dependent uniform substitute-Rotterdam model
Two-Stage Block Independent Rotterdam Model
First-stage (Group) demand. Following Theil (1980a) the first stage or group
demand equation of the block independent Rotterdam model can be given by
(1) ( ) ( ) ⎟⎟⎠
⎞⎜⎜⎝
⎛Θ+Θ= '
'
logloglogPP
dQdQdW ggggg φ ,
where ∑∈
=gSi
ig wW is the average expenditure share of group g ; ∑∈
=ΘgSi
ig θ is the
marginal expenditure share of group g; ( ) ( )iSi g
ig qd
Ww
Qdg
loglog ∑∈
= ; is the group Divisia
volume index; φ is the income flexibility of the marginal utility of income,
( ) ( )iSi g
ig pdPd
g
loglog ' ∑∈ Θ
=θ
is the Frisch price index of group g ;
( ) ( )∑=
=G
ggg QdWQd
1loglog is the overall Divisia volume index of all groups in the
system; ( ) ( ) ( )j
N
jj
G
ggg pdPdPd loglog'log
11
' ∑∑==
=Θ= θ is the overall Frisch price index.
Second-stage demand. Continuing to follow Theil (1980a), the second stage
demand equation can be given by.
168
(2) ( ) ( ) ∑ ∈ ⎟⎟⎠
⎞⎜⎜⎝
⎛+=
gSjg
jijggiii P
pdvQdWqdw '
' logloglog θ ,
where gii Θ=θθ ' is the marginal share of good i within its group; ijν is the relative
price coefficients. The share 'iθ answers the question that if the consumer’s income
increases by one dollar, resulting in an additional amount of gΘ dollars spent on gS , what
is the proportion of this additional amount that is allocated to good i ?
The first stage function is related to the second stage function through the quantity
demanded of the product group term ( )gg QdW log which is found in both stage functions.
It is assumed that this term found in the second stage demand equation is exogenously
determined.
Two-stage Block-wise Dependent Rotterdam Model
Like the case with block independence, we can apply the two-stage utility
maximization procedure in a block-wise dependence framework and estimate as a system
of first and second-stage demand equations.
First-stage demand. Following Theil (1980a), the first stage demand equation of
the block-wise dependent Rotterdam model can be given as
(3) ( ) ( ) ∑=
⎟⎟⎠
⎞⎜⎜⎝
⎛+Θ=
G
h
hghgggg P
PdVQdQdW
1'
'
logloglog
where ghV is the group relative price coefficients defined as. ∑∑∈ ∈
=gi hj
ijghV ν .
Second-stage demand. The second stage demand equation of the block-wise
dependent Rotterdam model is the same as that of the block independent Rotterdam
model. So, we can reproduce equation (2) here.
169
(4) ( ) ( ) ∑ ∈ ⎟⎟⎠
⎞⎜⎜⎝
⎛+=
gSjg
jijggiii P
pdvQdWqdw '
' logloglog θ .
Two-stage Block Independent Uniform Substitute-Rotterdam Model
First stage demand. Since the uniform substitute hypothesis is imposed on a set
of goods within the same group, the first stage block independent uniform substitute-
Rotterdam model is the same as that of the first-stage equation of the block independent
Rotterdam demand model. Thus, we reproduce it here as it is.
(5) ( ) ( ) ⎟⎟⎠
⎞⎜⎜⎝
⎛Θ+Θ= '
'
logloglogPP
dQdQdW ggggg φ .
Second stage demand. Following Theil (1980a), the second-stage equation of
the block independent uniform substitute- Rotterdam model can be given as
(6) ( ) ( ) ⎟⎟⎠
⎞⎜⎜⎝
⎛
Θ−
Θ+= '
'' log
1loglog
g
i
g
igggiii P
pd
kQdWqdw
θφθ .
Two-stage Block-wise dependent Uniform Substitute-Rotterdam Model
Since the uniform substitute hypothesis is imposed on a set of goods within the
same group, the first stage equation of the block-wise dependent uniform substitute-
Rotterdam model is the same as that of the block-wise dependent Rotterdam model. Thus,
we reproduce equation (3) here as it is.
(7) ( ) ( ) ∑=
⎟⎟⎠
⎞⎜⎜⎝
⎛+Θ=
G
h
hghgggg P
PdVQdQdW
1'
'
logloglog .
Following Seale (2003) the second-stage equation of the block-wise dependent
uniform substitute-Rotterdam model can be given as
170
(8) ( ) ( ) ⎟⎟⎠
⎞⎜⎜⎝
⎛
Θ−
Θ+= '
'' log
1loglog
g
i
g
iggggiii P
pd
kQdWqdw
θφθ .
If two-stage utility maximization is appropriate, then the two-stage demand
system and the unconditional demand systems should yield the same results. Edgerton
(1997) derived the formulae which convert the calculated two-stage demand parameters
to corresponding unconditional demand elasticities. Hence, the total expenditure
elasticity of the thi good within commodity group A is given by
(9) ( ) ( )AiAi EEE =
where ( )iAE is the second-stage elasticity of the thi good in group A and ( )AE is the first-
stage (group) expenditure elasticity for the thA group.
For two goods, i and j , belonging to commodity groups A and B , Edgerton
(1997) gives the total price elasticities for goods i and j as
(10) ( ) ( ) ( ) ( )( )( )BAABjBiAijAABij ewEee ++= δδ
where δ is the Kronecker’s delta equal to one when BA = and zero otherwise; ( )ijAe is
the second-stage cross price elasticity between goods i and j within group A and
( ) jBw the budget share of the thj good within group B .
Empirical Models of a Two-stage Demand System
Since the differential approach to consumption theory discussed in the previous
section does not postulate constancy for the coefficients of its demand equations, we
can’t talk about empirical estimation. In this section, we postulate that the coefficients
are constant and discuss the ways in which the theoretical models in the previous section
are parameterized so that they can be applied to statistical data. Since the nature of data
171
forces us to work with finite rather than infinitesimal changes, we replace the
infinitesimal changes of the theoretical models presented in the previous section by finite
changes. Finally, estimation procedures are presented for the different versions of the
Rotterdam model.
Two-Stage Block Independent Rotterdam Model
The system of first-stage (group) and second-stage (second-stage) demand
equations for the block independent Rotterdam demand model can be given by
(11)
⎪⎪
⎩
⎪⎪
⎨
⎧
+⎟⎟⎠
⎞⎜⎜⎝
⎛+=
+⎟⎟⎠
⎞⎜⎜⎝
⎛Θ+Θ=
∑∈
ig
j
Sjijggiii
gg
gggg
dPdp
vdQWdqw
dPdP
dQdQW
g
εθ
εφ
''
'
'.
where ( )
212, −+
= tggtgt
WWW is the average budget share of group g ;
6611 ... dQWdQWdQ ++= ; jSj
jg dpdPg
∑∈
= '' θ ; '6
1' g
gg dPdP ∑
=
Θ= ; gε is a group demand
disturbance equal to the sum of iε over gSi∈ .
Since the si'ε have multinomial distribution with zero means, so have the group
demand disturbances. If the error terms of the second-stage demand equation
disturbances are correlated with the group demand equation disturbances were correlated,
the consumer can no longer separate their allocation problem into two stages (Theil,
1980b). Also, a non-zero correlation of the Divisia volume index gdQ and the
disturbance of the second-stage equation iε imply that the volume index is not a
predetermined variable in the second-stage demand systems.
172
The first-stage or group demand function can be estimated after deleting one of
the six group demand equations as
(12)
( )( )( )( )( )
''''
'
55555
44444
33333
22222
11111
⎪⎪⎪
⎩
⎪⎪⎪
⎨
⎧
−Θ+Θ=
−Θ+Θ=
−Θ+Θ=
−Θ+Θ=
−Θ+Θ=
dPdPdQdQWdPdPdQdQWdPdPdQdQWdPdPdQdQW
dPdPdQdQW
φφ
φφφ
The second stage requires some manipulation to put it into an estimable form. In
order to estimate the second equation of equation (12) one of the three second-stage
demand equations is deleted from each of the second-stage demand systems. Using the
constraint iSj
ijg
v φθ=∑∈
, we write the own price coefficient iiv in terms of the other price
coefficients as ∑≠
−=ij
ijiii vv φθ so that the price term of equation (12) becomes
(13) ( ) ( )∑∑∈≠∈≠
−+−⎟⎟⎠
⎞⎜⎜⎝
⎛−=
gg Sijgjijgi
Sijiji dPdpvdPdpv '''φθ
( ) ( )∑∈≠
−+−=gSij
ijijgii dpdpvdPdp 'φθ
Now, using the constraint ∑≠
−=nk
kn θθ 1' we obtain
(14) ( ) ( )∑∑∈≠=
−+⎟⎠
⎞⎜⎝
⎛−−−=
gSijijij
n
knkknii dpdpvdpdpdpdp
1
'' θφθ .
Combining equation (13) and equation (14) yields
(15) ( ) ( ) iij
ijijiiiiii dpdpvBdqdqw εθφθθ +−++= ∑≠
''
173
where ( ) ( )⎟⎠
⎞⎜⎝
⎛−−−= ∑
=
n
knkiniii dpdpdpdpB
1
''' θθθ .
The estimable system of equations for the second-stage block independent
Rotterdam model can thus be given in equation (16) as
( ) ( ) ( )( ) ( ) ( )⎪⎩
⎪⎨⎧
−+−++=
−+−++=
2321
1312
2312211'222
1312111'
111
dpdpvdpdpvBdQWdqw
dpdpvdpdpvBdQWdqw
θφθ
θφθ
( ) ( ) ( )( ) ( ) ( )⎪⎩
⎪⎨⎧
−+−++=
−+−++=
5654
4645
56455221555
4645422'444
dpdpvdpdpvBdQWdqw
dpdpvdpdpvBdQWdqw
θφθ
θφθ
( ) ( ) ( )( ) ( ) ( )⎪⎩
⎪⎨⎧
−+−++=
−+−++=
8987
7978
89788331888
7978733'777
dpdpvdpdpvBdQWdqw
dpdpvdpdpvBdQWdqw
θφθ
θφθ
( ) ( ) ( )( ) ( ) ( )⎪⎩
⎪⎨⎧
−+−++=
−+−++=
11121110
10121011
1112101111441111111
101210111044'
101010
dpdpvdpdpvBdQWdqw
dpdpvdpdpvBdQWdqw
θφθ
θφθ
( ) ( ) ( )( ) ( ) ( )⎪⎩
⎪⎨⎧
−+−++=
−+−++=
14151413
13151314
1415131414551141414
131513141355'
131313
dpdpvdpdpvBdQWdqw
dpdpvdpdpvBdQWdqw
θφθ
θφθ
( ) ( ) ( )( ) ( ) ( )⎪⎩
⎪⎨⎧
−+−++=
−+−++=
17181716
16181617
171816171766'
171717
161816171666'
161616
dpdpvdpdpvBdQWdqw
dpdpvdpdpvBdQWdqw
θφθ
θφθ
Estimating equation (12) and equation (16) in one system will yield the parameter
estimates of the group demand function and second-stage demand functions. The error
terms of the second-stage and group demand functions are assumed to be uncorrelated,
and that the group Divisia volume index is predetermined.
174
Two-Stage Block-wise Dependent Rotterdam Model
The system of group and second-stage demand equations of the block-wise
dependent Rotterdam demand model can be given by
(17)
⎪⎪
⎩
⎪⎪
⎨
⎧
+⎟⎟⎠
⎞⎜⎜⎝
⎛+=
+⎟⎟⎠
⎞⎜⎜⎝
⎛+Θ=
∑
∑
∈
=
ig
j
Sjijggiii
gh
hghgggg
edPdp
vdQWdqw
dPdP
VdQdQW
g
''
6
1
'
'
θ
ξ
.
The estimation procedure of the block wise dependent Rotterdam model is similar
to that of the second-stage block independent Rotterdam model presented earlier. In
order to estimate the group demand equation (17) one of the six demand equations would
be deleted. Using the constraint ggh
ghV Θ=∑=
φ1
, we write the own price coefficient ggV in
terms of the other price coefficients as ∑≠
−Θ=6
ghghgggg VV φ and substitute it in equation
(17) so that the price term of equation (14) becomes
(18) ( ) ( )dPdPVdPdPV hgh
ghggh
ghgg −+−⎟⎟⎠
⎞⎜⎜⎝
⎛−Θ= ∑∑
≠≠
'6
'6
'φ
( ) ( )''6
ghgh
ghg dPdPVA −+Θ= ∑≠
φ
Now combining the first-stage of equation (17) and equation (18), we obtain
(19) ( ) ( ) ggh
ghijgggii dPdPVAdQdqw εφ +−+Θ+Θ= ∑≠
''
where ( ) ( )⎟⎠
⎞⎜⎝
⎛−Θ−−Θ=Θ ∑
=
6
1
'
knkkmhggg dPdPdPdPA .
175
The estimable systems of the first–stage (equation 20) for the block-wise
dependent Rotterdam demand model are given by
(20)
( ) ( ) ( ) ( )( )
( ) ( ) ( ) ( )( )
( ) ( ) ( ) ( )( )
( ) ( ) ( ) ( )( )
( ) ( ) ( ) ( )( )⎪
⎪⎪⎪⎪⎪⎪
⎩
⎪⎪⎪⎪⎪⎪⎪
⎨
⎧
−+−+−+−+−+Θ+Θ=
−+−+−+−+−+Θ+Θ=
−+−+−+−+−+Θ+Θ=
−+−+−+−+−+Θ+Θ=
−+−+−+−+−+Θ+Θ=
56565445
5335522551155555
46464545
4334422441144444
36363535
3434322331133333
26262525
2424232321122222
16161515
1414131312121111
()
()
()
()
()
dPdPVdPdPVdPdPVdPdPVdPdPVAdQdQW
dPdPVdPdPVdPdPVdPdPVdPdPVAdQdQW
dPdPVdPdPVdPdPVdPdPVdPdPVAdQdQW
dPdPVdPdPVdPdPVdPdPVdPdPVAdQdQW
dPdPVdPdPVdPdPVdPdPVdPdPVAdQdQW
φ
φ
φ
φ
φ
The estimable systems of the second-stage demand equations (equation 21) for
the relative price version of the Rotterdam model under block-wise dependence are given
by
(21)
( ) ( ) ( )( ) ( ) ( )⎪⎩
⎪⎨⎧
−+−++=
−+−++=
2321
1312
2312211'222
1312111'
111
dpdpvdpdpvBdQWdqw
dpdpvdpdpvBdQWdqw
θφθ
θφθ
( ) ( ) ( )( ) ( ) ( )⎪⎩
⎪⎨⎧
−+−++=
−+−++=
5654
4645
56455221555
4645422'444
dpdpvdpdpvBdQWdqw
dpdpvdpdpvBdQWdqw
θφθ
θφθ
( ) ( ) ( )( ) ( ) ( )⎪⎩
⎪⎨⎧
−+−++=
−+−++=
8987
7978
89788331888
7978733'777
dpdpvdpdpvBdQWdqw
dpdpvdpdpvBdQWdqw
θφθ
θφθ
( ) ( ) ( )( ) ( ) ( )⎪⎩
⎪⎨⎧
−+−++=
−+−++=
11121110
10121011
1112101111441111111
101210111044'
101010
dpdpvdpdpvBdQWdqw
dpdpvdpdpvBdQWdqw
θφθ
θφθ
( ) ( ) ( )( ) ( ) ( )⎪⎩
⎪⎨⎧
−+−++=
−+−++=
14151413
13151314
1415131414551141414
131513141355'
131313
dpdpvdpdpvBdQWdqw
dpdpvdpdpvBdQWdqw
θφθ
θφθ
176
( ) ( ) ( )( ) ( ) ( )⎪⎩
⎪⎨⎧
−+−++=
−+−++=
17181716
16181617
171816171766'
171717
161816171666'
161616
dpdpvdpdpvBdQWdqw
dpdpvdpdpvBdQWdqw
θφθ
θφθ
Two-Stage Block independent Uniform Substitute-Rotterdam Model
The system of group and second-stage demand equations for uniform-substitute
block independent Rotterdam demand model can be given by
(22)
⎪⎪
⎩
⎪⎪
⎨
⎧
+⎟⎟⎠
⎞⎜⎜⎝
⎛
Θ−
Θ+=
+⎟⎟⎠
⎞⎜⎜⎝
⎛Θ+Θ=
ig
i
g
igggiii
gg
gggg
edPdp
kdQWdqw
dPdP
dQdQW
'
''
'
1
'
θφθ
εφ
.
The estimable systems of the first–stage (equation 23) and second-stage demand
equations (equation 24) for the block independent Rotterdam demand model can be given
by
( )( )( )( )( )
''''
'
)23(
55555
44444
33333
22222
11111
⎪⎪⎪
⎩
⎪⎪⎪
⎨
⎧
−Θ+Θ=
−Θ+Θ=
−Θ+Θ=
−Θ+Θ=
−Θ+Θ=
dPdPdQdQWdPdPdQdQWdPdPdQdQWdPdPdQdQW
dPdPdQdQW
φφ
φφφ
(24) ( ) ( )( )
( ) ( )( )⎪⎪⎩
⎪⎪⎨
⎧
−−−−−Θ−
Θ+=
−−−−−Θ−
Θ+=
3231321
3231311
'2
'1
'2
11
111
'222
'2
'1
'1
11
111
'111
dpdpdpdpdpdpk
dQWdqw
dpdpdpdpdpdpk
dQWdqw
θθθφθ
θθθφθ
( ) ( )( )
( ) ( )( )⎪⎪⎩
⎪⎪⎨
⎧
−−−−−Θ−
Θ+=
−−−−−Θ−
Θ+=
6564651
6564641
'5
'4
'5
22
222
'555
'5
'4
'4
22
222
'444
dpdpdpdpdpdpk
dQWdqw
dpdpdpdpdpdpk
dQWdqw
θθθφθ
θθθφθ
177
( ) ( )( )
( ) ( )( )⎪⎪⎩
⎪⎪⎨
⎧
−−−−−Θ−
Θ+=
−−−−−Θ−
Θ+=
9897981
9897971
'8
'7
'8
33
333
'888
'8
'7
'7
33
333
'777
dpdpdpdpdpdpk
dQWdqw
dpdpdpdpdpdpk
dQWdqw
θθθφθ
θθθφθ
( ) ( )( )
( ) ( )( )⎪⎪⎩
⎪⎪⎨
⎧
−−−−−Θ−
Θ+=
−−−−−Θ−
Θ+=
1211121012111
1211121012101
'11
'10
'11
44
444
'111111
'11
'10
'10
44
444
'101010
dpdpdpdpdpdpk
dQWdqw
dpdpdpdpdpdpk
dQWdqw
θθθφθ
θθθφθ
( ) ( )( )
( ) ( )( )⎪⎪⎩
⎪⎪⎨
⎧
−−−−−Θ−
Θ+=
−−−−−Θ−
Θ+=
1514151315141
1514151315131
'14
'13
'14
55
555
'141414
'14
'13
'13
55
555
'131313
dpdpdpdpdpdpk
dQWdqw
dpdpdpdpdpdpk
dQWdqw
θθθφθ
θθθφθ
( ) ( )( )
( ) ( )( )⎪⎪⎩
⎪⎪⎨
⎧
−−−−−Θ−
Θ+=
−−−−−Θ−
Θ+=
1817181618171
1817181618161
'17
'16
'17
66
666
'171717
'17
'16
'16
66
666
'161616
dpdpdpdpdpdpk
dQWdqw
dpdpdpdpdpdpk
dQWdqw
θθθφθ
θθθφθ
Two-Stage Block-wise dependent uniform substitute-Rotterdam Model
The system of group and second-stage demand equations of the block-wise
dependent uniform substitute- Rotterdam demand model can be given by
(25)
⎪⎪
⎩
⎪⎪
⎨
⎧
+⎟⎟⎠
⎞⎜⎜⎝
⎛
Θ−
Θ+=
++Θ= ∑=
ig
i
g
iggggiii
gh
hghgggg
edPdp
kdQWdqw
dPdP
VdQdQW
1
''
6
1
'
1
'
θφθ
ξ
.
The estimable systems of the first–stage (equation 26) and second-stage demand
equations (equation 27) of the block-wise dependent uniform substitute- Rotterdam
model can be given by
(26)
178
( ) ( ) ( ) ( )( ) ⎥
⎦
⎤⎢⎣
⎡−+−
+−+−+−+−Θ+Θ=
16161515
141413131212111111 ()
'dPdPVdPdPV
dPdPVdPdPVdPdPVdPdPdQdQW
φ
( ) ( ) ( ) ( )( ) ⎥
⎦
⎤⎢⎣
⎡−+−
+−+−+−+−Θ+Θ=
26262525
242423232112222222 ()
'dPdPVdPdPV
dPdPVdPdPVdPdPVdPdPdQdQW
φ
( ) ( ) ( ) ( )( ) ⎥
⎦
⎤⎢⎣
⎡−+−
+−+−+−+−Θ+Θ=
36363535
343432233113333333 ()
'dPdPVdPdPV
dPdPVdPdPVdPdPVdPdPdQdQW
φ
( ) ( ) ( ) ( )( ) ⎥
⎦
⎤⎢⎣
⎡−+−
+−+−+−+−Θ+Θ=
46464545
433442244114444444 ()
'dPdPVdPdPV
dPdPVdPdPVdPdPVdPdPdQdQW
φ
( ) ( ) ( ) ( )( ) ⎥
⎦
⎤⎢⎣
⎡−+−
+−+−+−+−Θ+Θ=
56565445
533552255115555555 ()
'dPdPVdPdPV
dPdPVdPdPVdPdPVdPdPdQdQW
φ
( ) ( )( )
( ) ( )( )⎪⎪⎩
⎪⎪⎨
⎧
−−−−−Θ−
Θ+=
−−−−−Θ−
Θ+=
3231321
3231311
'2
'1
'2
111
1111
'222
'2
'1
'1
111
1111
'111
dpdpdpdpdpdpk
dQWdqw
dpdpdpdpdpdpk
dQWdqw
θθθφθ
θθθφθ
( ) ( )( )
( ) ( )( )⎪⎪⎩
⎪⎪⎨
⎧
−−−−−Θ−
Θ+=
−−−−−Θ−
Θ+=
6564651
6564641
'5
'4
'5
222
2222
'555
'5
'4
'4
222
2222
'444
dpdpdpdpdpdpk
dQWdqw
dpdpdpdpdpdpk
dQWdqw
θθθφθ
θθθφθ
( ) ( )( )
( ) ( )( )⎪⎪⎩
⎪⎪⎨
⎧
−−−−−Θ−
Θ+=
−−−−−Θ−
Θ+=
9897981
9897971
'8
'7
'8
333
3333
'888
'8
'7
'7
333
3333
'777
dpdpdpdpdpdpk
dQWdqw
dpdpdpdpdpdpk
dQWdqw
θθθφθ
θθθφθ
( ) ( )( )
( ) ( )( )⎪⎪⎩
⎪⎪⎨
⎧
−−−−−Θ−
Θ+=
−−−−−Θ−
Θ+=
1211121012111
1211121012101
'11
'10
'11
444
4444
'111111
'11
'10
'10
444
4444
'101010
dpdpdpdpdpdpk
dQWdqw
dpdpdpdpdpdpk
dQWdqw
θθθφθ
θθθφθ
( ) ( )( )
( ) ( )( )⎪⎪⎩
⎪⎪⎨
⎧
−−−−−Θ−
Θ+=
−−−−−Θ−
Θ+=
1514151315141
1514151315131
'14
'13
'14
555
555
'141414
'14
'13
'13
555
5555
'131313
dpdpdpdpdpdpk
dQWdqw
dpdpdpdpdpdpk
dQWdqw
θθθφθ
θθθφθ
179
( ) ( )( )
( ) ( )( )⎪⎪⎩
⎪⎪⎨
⎧
−−−−−Θ−
Θ+=
−−−−−Θ−
Θ+=
1817181618171
1817181618161
'17
'16
'17
66
666
'171717
'17
'16
'16
666
6666
'161616
dpdpdpdpdpdpk
dQWdqw
dpdpdpdpdpdpk
dQWdqw
θθθφθ
θθθφθ
180
APPENDIX F PARAMETER ESTIAMTES OF FRUIT JUCIES IN A TWO-STAGE ROTTERDAM
MODEL
Table F-1 Marginal value shares of fruit juices in a two-stage block independent Rotterdam model
Products Estimates SE t-statistics P-value First-stage ( )gΘ
Orange .784658 .028024 27.9994 0.000 Grapefruit .032615 .006289 5.18583 0.000 Apple .124181 .019870 6.24952 0.000 Pineapple .013296 .002634 5.04708 0.000 Grape .020134 .007185 2.80217 0.005 Other citrus .025117 .004603 5.45600 0.000
Second-stage ( )'iθ
U.S. orange .035610 .011509 3.09418 0.002 Brazilian orange .950692 .013339 71.2725 0.000 ROW orange .013697 .006854 1.99844 0.046 U.S. grapefruit .911801 .039109 23.3142 0.000 Israelis grapefruit .072429 .034661 2.08966 0.037 ROW grapefruit .015770 .019992 .788809 0.430 U.S. apple .324157 .029096 11.1411 0.000 Chinese apple .297490 .023956 12.4181 0.000 ROW apple .378353 .033876 11.1686 0.000 Thai pineapple .345634 .025559 13.5231 0.000 Philippine p. apple .591278 .029544 20.0137 0.000 ROW pineapple .063088 .036658 1.72099 0.085 U.S. grape .436304 .036702 11.8879 0.000 Argentinean grape -.004066 .019330 -.210375 0.833 ROW grape .567762 .034258 16.5729 0.000 Israelis other citrus .284424 .029932 9.50221 0.000 Italian other citrus .356137 .025032 14.2275 0.000 ROW other citrus .359439 .030920 11.6249 0.000 φ -1.81427 .274134 -6.61821 0.000
181
Table F-2 Relative price coefficients of fruit juices in a two-stage block independent Rotterdam model
Products Estimates SE t-statistics P-value First-stage ( )ggV
Orange -1.42358 .247270 -5.75721 0.000 Grapefruit -.059172 .011022 -5.36847 0.000 Apple -.225298 .039028 -5.77275 0.000 Pineapple -.024123 .004467 -5.39940 0.000 Grape -.036528 .012941 -2.82264 0.005 Other citrus -.045569 .006885 -6.61821 0.000
Second-stage ( )ijν
U.S. orange -.114667 .021029 -5.45290 0.000 U.S. orange/Brazilian orange .044702 .030979 1.44298 0.149 U.S. orange/ROW orange .005357 .006837 .783593 0.433 Brazilian orange -1.77649 .269315 -6.59635 0.000 Brazilian orange/ROW orange .006975 .015121 .461312 0.645 ROW orange -.037184 .004877 -7.62350 0.000 U.S. grapefruit -1.52703 .268499 -5.68727 0.000 U.S. grapefruit/Israelis grapefruit -.107800 .055164 -1.95419 0.051 U.S. grapefruit/ROW grapefruit -.019429 .032544 -.596994 0.551 Israelis grapefruit -.024248 .009785 -2.47800 0.013 Israelis grapefruit/ROW grapefruit .000642 .003121 .205710 0.837 ROW grapefruit -.163230 .033177 -4.92004 0.000 U.S. apple -.190613 .043210 -4.41137 0.000 U.S. apple/Chinese apple -.163230 .033177 -4.92004 0.000 U.S. apple/ROW apple -.234266 .037809 -6.19601 0.000 Chinese apple -.181944 .037199 -4.89110 0.000 Chinese apple/ROW apple -.194555 .037917 -5.13107 0.000 ROW apple -.257614 .068542 -3.75849 0.000 Thai. pineapple -.223952 .046327 -4.83419 0.000 Thai. pineapple/Philippines pineapple -.363552 .062524 -5.81460 0.000 Thai. pineapple/ROW pineapple -.039571 .022792 -1.73620 0.083 Philippines pineapple -.644166 .109387 -5.88887 0.000 Philippines pineapple/ROW apple -.065022 .038729 -1.67891 0.093 ROW pineapple -.009866 .008921 -1.10597 0.269 U.S. grape -.376426 .076567 -4.91628 0.000 U.S. grape/Argentinean grape .010298 .015761 .653395 0.514 U.S. grape/ROW grape -.425447 .071035 -5.98929 0.000 Argentinean grape -.008522 .003875 -2.19981 0.028 Argentinean grape/ROW grape .005604 .020168 .277910 0.781 ROW grape -.610234 .115943 -5.26324 0.000 Israelis citrus -.163454 .038524 -4.24295 0.000 Israelis citrus/Italian citrus -.175235 .034100 -5.13892 0.000 Israelis citrus/ROW citrus -.177333 .031369 -5.65313 0.000 Italian citrus -.244121 .048530 -5.03027 0.000 Italian citrus/ROW citrus -.226773 .039828 -5.69380 0.000 ROW citrus -.248015 .052568 -4.71796 0.000
182
Table F-3 Marginal value shares of fruit juices in a two-stage block independent uniform-substitute-Rotterdam model
Product estimates SE t-statistics p-value First-stage ( )gΘ
Orange .779670 .028880 26.9966 0.000 Grapefruit .033851 .006573 5.14986 0.000 Apple .125694 .020189 6.22585 0.000 Pineapple .013412 .002554 5.24986 0.000 Grape .021455 .007565 2.83594 0.005 Other citrus .025919 .004640 5.58519 0.000
Second-stage ( )'iθ
U.S. orange .034893 .010019 3.48271 0.000 Brazilian orange .953467 .013316 71.6023 0.000 ROW orange .011640 .003894 2.98891 0.003 U.S. grapefruit .851736 .036264 23.4872 0.000 Israelis grapefruit .092200 .025224 3.65530 0.000 ROW grapefruit .056065 .015162 3.69775 0.000 U.S. apple .283858 .027503 10.3212 0.000 Chinese apple .305403 .022680 13.4660 0.000 ROW apple .410739 .031858 12.8929 0.000 Thai pineapple .312187 .024819 12.5785 0.000 Philippine p. apple .610394 .026926 22.6695 0.000 ROW pineapple .077419 .031026 2.49529 0.013 U.S. grape .432947 .035615 12.1564 0.000 Argentinean grape .071831 .018053 3.97889 0.000 ROW grape .495222 .033521 14.7735 0.000 Israelis other citrus .337320 .027963 12.0630 0.000 Italian other citrus .339260 .024568 13.8088 0.000 ROW other citrus .323420 .027410 11.7992 0.000 φ -1.78992 .268601 -6.66385 0.000 K1 .733602 .182670 4.01601 0.000 K2 19.5402 6.19318 3.15512 0.002 K3 -18.1022 14.9335 -1.21219 0.225 K4 26.6874 14.1197 1.89009 0.059 K5 31.4253 16.4634 1.90880 0.056 K6 9.80751 6.01929 1.62935 0.103
.
183
Table F-4. Marginal value shares of fruit juices in a two-stage block-wise dependent Rotterdam model
Product estimates SE t-statistics P-value First-stage ( )gΘ
Orange .751464 .028163 26.6830 0.000 Grapefruit .037536 .007504 5.00159 0.000 Apple .168712 .020470 8.24180 0.000 Pineapple .002608 .003264 .798972 0.424 Grape .021447 .008919 2.40436 0.016 Other citrus .018233 .005562 3.27800 0.001
Second-stage ( )'iθ
U.S. orange .045940 .011898 3.86129 0.000 Brazilian orange .948666 .013595 69.7803 0.000 ROW orange .005393 .006911 .780370 0.435 U.S. grapefruit .970290 .037857 25.6301 0.000 Israelis grapefruit -.053392 .033523 -1.59268 0.111 ROW grapefruit .083102 .019281 4.30995 0.000 U.S. apple .309579 .028504 10.8608 0.000 Chinese apple .293365 .023683 12.3870 0.000 ROW apple .397056 .033429 11.8774 0.000 Thai pineapple .371217 .024475 15.1671 0.000 Philippine p. apple .578753 .028411 20.3710 0.000 ROW pineapple .050030 .034488 1.45064 0.147 U.S. grape .363085 .036141 10.0462 0.000 Argentinean grape .028087 .018890 1.48685 0.137 ROW grape .608828 .033910 17.9541 0.000 Israelis other citrus .290771 .028499 10.2027 0.000 Italian other citrus .345494 .024718 13.9772 0.000 ROW other citrus .363735 .029818 12.1985 0.000
184
Table F-5. Relative price coefficients of fruit juices in a two-stage block-wise dependent Rotterdam model
Product estimates SE t-statistics P-value First-stage ( )ghV
Orange -1.08115 .092591 -11.6766 0.000 Orange/grapefruit -.017084 .017447 -.979229 0.327 Orange/apple -.225097 .045249 -4.97466 0.000 Orange/pineapple .013484 .008623 1.56358 0.118 Orange/grape -.018079 .022379 -.807850 0.419 Orange/other citrus -.017143 .013417 -1.27770 0.201 Grapefruit -.033411 .012913 -2.58748 0.010 Grapefruit/apple .008405 .017001 .494383 0.621 Grapefruit/pineapple -.030916 .005324 -5.80600 0.000 Grapefruit/grape .015116 .011628 1.29997 0.194 Grapefruit/other citrus -.009295 .007211 -1.28886 0.197 apple -.122978 .045207 -2.72035 0.007 Apple/pineapple .029394 .008759 3.35558 0.001 Apple/grape -.011129 .021099 -.527482 0.598 Apple/other citrus .019422 .012870 1.50910 0.131 pineapple -.036891 .005172 -7.13208 0.000 Pineapple/grape .010691 .006959 1.53630 0.124 Pineapple/other citrus .009568 .004232 2.26061 0.024 grape -.062155 .021528 -2.88713 0.004 Grape/other citrus .027168 .009685 2.80498 0.005 Other citrus -.062355 .008608 -7.24365 0.000
Second-stage ( )ijν
U.S. orange/Brazilian orange .017866 .031392 .569128 0.569 U.S. orange/ROW orange .007783 .006843 1.13729 0.255 Brazilian orange/ROW orange .017496 .015275 1.14538 0.252 U.S. grapefruit/Israelis grapefruit .103170 .061379 1.68086 0.093 U.S. grapefruit/ROW grapefruit -.136896 .031218 -4.38522 0.000 Israelis grapefruit/ROW grapefruit .010289 .005612 1.83326 0.067 U.S. apple/Chinese apple -.151502 .017968 -8.43157 0.000 U.S. apple/ROW apple -.229964 .019423 -11.8395 0.000 Chinese apple/ROW apple -.195847 .021253 -9.21488 0.000 Thai. pineapple/Philippines pineapple -.377307 .028929 -13.0424 0.000 Thai. pineapple/ROW pineapple -.033435 .021830 -1.53162 0.126 Philippines pineapple/ROW apple -.050163 .033530 -1.49607 0.135 U.S. grape/Argentinean grape -.010914 .012270 -.889442 0.374 U.S. grape/ROW grape -.371457 .024538 -15.1378 0.000 Argentinean grape/ROW grape -.028606 .020595 -1.38900 0.165 Israelis citrus/Italian citrus -.171333 .017623 -9.72188 0.000 Israelis citrus/ROW citrus -.180312 .014563 -12.3816 0.000 Italian citrus/ROW citrus -.219075 .017856 -12.2687 0.000 ROW citrus
185
Table F-6 Marginal value shares of fruit juices in a two-stage block-wise dependent uniform-substitute-Rotterdam model
Products Estimates SE t-statistics P-value First-stage ( )gΘ
Orange .756406 .027528 27.4781 0.000 Grapefruit .040781 .007866 5.18436 0.000 Apple .156724 .019876 7.88518 0.000 Pineapple .004118 .003230 1.27475 0.202 Grape .021782 .008928 2.43948 0.015 Other citrus .020188 .005502 3.66866 0.000
Second-stage ( )'iθ U.S. orange .037614 .010406 3.61459 0.000 Brazilian orange .950939 .013520 70.3377 0.000 ROW orange .011447 .003791 3.01952 0.003 U.S. grapefruit .887821 .034415 25.7974 0.000 Israelis grapefruit .065535 .022347 2.93255 0.003 ROW grapefruit .046644 .015161 3.07647 0.002 U.S. apple .251557 .027094 9.28458 0.000 Chinese apple .310132 .022592 13.7275 0.000 ROW apple .438311 .031785 13.7898 0.000 Thai pineapple .366144 .023983 15.2667 0.000 Philippine pineapple .588128 .026478 22.2123 0.000 ROW pineapple .045727 .030486 1.49996 0.134 U.S. grape .329439 .034654 9.50646 0.000 Argentinean grape .070893 .017576 4.03346 0.000 ROW grape .599668 .033227 18.0478 0.000 Israelis other citrus .339165 .026889 12.6135 0.000 Italian other citrus .322025 .024246 13.2814 0.000 ROW other citrus .338810 .026834 12.6262 0.000 K1 .656030 .211131 3.10722 0.002 K2 15.4812 5.81309 2.66316 0.008 K3 -17.1306 10.9864 -1.55925 0.119 K4 182.938 190.418 .960718 0.337 K5 30.5385 19.0871 1.59996 0.110 K6 19.5314 13.7109 1.42451 0.154
186
Table F-7 Relative price coefficients of fruit juices in a two-stage block-wise dependent uniform-substitute-Rotterdam model
Products Estimates SE t-statistics P-value Orange -1.21855 .095938 -12.7014 0.000 Orange/grapefruit -.018566 .018860 -.984413 0.325 Orange/apple -.229599 .045051 -5.09648 0.000 Orange/pineapple .011628 .855837E-02 1.35863 0.174 Orange/grape -.021365 .023077 -.925846 0.355 Orange/other citrus -.020387 .013641 -1.49452 0.135 Grapefruit -.041094 .014669 -2.80141 0.005 Grapefruit/apple -.153681E-02 .018222 -.084339 0.933 Grapefruit/pineapple -.030683 .570167E-02 -5.38142 0.000 Grapefruit/grape .021011 .012734 1.65002 0.099 Grapefruit/other citrus -.983250E-02 .777688E-02 -1.26432 0.206 apple -.102394 .045164 -2.26717 0.023 Apple/pineapple .024435 .859537E-02 2.84279 0.004 Apple/grape -.020097 .021009 -.956610 0.339 Apple/other citrus .019055 .012798 1.48885 0.137 pineapple -.035858 .471975E-02 -7.59752 0.000 Pineapple/grape .011926 .696278E-02 1.71287 0.087 Pineapple/other citrus .010403 .425146E-02 2.44689 0.014 grape -.057979 .021841 -2.65464 0.008 Grape/other citrus .023401 .982678E-02 2.38131 0.017 Other citrus -.062588 .851760E-02 -7.34811 0.000
187
LIST OF REFERENCES
Abbot, P. and P. Paarlberg. 1986. “Modeling the Impact of the 1980 Grain Embargo.”
Embargoes, Surplus Disposal and U.S. Agriculture. U.S. Department of Agriculture, Economic Research Service Staff Report No. AGES860910. Washington, D.C.: ERS/USDA.
Alston, J.M., C.A Carter, R. Green, and D. Pick. 1990. “Whither Armington Trade
Models.” American Journal of Agricultural Economics 72(1990): 455–67. Armington, P. 1969. “A Theory of Demand for Products Differentiated by Place of
Production.” International Monetary Fund Staff Papers, 16:159-178. Babula, R. 1987.“An Armington Model of U.S. Cotton Exports.” Journal of Agricultural
Economics Research 39:12–22. Barten, A. 1964. “Consumer Demand Functions under Conditions of Almost Additive
Preferences.” Econometrica 32:1–38. _____. 1969. “Maximum Likelihood Estimation of a Complete System of Demand
Equations.” European Economic Review 1:7–73. _____, 1977. “The Systems of Consumer Demand Functions Approach: A Review.”
Econometrica 45:23–51. _____. 1993. “Consumer Allocation Models: Choice of Functional Form.” Empirical
Economics 18:1:129–58. Blundell, R. and J.M. Robin 2000 “Latent Separability: Grouping Goods without
Weak Separability” Econometrica 68: 53-84. Brown, M. 1993. “Demand Systems for Competing Commodities: An Application of
the Uniform Substitute Hhypothesis.” Review of Agricultural Economics 15(3): 577-589.
Brown, M., J. Lee and J. Seale. 1994. “Demand Relationships among Juice
Beverages: A Differential Demand System Approach.” Journal of Agricultural and Applied Economics 26(2): 417-29.
188
Brown, M. and J., Lee. 1997. Incorporating Generic and Brand Advertising Effects in the Rotterdam System. International Journal of advertising 16:211-220.
Brown, M., and J. Lee and J. Seale. 2000. “A Uniform Substitute Demand Model
with Varying Coefficients.” Journal of Agricultural and Applied Economics 32(1): 1-9.
Clements, K. and H. Theil. 1978.“A Simple Method of Estimating Price
Elasticities in International Trade.” Economics Letters 1: 133–37. Clements, K. and L. Johnson. 1983. “The Demand for Beer, Wine and Spirits: A
System-Wide Analysis,” Journal of Business 56:273-304. Clements, K and E. Selvanathan, 1988. “The Rotterdam Model and its Application in
Marketing.” Marketing Science, 7, 1:60-75. Davis, G. and N. Kruse. 1993. “Consistent Estimation of Armington Demand
Models.” American Journal of Agricultural Economics 75(3): 719–23. Duffy, M. 1987. “Advertising and the Inter-product Distribution of Demand: A
Rotterdam Approach.” European Economic Review 31:1051-1070. Duffy, P., M. Wohlgenant, and J.Richardson.1990. “The Elasticities of Export
Demand for U.S. Cotton.” American Journal of Agricultural Economics 72:468-74.
Eales, J., C. Durham, and C. Wessels 1997. “Generalized Models of Japanese Demand for Fish.”. American Journal of Agricultural Economics 79:1153–63.
Fabiosa, J., and J. Ukhova. 2000. “New Aggregate and Source-Specific Pork Import
Demand Elasticity for Japan: Implications to US Exports.” Working Paper 00-WP 253. Center for Agricultural and Rural Development, Iowa State University, Ames, Iowa 50011-1070.
Fan, S., Wailes, E. and Cramer, G. 1995. “Household Demand in Rural China: A Two-
stage LES-AIDS Model.” American Journal of Agricultural Economics 77 (1):54-62.
FAO (Food and Agriculture Organization). Agricultural data. Available at http://faostat.fao.org/faostat/collections?version=ext&hasbulk=0&subset=agriculture. (20 March 2006)
Fousekis, P and B., Revell. 1990. “Meat Demand in the UK: A Differential
Approach.” Journal of Agricultural and Applied Economics. 32 (1):11–19
189
Frisch. 1959. “A Complete Scheme for Computing all Direct and Cross Demand Elasticities in a Model with Many Sectors.” Econometrica 27:177-196
Gao, X., E. Wailes, G. Cramer. 1996. “A Two-stage Rural Household Demand Analysis: Micro Data Evidence from Jiangsu Province, China.” American Journal of Agricultural Economics. 78(3):604-613.
Goodwin, J. 1994. Agricultural Price Analysis and Forecasting. New York: John Wiley
and Sons, Inc. Greene, W., 2000. Economic Analysis. Upper Saddle River, NJ: Prentice Hall.
Grennes, T., P.R. Johnson, and M. Thursby. 1977. The Economics of World Grain
Trade. New York: Praeger Publishers.
Hall, B and C. Cummins. 1999. Time Series Processor version 4.5. Reference Manual. TSP International Palo Alto, California, USA
Haniotis, R. 1990. “European Community Enlargement: Impact on U.S. Corn and
Soybean Exports.” American Journal of Agricultural Economics 72: 289–97. Huang, S. 2004. Global Trade Patterns in Fruits and Vegetables. Agriculture and Trade
Report No. WRS-04-06. Available at http://www.errs/usda-gov/publications/wrs0406/wrs040dfm.pdf (11 April 2006)
JETRO (Japan External Trade Organization). Market Information for Fruit Drinks.
Available at http://www.jetro.go.jp/en/market/reports/jmr/024.pdf. (3 March 2005)
Lee, J., J. Seale, and P. Jierwiriyapant. 1990. “Do Trade Agreements Help U.S. Exports? A Study of the Japanese Citrus Industry.” Agribusiness 6: 505-14.
Lee, J., M. Brown, and J. Seale. 1992. “Demand Relationships among Fresh Fruit and
Juices in Canada.” Review of Agricultural Economics 14:255–62. Lee, J., M. Brown, and J. Scale 1994. “Model Choice in Consumer Demand Analysis:
Taiwan 1970–89.” American Journal of Agricultural Economics 76:504–12.
Sarris, A. 1981. “Empirical Models of International Trade in Agricultural
Commodities,” in Imperfect Markets in Agricultural Trade, edited by A. McCalla and T. Josling. Montclair: Allenheld, Osmun and Co.
_____. 1983. “European Community Enlargement and World Trade in Fruits and Vegetables.” American Journal of Agricultural Economics 65: 235–46.
190
Schmitz, T. and J. Seale. 2002. “Import Demand for Disaggregated Fresh Fruits
Japan.” Agricultural and Resource Economics Review 34(3): 585-602. Seale, J. 1996. Import Demand for Products Differentiated by Place of Production.
Intentional Working Paper Series IW96-13, University of Florida, Gainesville, FL 32611
Seale, J. 2003. “Uniform Substitute When Group Preferences are Block-wise
Dependent.” Journal of Agricultural and Applied Economics 35(0): 51-55.
Seale, J., A. Sparks and B. Buxton. 1992. “A Rotterdam Application to International Trade in Fresh Apples: A Differential Approach.” Journal of Agricultural and Resource Economics 17(1): 138-49.
Selvanathan, E. and S. Selvanathan, . 2004. “Economic and Demographic Factors in
Australian Alcohol Demand.” Applied Economics 36:2405-2417. Soshnin, A., W. Tomek and H. Gorter. 1999. “Elasticities of Demand for Imported
Meats in Russia.” Working Paper 19, Department of Applied Economics and Management, Cornell University, Ithaca, New York 14853-7801.
Spreen,T.,R. Barber, M.Brown,, A. Hodges, J.Malugen, D. Mulkey, R., Muraro, R.
Norberg, M. Rahmani, F. Roka and R. Rouse 2006. An Economic Assessment of the Future Prospects for the Florida Citrus Industry. Available at http://www.fred.ifas.ufl.edu/economic_assess_flciturus_indus.pdf. (14 April 2006)
Statistics Bureau of Japan’s Ministry of Internal Affairs and Communications Population
Eestimates. Available at (http://www.stat.go.jp/english/data/jinsui/2-2.htm (12 March 2006)
Theil, H. 1965. “The Information Approach to Demand Analysis”. Econometrica
33:67–87. ______. 1971. Principles of Econometrics. 1971. New York: John Wiley and Sons, Inc. ______. 1975. Theory and Measurement of Consumer Demand, Vol. I and II.
Amsterdam: North Holland. _______. 1980(a). The System-Wide Approach to Microeconomics. Chicago: University
of Chicago Press. _______. 1980(b). System-Wide Explorations in International Economics, Input-Output
Analysis and Marketing Research. North-Holland Publishing Company, Amsterdam.
191
Theil, H. and Clements, K.1978. “A Differential Approach to U.S. Import demand.”
Economics Letters 1:249-252.
UNCTAD (United Nations Conference on Trade and Development). Citrus fruit. Available at. http://r0.unctad.org/infocomm/anglais/orange/market.htm (12 March 2006)
University of Pretoria. Global citrus industry. Available at http://www.up.ac.za/academic/fabi/citrus/global.html. (12 March 2006)
USDA (United States Department of Agriculture). Japan: Trade. Available at http://www.ers.usda.gov/Briefing/Japan/trade.htm. (12 March 2005)
Washington, A. 2000. The derived demand for imported dairy products in selected
international markets. PhD dissertation, University of Florida, Gainesville, FL 32611
Washington, A. and R. Kilmer. 2002. “The Production Theory Approach to Import
Demand Analysis: A Comparison of the Rotterdam Model and the Differential Production Approach.” Journal of Agricultural and Applied Economics 34(3): 431-43.
Weatherspoon, D., and J. L. Scale. 1995. “Do the Japanese Consumers Discriminate
AgainstAustralian Beef Imports? Evidence from the Differential Approach.” Journal of Agricultural and Applied Economics 27(2):536–543.
Yang, R. and W. Koo. 1994. “Japanese Meat Import Demand Estimation with the Source Differentiated AIDS Model.” Journal of Agricultural and Resource Economics.19: 396:408.
Zhang, P., S. Fletcher and D. Carley. 1994. “Japan’s Peanut Import Demand:
Implications for United States Exports.” Agricultural Economics 11 (1):51-59.
192
BIOGRAPHICAL SKETCH
Shiferaw Tesfaye Feleke was born in Ethiopia on 13 July 1971. He joined the
Alemaya University of Agriculture in Ethiopia in 1988 and received a Bachelor of
Science degree in agricultural economics in 1992. He worked for the Awassa
Agricultural Research Center (Southern Ethiopia) for eight years before he joined the
Food and Resource Economics Department at the University of Florida in 2000 for
further education. He received the Master of Science degree in agricultural economics in
2002 and continued a PhD program in the same field in the same department and
received a PhD in 2006.