Criminal Convictions and Labor Market Outcomes in Young Men : A Matching Analysis with U.S. Data
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
This paper investigates the causal effect of criminal convictions on earnings in young men using data from the 2006 wave of the 1997 National Longitudinal Survey of Youth (NLSY 1997). The treatment effect of interest (the average treatment effect on the treated) is estimated using a propensity score matching algorithm. Matching is done with both a nearest neighbor and a nearest K neighbor matching technique. The main assumption is that of conditional independence of potential outcomes and treatment and is ensured by controlling for a large array of personal and family characteristics. Both matching algorithms lead to an estimated ATT of about -$2,200 which is markedly lower than estimates from standard regressions which yield an effect of -$3,300. These results align well with previous findings investigating older cohorts and is in contrast to anticipated declines in the effect given the surge in the correctional populations. An increase in the discriminated group should have acted to reduce the penalty and the lack of a decline suggests perhaps the proportion of discriminating firms has also increased which might reflect a social shift in the perception of having a criminal record. An additional analysis with matching preprocessing and subclassified regression suggest there is heterogeneity in the effect with those with low probability of convictions having effects in the range of -$4,000 to -$4,500 while those with high probability of convictions having positive (insignificant) effects.
Abstract
This paper investigates the causal effect of criminal convictions on earnings in young men using data from the 2006 wave of the 1997 National Longitudinal Survey of Youth (NLSY 1997). The treatment effect of interest (the average treatment effect on the treated) is estimated using a propensity score matching algorithm. Matching is done with both a nearest neighbor and a nearest K neighbor matching technique. The main assumption is that of conditional independence of potential outcomes and treatment and is ensured by controlling for a large array of personal and family characteristics. Both matching algorithms lead to an estimated ATT of about -$2,200 which is markedly lower than estimates from standard regressions which yield an effect of -$3,300. These results align well with previous findings investigating older cohorts and is in contrast to anticipated declines in the effect given the surge in the correctional populations. An increase in the discriminated group should have acted to reduce the penalty and the lack of a decline suggests perhaps the proportion of discriminating firms has also increased which might reflect a social shift in the perception of having a criminal record. An additional analysis with matching preprocessing and subclassified regression suggest there is heterogeneity in the effect with those with low probability of convictions having effects in the range of -$4,000 to -$4,500 while those with high probability of convictions having positive (insignificant) effects.
참고문헌 (18)
- Abadie, Alberto, "Large Sample Properties of Matching Estimators for Average Treatment Effects", Econometrica, 74(1), pp.235~267 (2006)
- Abadie, Alberto, "On the Failure of the Bootstrap for Matching Estimators", Econometrica, 76(6), pp.1537~1557 (2008)
- Abadie, Alberto, "Bias-Corrected Matching Estimators for Average Treatment Effects", Journal of Business and Economic Statistics, 29, pp.1~11 (2011)
- Allgood, Sam, "The Impact of Youth Criminal Behavior on Adult Earnings" (2006)
- Black, Dan, "Discrimination in an Equilibrium Search Model", Journal of Labor Economics, 13, pp.309~334 (1995)
- Black, Dan, "How Robust is the Evidence on the Effects of College Quality? Evidence from Matching", Journal of Econometrics, 121, pp.99~124 (2004)
- Crime and the Employment of Disadvantaged Youths, "Freeman, Richard", National Bureau of Economic Research, ~, pp.Working Paper No.3875 (1991)
- Grogger, Jeffrey, "The Effect of Arrests on the Employment and Earnings of Young Men", Quarterly Journal of Economics, 110, pp.51~71 (1995)
- Harder, Valerie, "Marijuana Use and Depression Among Adults: Testing for Causal Associations", Addiction, 101, pp.1463~1472 (2006)
- Collateral Costs: The Effects of Incarceration on the Employment and Earnings of Young Workers, "Holzer, Harry J.", IZA, 3118~, pp.Discussion Paper No. 3118 (2007)
- Jones, Alison Snow, "Causal Effects of Alcoholism on Earnings: Estimates from the NLSY", Health Economics, 15, pp.849~871 (2006)
- Kling, Jeffrey R., "Incarceration Length, Employment, and Earnings", American Economic Review, 96, pp.863~876 (2006)
- Politis, Dimitris, "Large Sample Confidence Regions Based on Subsamples under Minimal Assumptions", The Annals of Statistics, 22(4), pp.1635~2134 (1994)
- Richey, Jeremiah, "The Causal Effects of Criminal Convictions on Labor Market Outcomes in Young Men: A Nonparametric Bounds Analysis" (2013)
- Rosenbaum, P. R., "The Central Role of the Propensity Score in Observational Studies for Causal Effects", Biometrika, 70(1), pp.S.1~S.55 (1983)
- U.S. Department of Justice, "Office of Justice Programs. Bureau of Justice Statistics. Bulletin: Correctional Population in the United States, 2009", U.S. Government Printing Office, ~ (2010)
- Western, Bruce, "The Labor Market Consequences on Incarceration", Crime & Delinquency, 47, pp.410~426 (2001)
- Waldfogel, Joel, "The Effect of Criminal Conviction on Income and the Trust “Reposed in the Workmen", he Journal of Human Resources, 29, pp.62~81 (1994)