Social Preferences Among Clinicians in Tanzania: Evidence from the Lab and the Field

Thumbnail Image
Publication or External Link
Brock, J. Michelle
Leonard, Kenneth L.
Health worker effort can have a dramatic influence on patient outcomes. This is especially true in developing countries, where poor overall quality of healthcare systems is the norm. There is evidence, however, that despite low levels of education and experience, clinicians in Tanzania underperform relative to their ability (Leonard and Masatu, 2005). Understanding clinicians' intrinsic motivations may help us identify nonmonetary incentives for improving quality of care. To this end, this dissertation considers how risk, pride and social information impact altruism among Tanzanian clinicians. In Chapter 4, we study how risky environments impact social preferences. With experimental evidence from games with risky outcomes, we establish that social preferences of players who give in standard dictator games are best described by consideration of equating ex ante chances to win rather than of ex post payoffs. The more money decision-makers transfer in the dictator game, the more likely they are to equalize payoff chances under risk. Risk to the recipient does, however, generally decrease the transferred amount. Also, while some people behave generously regardless of the attributes of others, pride and knowledge about the recipient characteristics may also motivate altruistic behavior. In Chapter 5, we explore the role of social information and pride in determining pro-social behavior among clinicians in Tanzania. We find that making someone feel proud increases the number of "fair" allocations (50/50 giving) and that those who do not respond to decreased partner anonymity are less responsive to induced pride. Chapter 6 combines laboratory data on social preferences and field data on clinicians' workplace effort. This study is unique in that we observe the same subjects from the laboratory in a field setting, where pro-social behavior has large welfare impacts. We use modified dictator games to define subjects as fair types, social information responsive types and pride responsive types and test how those characteristics are correlated with effort in the workplace. We find that clinicians responsive to both pride and social information provide higher than average effort in the workplace. These results are suggestive of Ellingsen and Johannesson's (2008) theory of social preferences wherein social identity and esteem interact to motivate altruism.