192
H. Mori et al. /International Food and Agribusiness Management Review 3 (2000) 189-205
Table 2
Per capita food supply, 1955, 1965, 1975 and 1995
Fiscal Year |
(kg/year) | |||
1955 |
1965 |
1975 |
1995 | |
Cereals |
155.5 |
145.0 |
122.2 |
102.0 |
(Rice) |
(110.4) |
(111.7) |
(88.1) |
(67.8) |
Meat |
3.2 |
9.2 |
17.9 |
31.3 |
Hen Eggs |
3.4 |
11.6 |
14.0 |
17.6 |
Milk and Dairy Products |
12.0 |
37.4 |
53.3 |
91.2 |
Fish and Shell Fish |
26.2 |
29.2 |
34.9 |
38.2 |
Vegetables |
82.2 |
109.6 |
111.3 |
105.5 |
Fruits |
12.3 |
28.5 |
43.0 |
40.7 |
Caloric Intake (kc/day) |
2,000 |
2,460 |
2,520 |
2,638 |
Sources: Food Balance Sheet, Ministry of Agriculture, Forestry and Fisheries, various issues.
economy leaped forward since then, people’s diets drastically changed toward a “western”
type, namely, greatly decreased cereal consumption, particularly rice, and increased con-
sumption of meat and dairy products (Table 2).
Those born before and during WWII came of age during the food shortage periods and/or
“periods of much cereal and little meat.” They may have formed their eating habits, if not
willingly, under circumstances quite opposed to those who came of age, say after the late
1960s when the economy reached the level of that of a developed economy, or those born
after 1970 or so who experienced affluence of everything in life during their formative years.
These generations may have different preferences for and patterns of food intake, the age
factor set aside.
Cohort analysis has been used in demography, sociology, and medical studies (Meredith
& Schewe, 1994; Ryder, 1965; Mason & Fienberg, 1985). Its use in other disciplines has
been much more limited. In economics, cohort analysis has been employed in rather
primitive form to examine the declining savings in the United States (Attanasio, 1997;
Gokhale et al., 1996). Cohort analysis of food consumption is relatively unexplored (Rentz
& Reynolds, 1991).
3. Research methods
Analysis of food consumption trends from the viewpoint of age is limited by the
availability and nature of the data. The Japanese government publishes yearly Family Income
and Expenditure Surveys (FIES). These surveys include consumption data for approximately
8,000 households. Beginning in 1979, the surveys report consumption and expenditures
according to age groups of head of household (HH). This makes it possible to address the age
issue.
It is necessary to derive individual consumption of household members by age from the
household consumption data. A common method of addressing this issue is simply to divide
household consumption by the average household size. This method has severe shortcom-