“Altruism, Favoritism and Guilt…: Sophie’s Choice in Mao’s Mass Send-Down Movement,” H. Li, M. Rosenweig and J. Zhang (2010)

From 1966 to 1976, enormous numbers of urban youth in China were “sent down” to poor rural villages to do backbreaking farm labor. In many cases, families were required to send one of their members. Further, because during this period schools were essentially nothing but propaganda machines, the “sent down” youth acquired enough skills that being sent away positively biased future earnings potential versus staying in school. The authors use these facts, in conjunction with a custom survey of Chinese twins that includes intersibling transfers and the amount of wedding gifts from parents, to estimate the sign of “altruism” and “guilt” on the part of parents and siblings. That is, controlling for innate ability and current earnings, do parents give more the kid they chose to “send down” than they would otherwise? And do the siblings not sent down give transfers to the sibling who was sent down? A straightforward fixed effects OLS model allows identification of the signs of these transfers. The twin data shows clear evidence that parents transferred money later in life because of “guilt” and that siblings who were not sent down transfer money to their sent down sibling for the same reason. The authors consider a number of potential alternative explanations (for instance, the sent down child is more likely to be in the Communist Party today, hence will be able to reciprocate gifts from a parent using the political office) and find them all wanting.

In general, identifying economic behavior due to “guilt” is very difficult, precisely because guilt-inducing behavior tends to occur when agents are in long-term relationships, and thus experimental-type research on pairs of random undergraduate subjects will tend not to evidence such behavior. If the results of this paper are externally valid, then there is a fairly straightforward explanation for a number of economic paradoxes – such as why agents who “defect” in a repeated prisoner’s dilemma tend to “cooperate” later even when the opponent is clearly playing a grim trigger.

http://www-leland.stanford.edu/group/SITE/SITE_2009/segment_4/segment_4_papers/rosenzweig.pdf (link to WP – final version in JPE 118.1 (2010))

%d bloggers like this: