An Investigation into the Nature of Nutrition in Post-War Occupied Japan | Aaron Goh
An Investigation into the Nature of Nutrition in Post-War Occupied Japan Aaron Goh
This paper explores issues related to nutrition in Japan during the post-war period of 1945 to 1952. While there is evidence supporting various historians’ claims of large shortfalls in food, there is an absence of data to sufficiently back the stronger claim that this resulted in massive starvation and famine. This paper takes a closer look at statistics collected during the time period, available through the Historical Statistics of Japan, as well as the School Health Examination Survey. The former contains information about food consumption (an input variable), whilst the latter concerns the physiques of youths (an output variable). Given the problematic nature of data collection in the post-war period, these sources of data are used to corroborate and challenge each other in order to get a better picture of the situation. It is important to keep in mind that the analysis can only be regarded as accurate only to the extent that the original data can be regarded as reliable. Documents from the General Headquarters of the Supreme Commander of the Allied Powers (GHQ-SCAP) reveals that the GHQ-SCAP faced similar issues with regards to the reliability of data as we do now. Data analysis reveals that starvation was likely to be significantly less than thought, insofar as average caloric consumption levels were close to the recommended daily intakes. However, output data reveals that under-nutrition was common. Youths outside the purview of protein-rich school lunch programs experienced poorer growth rates in relation to historical averages, while those under the program experienced increased weights in relation to the same averages. This paper attempts to study such differences in data in relation to the political circumstances of the moment, such as the pressure for targeted food aid for younger children. top of page

• • •

The occupation of Japan by the Allied Forces following the end of World War II lasted for a period of slightly under seven years, between 1945 and 1952. This primarily American-led occupation was perceived by then-contemporary intellectuals as a “[free hand] to experiment” (Kawai 1951, 23) in the governance of a post-war Japan. It was during this period that Japan faced threats regarding the nutrition of its citizens due to a reduction in cultivated land and agricultural output right after the war1, and also due to the initial hesitation by Occupation authorities to provide more food aid because of the perceived “alarmist” nature of the Japanese government (Aldous 2010, 231). Historian Andrew Gordon keenly notes this food shortfall, stating that “millions of people faced starvation” (Gordon 2003, 228), and researchers in other fields similarly note “[high] mortality rates of children because of malnutrition” (Kibayashi, et al. 1998, 2)

 This paper will examine the claims of food shortages during the Occupation era of 1945 to 1952 by examining both primary data and secondary literature on the matter. In particular, the Historical Statistics of Japan will be used to track food consumption during the time period. It will also observe possible impact of changes in nutrition by corroborating the nutritional data with statistics on the weights of youths as tracked by the Japanese Ministry of Education in its annual School Health Examination Survey. It must be noted that it is not possible to disentangle nutrition from politics; the nature of Occupation era politics resulted in attempts by parties within and outside of the Allied Occupation authorities pursuing their own objectives, possibly at the expense of others. Because of this, we see questionably recorded data. My task here will be to carefully examine the statistics collected, and corroborate it with as many other forms of statistics and primary sources as possible. Furthermore, I will examine secondary sources regarding Occupation era politics so as to better understand the dynamics of the Occupation.

At this point, it should be noted that the recorded statistics are incomplete. Daily average caloric consumption was not recorded during the entire period of Occupation. This leaves us with having to extrapolate how many calories were consumed, based on the data from 1946 to 1952 on the weight of each type of food consumed. Though this is an imprecise science, it is the best information that the data can yield; it should be examined in relation to other more complete sources of data. More troubling, however, is the complete lack of recorded statistics for the year of 1945. Although the surrender and subsequent Occupation began only in late 1945, our inability to acquire data for this year is highly problematic, as 1945 was considered one of the “crisis years” (Aldous 2010, 255) in which food supplies and distribution were incomplete.

 This paper generally agrees with the findings in Contesting Famine: Hunger and Nutrition in Occupied Japan (Aldous 2010). The claims of famine can be contested, insofar as there is a significant amount of data, in the form of average caloric consumption and student physiques that rejects the assertion that Japan experienced severe food deprivation during the Occupation. Yet, the claims of slight levels under-nutrition are backed by a review of the statistics, which revealed an average caloric count that potentially implies malnutrition amongst a portion of the population; furthermore, protein consumption from legumes and animal-products remain low throughout the period2, and more than 80% of calories were obtained from cereal or potatoes. As Aldous asserts, there was indeed “much suffering and deprivation among Japanese” (Aldous 2010, 255) during that period. However, the claims to a massive famine are largely unfounded based on the data.

 To begin with, it is important to examine the reliability of the various claims and data that I examined. This paper finds that there is conflicting information regarding the actual consumption of food during the period of 1945 to 1952. Whilst there were numerous reports of starvation or malnutrition (Gordon 2003, 228), we are unable to find corroborating data regarding caloric consumption in the Historical Statistics of Japan. Furthermore, domestic agricultural output growth far outpaced population growth during the time period, suggesting an increase in food availability. The only piece of evidence in the dataset that suggests a problem with nutrition is the falling average weights of students aged 14 to 20, during 1945 to 1952.

 Results from simple data analysis from the Historical Statistics of Japan, published in 1987, show that there was at least an approximate average consumption of 2000 calories a day, above the level of starvation for an average person. This is set in contrast to the 1800 calories calculated by the Occupation authorities in late 1945 (Aldous 2010, 239), and significantly above the 1280 calories that was defined as the bare minimum for a Japanese adult (Aldous 2010, 239). This calls into question which data is more accurate. Aldous suggests that the Occupation data might be mired by the fact that Japanese authorities had the incentive to “distort the picture to their advantage” (Aldous 2010, 241) to request for more aid. Such discrepancies alert us to the danger of relying on a singular source for statistical data.

In order for us to make sense of this conflicting information, I venture the hypothesis that most Japanese faced levels of slight under-nutrition during the period of the Occupation. This could only be possible if the normal average calorie consumption levels belie the differences across regions; it is highly likely that people living in rural areas had higher daily calorie intake than those in urban areas, thus averaging out to an approximate intake of 2000 calories daily across Japan. The nature of Japanese demography during this time period allows for this mathematical process to happen. In the 1950 census, summed population in all cities was 31.3 million, out of a nation-wide total of 84.1 million people3. This means that the urban population formed 37% of the country, leaving the remaining 63% as either rural or living in small towns. Given how the 1947 Food and Fertilizer Mission argued that the minimum calorie consumption for people living in rural regions should be at least “2120 calories per person per day” (Aldous 2010, 247), this suggests that the average amount of food consumed by an urban-dweller would be lower than 2000 calories, the amount recommended by today’s USA FDA. It is crucial to note that such a projection is mainly conjecture based on the data available. Without a breakdown of the consumption of food by regions, we are unable to make conclusive claims regarding the true nature of how caloric consumption was distributed across the different regions and peoples.

As such, in order to buttress our argument with more sources of data, we will turn to proxies for food consumption as well, namely agricultural production and the weights of youths. We note an agricultural collapse in 1945 relative to preceding years. The total area of cultivated land is reported to be 5301 thousand acres, compared to a peak of 6048 thousand acres in 1937. Even more significantly, milk production halved, from more than 300 thousand tons in 1944, to slightly over 150 thousand tons in 1945. Hen’s egg production, along with pig and cattle, also plummeted between 1944 and 1945. Yet, after the collapse of 1945, there was a massive expansion of paddy fields between 1945 and 1951, and the 1946 rice harvest “measured up to prewar standards” (Aldous 2010, 245). Furthermore, hen egg, pig and cattle production was restored to former peak levels at the end of the Occupation4.

The increase in food production should be set in the context of population growth during the same time period, of approximately 14% from 1945 to 19525 based on historical statistics. Such a rate of population growth was vastly outpaced by the rapid rise in food production, with a more than 30-fold increase in hen’s eggs and pig carcasses produced, and a more than 3-fold increase in cow carcasses and milk. This suggests an improvement in the levels of food available for domestic consumption, holding other factors constant. In order to investigate this further, we will have to explore the changes in levels of food imports and aid during this time period. However, we note no significant change in the amount of general food aid that was provided to Japan, other than an early initial rise in July 1946, following repeated pleas by the head of SCAP’s special Japanese food commission, Colonial Harrison (Aldous 2010, 241).

In reference to food supply, it might be useful to examine the claim of William Ball, the British Commonwealth representative on the Allied Council for Japan, namely that the crisis of food during the Occupation years “has mainly been a crisis of distribution, not of production” (Ball 1948, 103). If assertion were true, then the increases in agricultural output might not have had an appreciable impact on food consumption. Evidence of this exists in the form of the high prices of food in the black market, which Ball claims to have been “up to twenty times” (Ball 1948, 158) the official price. Furthermore, even official prices for food quadrupled between June 1946 and June 1947 (Ball 1948, 158). Slightly later data from the Historical Statistics of Japan corroborates Ball’s early claim; official prices for food almost doubled between 1947 and 19526 within the city of Tokyo despite early attempts at price control and “fifty per cent [defrayment] of this cost from government sources”7. This suggests to us that a possible inequitable distribution of food led to malnutrition among the groups who were least able to access or afford food during this time period.

We have thus far explored the use of official calorie consumption data and agricultural production statistics as instrumental variables for us to tease out the actual impact of the Occupation on nutrition in Japan. However, as we have noted, both measures are highly problematically. The former does not account for differences between consumption in rural and urban areas. Since the daily caloric needs vary for people performing differing tasks, it is difficult to tell if the average consumer was under-nourished. The latter data of agricultural production only succeeds as an input variable; it does not necessarily reveal consumption of agricultural produce, especially since there were problems with food distribution. This signals the need to adopt a different tack with regards to nutrition. As such, we will examine an output variable instead, in the form of the weights of youths during this time period.

We observe increases in the average weight of both male and female youths between 6 and 10 years old, during the period of 1945 to 1952. However, the average weights of almost all cohorts of youths between 12 and 20 fell during the same time period8. The maximum increase was among 8-year old males, in which the 1952 cohort displayed a 4.1% weight gain over the 1945 cohort; the greatest decrease was among 20-year old males, who displayed a 3.5% decrease in the same time period. This curiosity can be explained by the fact that the school lunch programs promoted by the General Headquarters of the Supreme Commander for the Allied Powers (GHQ-SCAP) in 1947 (Kibayashi, et al. 1998, 15) were open only to elementary school children (Parker 2010, 27). These programs were designed to raise the consumption of protein through skimmed milk powder, and were conducted at least twice a week. Aside from providing “more than 300 calories per meal” (Parker 2010, 28), the school lunch program also served to deliver other “much needed proteins and vitamins” (Parker 2010, 28) among school children. Amongst these food items were “canned meats, rice… [and] occasionally fresh whale meat and other fish” (van Staaveren 1994, 93). The fact that average weights of elementary school children went up during the time period suggests that the program was successful. Furthermore, evidence of the program’s efficacy emerges from the revelation that the average weights of children between 12 and 20 fell. The combination of both factors suggests that the independent variable of school lunches made a difference to the nutritional intake of the youths.

On the flip side, the data trend provides another equally important insight. If we are to remove the confounding factor of provided school lunches, we notice a drop in average weights over the time period. This suggests strongly that there was a general under-consumption of food between 1945 and 1952. There are indeed clear problems with the data collected, namely that it only looks at the weights of school-going youths. This makes our data unrepresentative of the population at large. Yet, despite the limitations of data, it is the best analogue to the general population that is available. Furthermore, it is highly likely that the weights for 12 to 20 year olds were biased upwards. Even in 1970, only 54 percent of the lowest fifth of the income group attending high school whereas more than 80 percent of the rest of the population did (Sato 1987, 14); it is highly suggestive that this was also the case in the immediate post-war years. Since the statistics of weights were recorded from youths attending schools, it suggests to us that a more complete sample that included those of lower income groups would be more likely to present an even steeper trend of decline in weights. Our limited data, along with some understanding of the historical circumstances, supports the idea that there was some level of under-nutrition that was taking place in Occupation-era Japan.

Having parsed through most of the readily available data, it is fruitful to look at the discourse surrounding nutrition within the GHQ-SCAP during the period of Occupation. It is clear that much attention was paid to issues of food, as evidenced by internal GHQ-SCAP memoranda about how the Imperial Japanese Government was to inform the GHQ if “stocks of staple foodstuff in any city… [were to] fall below ten days supply”9; furthermore, the GHQ commissioned studies on nutrition in civilian populations early in the Occupation period, in December 194510. Much of the discussions within the SCAP reveal that the authorities faced similar issues regarding data to those we have explored thus far. In particular, there are three broad issues worth examining, namely: first, the discussion amongst Allied administrators to adjust food aid levels; second, the definition of the supposed minimum caloric intake for a “normal” Japanese; third, the shifting discourse amongst Allied administrators on caloric intake, versus a broader approach to nutrition that includes different food types.

In reference to the discussion on the levels of food aid, there were considerable differences in the positions taken by the various key actors. Whilst the Japanese Ministry of Agriculture and Forestry requested 6.12 million metric tons in rice equivalents in 1946, Supreme Commander MacArthur only conveyed a request of 3.7 million tons to the Americans (Aldous 2010, 240). Yet, even this significantly reduced sum of 3.7 million metric tons was called into question by Robert Cogswell, chief of the Nutrition Branch of the Pentagon’s Civil Public Health and Nutrition Division, who privately remarked that MacArthur’s requests for food were “stupendous” (Aldous 2010, 240). At this point, Cogswell’s analysis of the data on caloric consumption reaches very similar conclusions to the ones held by this essay. He questions the accuracy of the Occupation’s cables, and wonder “if urban Japanese were reported to be consuming 2000 calories… [why] the ration could not be reduced so as to ‘extend locally-held food supplies” (Aldous 2010, 241). This situation reflects the fact that several top-level administration officials were possibly working with similar statistics to the ones that we have looked at thus far, and thus faced the similar issue of seemingly contradictory data.

 Yet, data-driven arguments were still crucial in driving the eventual decisions made. As revealed earlier, the head of SCAP’s special Japanese food commission, Colonial Harrison, argued strongly for a per annum import of 2.4 million metric tons of rice equivalents (Aldous 2010, 242) due to bad domestic rice harvests. Furthermore, this number was based on a calculation of the amount of rice and other grains available to the Japanese population; if the Americans failed to support his plan, the total amount of food available to the average Japanese would go down by 41.4 percent as compared to prewar levels (Aldous 2010, 242). This meant the threat of widespread starvation, which would then lead to a loss of prestige for the United States. Such an argument provided effective, and food aid was increased to levels above what was initially requested.

 The debate on what entailed a suitable “minimum” calorie consumption involved various numbers that were supposedly “scientifically determined” (Aldous 2010, 239). Yet, it remains unclear how these figures were actually derived. The Occupation’s Natural Resources Section briefing paper to ex-President Herbert Hoover in May 1946 detailed the minimum calorie consumption for the average Japanese to be 1280 calories, approximately 80 percent of the requirements for a Westerner (Aldous 2010, 239). However, the report did allow that the energy requirements were to be higher for normal consumers who had to conduct the various activities of daily life; the figure for such a consumer was put at 1800 calories per person per day. About the same period in May 1946, Cogswell, the chief of the Pentagon’s Nutrition branch, made independent estimates that the daily caloric needs of the “‘normal’ Japanese to be 1400 calories per day” (Aldous 2010, 241).

 The discrepancies in such numbers might have been cause by various reasons. First of these reasons is the issue of how the minimum caloric need was defined. Second, and perhaps more useful to this paper, is the issue of how the idea of the average Japanese was defined. Cogswell complained about the issue of creating a “caloric statement for Japan” (Aldous 2010, 241) because this included all kinds of people such as infants, children, urban dwellers and agricultural workers, all of whom have different caloric needs. This is reminiscent of some of the issues we have had in the examination of statistics of average caloric intake during the time period. A national average only does not tell us enough information to make detailed claims about which groups suffered from under-nutrition, if at all.

 Finally, Allied officials also broached the topic of diversifying Occupation-era metrics on the level of nutrition received. While mass starvation was avoided in 1946 by large imports of cereals, this was “a recipe for survival, not healthy living” (Aldous 2010, 248). The above-mentioned briefing paper for ex-President Hoover warned of the erroneous belief that “weight loss alone is indicative of malnutrition” (Aldous 2010, 249). Indeed, the American Kittredge-Tuck report of 1948 stated that 25 percent of adults in Japan showed “evidence of undernutrition” (Aldous 2010, 249). We have earlier noted this in our data, in reference to the protein intake by the average Japanese consumer. Consumption of both animal and vegetable protein was low throughout the period of 1945 to 1952. Less than 120 grams of animal products, and less than 70g of legumes consumed a day from 1945 to 1952. This works out to be less than the FDA recommended daily intake of 46g of protein for women and 56g for men.  This led to GHQ-SCAP policies such as the expansion of fisheries, which will not be covered in depth in this paper. In general, however, this internal discussion reveals an awareness of the fact that calories alone were an insufficient measure of the nutritional status for the Japanese.

 As has been demonstrated in this essay, a data-driven analysis of the Occupation problematizes the prevailing historical narrative of severe starvation and malnutrition among the Japanese. In particular, we observe an approximate average intake of about 2000 calories per person per day during the Occupation. Whilst this may be an under-consumption of food for certain individuals due to the nature of averages, the data hints that the issue of starvation was likely to be less widespread than thought. This statistic on a suitable average caloric intake is supported by data on increasing domestic agricultural output during the same time period. Yet, our observations on youths during this time period reveal a worrying trend of decreasing weights amongst 12 to 20 year olds, the demographic that did not have access to school lunches. Such an observation of decreasing weights amongst youths is likely correlated with the falling weights of the general population. This is a sign of under-nutrition in the wider population, although our other data hints that this was not significant enough to be considered a severe famine.

On the other hand, the weights of 6 to 10 year old youths, recipients of Occupation provided protein-rich school lunches, increased during the Occupation. This provides a highly tangible example of how GHQ-SCAP policies on nutrition directly impacted health outcomes. The brief examination of the discussions within the SCAP reveals official concerns regarding nutritional data that are similar to those raised by this paper’s examination of official statistics. These included the problematic nature of creating an average caloric intake for Japan and whether calorie intake represented a suitable metric to measure nutrition by. Regardless of this, however, there was little evidence in the SCAP documents that revealed nation-wide starvation. As such, the statistics and the information on SCAP discussions lead us to conclude that whilst there was a possibility of slight under-nutrition nationwide, and of more significant under-nutrition in isolated regions, the claim of mass starvation was likely to be overblown.

• • •


1. Historical Statistics of Japan, Area of Cultivated Land (1904—2004), Tokyo: Japan Statistical Association, 2012.

2. Less than 120g of animal products consumed a day between 1946 and 1952, as compared to 320g in 1985. Less than 70g of legumes consumed a day in the same time period. This works out to be less than the FDA recommended daily intake of 46g of protein for women and 56g for men.

3. Japan Statistical Handbook, Number of Cities, Towns and Villages and Population by Size of Population, Tokyo: Japan Statistical Association, 2013

4. Historical Statistics of Japan, Production of Raw Milk, Hen's Eggs, Chicken and Carcass (1894--2003), Tokyo: Japan Statistical Association, 2012.

5. Historical Statistics of Japan, Population by Sex, Population Increase and Decrease, Population Density (1872--2009), Tokyo: Japan Statistical Association, 2012.

6. Historical Statistics of Japan, Consumer Price Index by Five Major Groups (Prewar Base) for Ku-area of Tokyo (1947--2010), Tokyo: Japan Statistical Association, 2012.

7. General Headquarters Supreme Commander of the Allied Forces, Price Schedule for Rice, by Colonel Allen H. W., Tokyo, Japan, 1945.

8. Historical Statistics of Japan, Weight by Age and Sex, Tokyo: Japan Statistical Association, 2012.

9. General Headquarters Supreme Commander of the Allied Forces, Eeport on Stocks of Staple Foodstuff, by Colonel Allen H. W., Tokyo, Japan, 1945.

10. General Headquarters Supreme Commander of the Allied Forces, Nutrition Surveys of Civilian Populations, by Colonel Allen H. W., Tokyo, Japan, 1945.

• • •

Works Cited

Primary Sources

Government Sources

Historical Statistics of Japan. Tokyo: Japan Statistical Association, 2012.

Japan Statistical Yearbook. Tokyo: Japan Statistical Assocation, 2014.

Japan Statistical Handbook, Number of Cities, Towns and Villages and Population by Size of Population, Tokyo: Japan Statistical Association, 2013

General Headquarters Supreme Commander of the Allied Forces, Eeport on Stocks of Staple Foodstuff, by Colonel Allen H. W., Tokyo, Japan, 1945.

General Headquarters Supreme Commander of the Allied Forces, Price Schedule for Rice, by Colonel Allen H. W., Tokyo, Japan, 1945.

General Headquarters Supreme Commander of the Allied Forces, Nutrition Surveys of Civilian Populations, by Colonel Allen H. W., Tokyo, Japan, 1945.

Non-Governmental Sources

Kaisha, Henshūbu, and Kabushiki Santōsha. Shokuseikatsu dēta sōgō tōkei nenpō = Year of food consumption statistics.. Tokyo: , 2012.

W. MacMahon Ball Japan: Enemy or Ally? (London: Cassell, 1948)

van Staaveren, Jacob. An American in Japan, 1945-1948: A Civilian View of the Occupation. Seattle: University of Washington Press, 1994.

Secondary Sources

国立教育政策研究所 (National Institute for Education Policy Research). "School Lunch Program in Japan." Education_in_Japan_files/201303SLP.pdf (accessed June 3, 2014).

Aldous, Chris. "Contesting Famine: Hunger and Nutrition in Occupied Japan, 1945-1952." Journal of American-East Asian Relations 17: 230-256.

Gordon, Andrew. A modern history of Japan: from Tokugawa times to the present. New York: Oxford University Press, 2003.

Kawa, Kazushi. Agriculture and economic growth: Japan's experience. Princeton, N.J.: Princeton University Press, 1970.

Kawai, Kazuo. "American Influence on Japanese Thinking." The Annals of the American Academy of Political and Social Science 278: 23-31.

Kiyabashi, Etsuko, et al. "The change of dietary habits and the resultant changes in health of Japanese children." Ecology of Food and Nutrition 38: 1-19.

Parker, J. D. (2010). Policy and Reform in Public Schools during the Occupation of Japan. (Master’s dissertation). Retrieved from KU ScholarWorks.

Sato, J. (1987). Family Background and School Continuation Decisions in Japan: A Comparison Between Pre-War and Post-War Educational Systems. (Master’s dissertation). Retrieved from University of Wisconsin Social Science Computing Cooperative.

Wen, Chern. Analysis of the Food Consumption of Japanese Households. Rome: Food and Agriculture Organization of the United Nations, 2003.

Yamada, Tetsuji, Tadashi Yamada, and Chaloupka Frank. "Nutrition and Infant Health in Japan." Journal of Human Resources 24: 725-736.

Copyright © 2014 by Dartmouth Quarterly of East Asian Studies. All Rights Reserved.