Working Papers

Juvenile Crime and Anticipated Punishment

Recent research suggests that the threat of harsh sanctions does not deter juvenile crime. This is based on the finding that criminal behavior reduces only marginally as individuals cross the age of criminal majority, the age at which they are transferred from the juvenile to the more punitive adult criminal justice system. Using a model of criminal capital accumulation, I show theoretically that these small reactions close to the age threshold mask sizable reactions away from, or in anticipation of, the age threshold. The key prediction of this framework is that when the age of criminal majority is raised from seventeen to eighteen, all individuals under eighteen will increase criminal activity, not just seventeen year olds. I exploit recent changes to the age of criminal majority in the United States to show evidence consistent with this prediction - arrests of 13-16 year olds rise significantly for offenses associated with street gangs, including homicide, robbery, theft, burglary and vandalism offenses. Consistent with previous work, I find that arrests of 17 year olds do not rise systematically in response. I provide suggestive evidence that this null effect may be due to a simultaneous increase in under-reporting of crime by 17 year olds.

Elections by Consensus: Effects on Politician Identity & Government Size

This paper examines the effects of election via community consensus on politician identity and government performance. Identification comes from a unique policy experiment in rural India that incentivizes village residents to reach a consensus on who their political representatives should be, eliminating the need for the state government to organize official elections. I find that this policy crowds in younger, more educated candidates into political office. This move towards inexperienced politicians exerts a negative influence on government size, measured by the collection of tax revenue, government expenditure and employment creation by the local government.


The Career Impact of First Jobs: Evidence and Labor Market Design Lessons from Randomized Choice Sets (with Jonas Hjort)