Local Nutritional Availability and Adult Height: Evidence from South Korea, 1946-1977
It is generally accepted that the rapid economic growth in South Korea during the second half of the twentieth century greatly improved the standards of living of the country’s population. However, it is not fully understood how the living conditions in South Korea changed over time, and what are the major factors that produced the changes. For instance, there is still much to be learned about how the process of improvements differed by socioeconomic characteristics, and about how the experiences of each birth cohort differed with each other. We do not fully understand how particular economic or social changes affected the wellbeing of the population, either.
Rigorous research on these issues is often seriously hampered by shortage of appropriate data, especially for the periods prior to the 1970. In our previous study (Lee 2016), supported by the 2015 Cliometric Study Program of the KDI School of Public Policy and Management, we investigated how biological indices of standards of living, such as height, weight, and body-mass-index (BMI, hereafter) changed across birth cohorts born between 1946 and 1957, based on a newly collected sample of military records for more than 18,000 conscripts. We found that the mean height at conscription slightly declined from the 1946 cohort to the1951 birth cohorts before it rapidly increased across cohorts. The mean height increased by more than two centimeters in just 6 years between 1951 and 1957. We also found considerably large variations across provinces in the patterns of changes in anthropometric measures as well as the levels of the measures. In the previous study, however, we were unable to determine what factors produced the changes in anthropometric measures across different birth cohorts and across different regions. In the present study, we investigated how nutritional availability in two crucial periods for human growth, namely, early childhood (from conception to age 2) and adolescence (from age 12 to age 16), affected the heights of Korean conscripts born from 1946 to 1957.
For the purpose, we constructed province- and county-level data on agricultural productions, and matched the dataset with the sample of conscripts using the information on place of residence. We also explored how much improved nutrition during infancy and adolescence contributed to the increase in heights between the 1951 cohort and the 1957 cohort. We used the amounts of calories and three major nutrients (protein, fat, and carbohydrate) per farm household adult equivalent in a given province or county as our primary measures of local nutritional availability. In addition to nutrition variables, we considered local environmental conditions indicated by population density and nonfarm population share as well as personal and family characteristics such as season of birth, family size, and father’s occupation. We selected several different samples for whom the effects of our measure local nutritional availability on heights might differ with one another. These samples include: men from counties with information on nutritional availability, men from rural counties (baseline sample), and farmers’ sons living in rural counties. The regression results suggest that variables on local nutritional availability are generally positive and statistically significant. According to the results from the rural county sample, a person who spent infancy in a province that produced calories per farm population one standard deviation above the mean would have been about 0.2 centimeters taller at the time of conscription, if other things equal. A one standard deviation increase in protein or carbohydrate in infancy would result in an increase in height by a similar magnitude. The estimated coefficients for nutritional availability variables in adolescence are generally smaller in magnitude and statistically less significant. A major drawback of our analyses is that the true birth place of a conscript is unknown. To reduce potential measurement errors arising from geographic mobility between birth and conscription, we used a subsample of men from rural counties whose province of residence at the time of conscription is the same as the province of “original family place (Bonjeok).” If a conscript’s current address and Bonjeok are identical, it is likely that he was born in the current province of residence. About 79 percent of the full sample and 91 percent of the rural county sample reported the same current and Bonjeok province, which implies that prior geographic mobility of the conscripts in the sample was probably low. The regressions conducted based on the subsample reveal larger effects of nutritional availability on heights, compared to the baseline results. We performed height regressions with various samples alternative specifications. Adding the conscripts’ own education does not noticeably change the results. By contrast, inclusion of province fixed effect greatly increases the size of the effects of nutritional availability in early childhood.
The effects of nutrition availability variables in adolescence are stronger for the 1946-1950 birth cohorts than those for the 1951-1957 birth cohorts. If the sample is limited to farmers’ sons from rural areas, the estimated coefficients for nutrition variables become larger. If the sample is extended to include men from urban counties with minimum agricultural productions, the baseline results remain little changed. We finally explored how much improved nutrition contributed to the increase in height between the 1951 and 1957 birth cohorts. The cohort trends of nutritional availability and height match well, which is consistent with the hypothesis that early-life nutrition mattered for growth of stature for the birth cohorts under investigation. For estimating the contribution of improved nutrition to the increase in height, we computed the change in height predicted by change in nutrition using height regression results and estimated change in each nutritional availability variable between 1951 and 1957. The result suggests that improved nutrition in early childhood and adolescence accounts for 30 percent to 50 percent of the increase in adult height that was gained from the 1951 and 1957 birth cohorts. Increased nutritional availability during early childhood explains a lion’s share of the contribution.
The results of this study strongly suggest that nutrition was an important determining factor of biological standards of living indicated by adult height. In particular, provisions of calories and protein were strongly associated with larger stature. Food availability during early childhood was more critical for human growth than nutrition in adolescence. This result is consistent with the consensus that catch-up growth in adolescence is insufficient to fully make up for deficiencies in early childhood. Although our estimate is highly preliminary and subject to errors, it is likely that improvements in the quantity and quality of nutritional intakes that were made possible by increased incomes and enhanced agricultural productivity significantly contributed to the rise of biological standard of living indicated by changes in anthropometric measures.
Click the button and follow the links to connect to the full text. (KDI CL members only)
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.