202
H. Mori et al. /International Food and Agribusiness Management Review 3 (2000) 189-205
groups tending to consume less than the older age groups. There are very strong negative
cohort effects for newer cohort groups. On the whole, older cohorts have apparently positive
cohort effects. There is a negative trend in period effects, although not a substantial one. The
negative cohort effects for newer cohorts will lead to continued decline in overall at-home
sake consumption.
4.5. Beer (Fig. 5)
The analysis resulted in a grand mean of 19.3 l of beer consumed (at home) annually. The
age effects showed no clear trend, but were close to 0 for each age group. The currently
younger age groups had strong positive cohort effects of from 4 to 6 l. Those born before
1932 had increasingly negative cohort effects. There was a slight tendency for increasing
positive period effects. With the larger number of newer cohorts preferring beer than older
cohorts, consumption should continue to increase if the pure period effects do not operate
quite favorably.
4.6. Fresh fruit (Fig. 6)
Per capita fresh fruit consumption has been declining steadily from 45.2 to 31.7 kg in the
past 20 years. The grand mean from the analysis was 38.95 kilograms per year. The age
effects are negative for those between 20 and 50, other effects controlled. Older “pure” age
groups have positive age effects. More significantly than the case of rice and fish, those born
before 1952 have strong positive cohort effects, while those born after 1952 have sharply
increasingly negative cohort effects. The newest cohort group has a 232 kilogram cohort
adjustment to fruit consumption. There is no apparent trend in period effects. Most of the
decline in overall consumption levels can be explained through the cohort effects, the older
cohort groups being replaced by newer ones. It is expected that fruit consumption will
continue to decline at an increasing rate, unless the industry launches effective marketing
actions to reverse the negative cohort effects of the newer generations immediately, which
could include subsidized provisions of fresh fruits to school lunch programs for primary- and
middle-school children, nationwide (Rentz & Reynolds, 1991).
4.7. Appendant (Table 5)
The resulting period effects from the analysis were used to perform a least-squares
regression of fresh fruit consumption over time from 1979 to 1997. The grand mean was
adjusted by the amount of the period effect for each survey year. The resulting quantities
were regressed against real prices and per capita living expenditures. The same regression
was performed with actual per capita figures from the FIES instead of the estimated
quantities from our analysis.
Both regressions resulted in significant models. The major difference between the two was
the income elasticity estimates. Using the actual figures from the FIES resulted in a 20.59
income elasticity. The quantities from our cohort analysis resulted in a 10.38 income
elasticity (Table 5). Fresh fruit, however, is not considered an inferior good. Cross section-