Essays on Labor and Behavioral Economics
Martin, David Chase
MetadataShow full item record
CitationMartin, David Chase. 2020. Essays on Labor and Behavioral Economics. Doctoral dissertation, Harvard University, Graduate School of Arts & Sciences.
AbstractThis dissertation combines two streams of research. The first uses quasi-experimental techniques to address questions of production and labor supply in public-sector markets. The second uses experimental methods to assess the effect of monetary incentives on classic decision-making biases.
In Chapter 1 (joint with Ashley Craig), I look at the effect of relaxing strict school discipline policies on student outcomes. We evaluate a reform in New York City that eliminated suspensions for non-violent, disorderly behavior using a difference-in-differences framework and natural variation in the reform's impact. In schools that were more affected, we find modest, but meaningful, achievement gains that are linked to improvements in school culture as a result of the discipline reform. These improvements benefited students even if they were unlikely to be suspended themselves.
In Chapter 2, I measure the effect of mayoral partisanship on labor market outcomes for fire fighters, police officers, and teachers. Average earnings in these occupations are highly correlated with both Democratic vote share and union strength at the state level. Using two complementary methodologies, I test whether partisan differences in collective bargaining at the local level can explain these patterns. I provide suggestive evidence that mayoral partisanship affects labor market outcomes for fire fighters, police officers, and teachers. But it is not the case that these workers do consistently better under Democratic mayors.
In Chapter 3 (joint with Benjamin Enke, Uri Gneezy, Brian Hall, Vadim Nelidov, Theo Offerman, and Jeroen van de Ven), I study the effect of large incentives on four widely documented cognitive biases: base rate neglect, anchoring, failure of contingent thinking, and intuitive reasoning. In pre-registered laboratory experiments with 1,236 college students in Nairobi, Kenya, we implement three incentive levels: no incentives, standard lab payments, and very high incentives that increase the stakes by a factor of 100 to more than one month of average income. Although cognitive effort as measured by response times increases by 40% with very high stakes, performance improves only mildly or not at all. These results contrast with expert predictions that forecast larger performance improvements.
Citable link to this pagehttps://nrs.harvard.edu/URN-3:HUL.INSTREPOS:37365909
- FAS Theses and Dissertations