- The purpose of this study is to comprehend food demand
structure and its changes under rapid economic development
theoretically and statistically. The recently developed Far
East Asian countries, Japan, Korea, and Taiwan were chosen
to obtain implications for the food demand patterns in the
future industrialized countries around the world. Food
demands for nine food commodities, rice, bread/wheat,
barley, beef, pork, chicken, fish, eggs, and milk were
analyzed. The first three commodities are plant origin, the
rest are animal origin. Study periods are from the early
1910's to the end of 1980's for Japan (the end of 1930's to
the early 1950's were excluded because of the war period),
and from the early 1960's to the end of the 1980's for both Korea and Taiwan. The importance of income growth on food
demand changes in developing countries has been stressed.
Many studies have been done based on a simple model using
per capita income as the only explanatory variable, or at
most including the prices of own and closely related
commodities. This study employed a more versatile
analytical framework, incorporating a wider range of cross
price effects. This study has two main objectives; the
first is to reconsider the effect of income growth on food
demands, particularly to examine whether income elasticities
change between various stages of economic development. The
other is to evaluate non-economic factors that cause changes
in food consumption patterns under economic development.
Age-population composition and household size were two of
the explanatory variables. A complete demand system by
adding dynamic and demographic features to DEATON and
MUELLBAUER's (1980a, b) LA/AIDS model. Data were complied
from various secondary sources. Price and quantity data
sets passed nonparametric tests of stability of preferences.
Two different estimation techniques; an iterative SUR
(maximum likelihood) estimation and a single equation
estimation using the homogeneity condition and first order
autocorrelation were applied for the demand system.
Assuming weak separability for the group of foods in the
study, the expenditure elasticities calculated by the demand
system were converted to the ones equivalent to income
elasticities. Major findings were:
1) the impact of the "pure" income effect was not
significant. However, effects of age-population composition
changes and own and cross price effects were significant.
The impacts from changes in own price level and/or age-population
composition exceeded the impacts from changes in
expenditure level most frequently for animal origin foods.
Significant cross price effects between animal and plant
origin foods were observed.
2) Various patterns of changes in income elasticities
were observed. Unexpectedly, some animal origin foods such
as beef, chicken, and eggs showed negative expenditure
elasticities at low income level. This phenomenon was
observed across countries and across time periods.