Effects of income, education, and time preference on youth crime
- This thesis explores whether impatience is a significant predictor of criminal activity. I develop a simple lifecycle model that links income level and educational attainment to criminal propensity, then evaluate its predictions using the National Longitudinal Youth Survey of 1997 (NLSY97). In this model, an individual's discount parameter may distort his ability to weigh the potential benefits and costs of committing a crime. Estimations of the effect the discount parameter has on propensity to commit crime typically rely on proxies for impatience which involve either directly surveying an individual's self-assessed impatience, or imputing an individual's impatience through his responses to a series of hypothetical financial choices. I propose using a method which relies entirely on observed decisions -- specifically, a semi-parametric structural approach proposed by Fang and Wang (2015) for the estimation of discount and impatience parameters. With the goal of estimating impatience in mind, I examine two other important predictors of criminal propensity in the NLSY97: annual current income and educational attainment in the form of high-school graduation. I focus on these two variables both because of their strong correlation with crime in the NLSY97 dataset, as well as their link to time preferences. Annual income provides a simple metric for evaluating cost-benefit analyses across time periods, and the positive correlation between education and patience has long been recognized -- in particular, there is growing evidence that education causally increases the ability to delay gratification. Using pooled and fixed-effect reduced form regressions, I observe a strong negative correlation both between current annual income level and crime, and between high-school graduation and crime. To elucidate the nature of this negative relationship, I use a structural estimation proposed by Bajari et al. (2012) to evaluate the utility functions of my model and compare instantaneous and long-term criminal incentives across income groups. Curiously, I find no variation in these incentives. This suggests that the negative correlation between income and crime in the NLSY97 is primarily driven by current utility. Unfortunately, this lack of variation does not allow me to estimate discount parameters using the methodology proposed. As a result, I cannot rule out the possibility that individuals who do and do not commit crimes are equally myopic. This unexpected lack of variation raises the question of whether criminal propensity is driven primarily by early ingrained behaviours, rather than by differing incentives across income groups. I attempt to answer this through further reduced form analysis. This analysis reveals that, controlling for individual traits, criminal involvement before the age of 15 is the strongest predictor of crime among regressors in the NLSY97 dataset, suggestive of path-dependence for young criminals. Combined with evidence of successful adolescent intervention programs that reduce future criminal propensity by targetting impulsive behaviour (Heller et. al 2015), this reinforces the hypothesis that adolescence is a crucial period for influencing crime-relevant characteristics -- in particular, impulsivity. Though I am unable to test this hypothesis in the NLSY97, empirical work on impatience and crime is currently limited, and structural estimation of criminal time preferences is an important avenue of future research.
|Type of resource
|electronic; electronic resource; remote
|1 online resource.
|Dagenais, Genevieve Erin Elan
|Stanford University, Department of Economics.
|Polinsky, A. Mitchell
|Polinsky, A. Mitchell
|Statement of responsibility
|Genevieve Erin Elan Dagenais.
|Submitted to the Department of Economics.
|Thesis (Ph.D.)--Stanford University, 2017.
- © 2017 by Genevieve Erin Elan Dagenais
- This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).
Also listed in
Loading usage metrics...