Body
A familiar trope of the North Korean famine literature is the assertion that the event would have significant long-term effects on the physical and mental capacities of individuals who were nutritionally deprived as children. While the notion has enormous intuitive plausibility, I know of no actual evidence on the North Korean case, and studies of previous famines have yielded mixed results.
A recent paper by George Mason University economists Sven Neelsen and Thomas Stratmann examine this issue using panel data from Greece. From a statistical standpoint, the 1941-42 Greek famine is interesting because it was relatively short, and had significantly different incidence in urban and rural areas. Those characteristics lend themselves to enabling Neelsen and Stratmann to try to extract the differential effect of famine on individuals who were one year olds at the time of the famine, those who were born during the famine, and those who were fetuses during the famine period.
Neelsen and Stratmann focus on educational attainment and find a statistically significant negative effect of famine exposure, but the results are relatively small, typically on the order of a 2-3 percent reduced likelihood of completing secondary education. However, their result is confounded by a sample selection issue of unknown magnitude, namely the fall in fertility that accompanies famine. If famine exposure is negatively correlated with socio-economic status as one might expect, this means that individuals born or conceived during famines may be disproportionately from relatively unaffected high status groups. Hence, they interpret their estimates as a floor on the actual effects.
Of course the brevity of the Greek famine experience that makes it attractive from the standpoint of statistical analysis may make is a poor comparator for the North Korean famine which was longer in duration. Research on the famine that accompanied the Chinese Great Leap Forward which may represent a more apt comparison, has found that famine exposure is associated with reductions in later life height and weight, labor market participation, income, and likelihood of marriage, as well as an increase in the likelihood of adult schizophrenia.
