Research

Working Papers

Life-cycle Bias and Intergenerational Associations in Crime 

The lack of complete lifetime data causes bias in intergenerational associations estimates due to variations in the short-run proxy throughout the life cycle. We develop a new methodology based on the generalized errors-in-variables (GEiV) model to correct the life-cycle bias and estimate the intergenerational associations in crime. Our results suggest that intergenerational associations are underestimated by 30%–50% even when the criminal behaviors of both generations are observed in teenage years and have the smallest life-cycle bias. The GEiV-corrected intergenerational elasticity in New Zealand between fathers and sons is around 0.7 for the likelihood of crimes and 0.5 for the number of crimes. Mother–child elasticities are generally smaller than father–child elasticities. The estimated elasticities are stable across ages, birth cohorts, and types or severity of crimes. Our GEiV correction performs well and can be applied to any outcomes with a strong age profile for both the extensive and intensive margins.


Motherhood Employment Penalty and Gender Wage Gap Across Countries: 1990–2010 (with Harold Cuffe and Nguyen Doan)

In this paper, we use twin birth as an instrument to estimate the effects of fertility on female labor force participation using 70 censuses from 36 countries in 1990–2010. We document a strong relationship between the gender wage gap and the size of the motherhood penalty. The penalty is smallest in countries with small gender wage gaps. Both cross- and within-country relationships between motherhood penalty and gender wage gap remain strong and negative even when we condition on per-capita GDP and educational attainment. Our estimates suggest that a reduction of 1-percentage-point in the gender wage gap is associated with a decrease of 0.45–0.65 percentage-points in the estimated motherhood employment penalty.


Intergenerational Mobility

The Apple Doesn't Fall Far from the Tree: Intergenerational Wealth Mobility in Taiwan (with Ming-Jen Lin and Huici Nian)

Labour Economics, forthcoming.

We estimate intergenerational wealth mobility using administrative records from Taiwan and find strong intergenerational rank correlations in household gross wealth, around 0.40 for sons and 0.30 for daughters. The wealth rank correlations are similar for single and married children, even though married children have greater household wealth and relatively strong assortative mating. The intergenerational wealth correlation is nonlinear and particularly strong for families in the top 10 % of household wealth. Different family sizes and gender compositions do not significantly affect intergenerational wealth mobility, regardless of the children's gender or marital status. Since our sample includes only parents who are still alive, these correlations are not driven by bequests. However, inter vivos transfers appear to play a significant role.


Intergenerational Earnings Mobility in Taiwan: 1990-2010 (with Ming-Jen Lin)

Empirical Economics, 59(1), July 2020, pp. 11-45. Working paper version

In this paper, we study the intergenerational earnings mobility between fathers and sons in Taiwan. We apply the two-sample approach developed by Björklund and Jäntti (1997) and find that the intergenerational earnings elasticity in Taiwan was around 0.4–0.5 in both the early 1990s and the late 2000s. We also estimate the intergenerational rank association in earnings to have been around 0.3 in both periods. Intergenerational earnings mobility in Taiwan is similar to that in less mobile countries such as the U.K. and the U.S., and it appears to remain stable during a period of rapid economic development. 


Student Loan

Do Academically Struggling Students Benefit From Continued Student Loan Access? Evidence From University and Beyond (with Harold Cuffe) 

Review of Economics and Statistics, 106 (1), January 2024, pp.68-84. Working paepr version

We estimate the effects of access to student loans on university students’ educational attainment and labor market returns in New Zealand. We exploit the introduction of a policy mandating a minimum pass rate of 50% for student loan renewals using a fuzzy regression discontinuity design and linked administrative records. For students around the threshold, retaining access to student loans increases their likelihood of re-enrollment and bachelor’s degree completion rate. The effects are observed primarily among female students due to a substantial gender difference in compliance with the pass rate criterion. We find that retaining student loans leads to large labor market returns for struggling female students. The additional student loan debt from further borrowing declines quickly due to faster repayment.

 

Medical Marijuana

Joint Culpability: The Effects of Medical Marijuana Laws on Crime (with Wilbur Townsend)

Journal of Economic Behavior and Organization, 159, March 2019, pp. 502-525. Working paper version

Most of the U.S. states have passed medical marijuana laws. In this paper, we study the effects of these laws on violent and property crime. We first estimate models that control for city fixed effects and flexible city-specific time trends. To supplement this regression analysis we use the synthetic control method which can relax the parallel trend assumption and better account for heterogeneous policy effects. Both the regression analysis and the synthetic control method suggest no causal effects of medical marijuana laws on violent or property crime at the national level. We also find no strong effects within individual states, except for in California where the medical marijuana law reduced both violent and property crime by 20%.

Brief summary at The Conversation (reprinted in Newsweek, Foundation of Economics Education, Business Insider)


High Times: The Effect of Medical Marijuana Laws on Student Time Use (with Seth Gershenson)

Economics of Education Review, 66, October 2018, pp. 142-153 . Working paper version

Medical marijuana laws (MMLs) represent a major change of marijuana policy in the U.S. Previous research shows that these laws increase marijuana use among adults. In this paper, we estimate the effects of MMLs on secondary and post-secondary students’ time use using data from the American Time Use Survey. We apply a difference-in-differences research design and estimate flexible fixed effects models that condition on state fixed effects and statespecific time trends. We find no effect of MMLs on secondary students’ time use. However, we find that college students in MML states spend approximately 20% less time on education-related activities and 20% more time on leisure activities than their counterparts in non-MML states. These behavioral responses largely occur during weekends and summer when students have more spare time. Finally, the impacts of MMLs are heterogeneous and stronger among part-time college students, who are more likely to be first-generation college goers and to come from underrepresented racial and ethnic groups.


What We Still Need to Know About the Impacts of Medical Marijuana Laws in the United States? Commentary on Sarvet et al. (2018)

Addiction, 113(6), June 2018, pp. 1017-1018.  

Past-month marijuana use among adolescents does not increase after the passage of medical marijuana laws in the U.S. It is crucial for future research to explore causal mechanisms affecting different types of marijuana users to bring a deeper understanding of behavioral responses to marijuana policies.


Do Medical Marijuana Laws Increase Hard Drug Use?

 The Journal of Law and Economics, 58(2), May 2015, pp. 481-517. Working paper version

Medical marijuana laws generate significant debates regarding drug policy. For one thing, if marijuana is a complement to hard drugs, then these laws would increase the usage not only of marijuana but also of hard drugs. In this paper I study empirically the effects of medical marijuana laws by analyzing data on drug arrests and treatment admissions. I find that medical marijuana laws increase these proxies for marijuana consumption by around 10–15%. However, there is no evidence that cocaine and heroin usage increases. From the arrest data, the estimates indicate a 0–15% decrease in possession arrests for cocaine and heroin combined. From the treatment data, the estimates show a 20% decrease in admissions for heroin treatment, although there is no significant effect for cocaine treatment. These results suggest that marijuana may be a substitute for heroin, but it does not have a strong relationship with cocaine.


The Effects of Medical Marijuana Laws on Illegal Marijuana Use

Journal of Health Economics, 38, December 2014, pp. 43–61. Working paper version

More and more states have passed laws that allow individuals to use marijuana for medical purposes. There is an ongoing, heated policy debate over whether these laws have increased marijuana use among non-patients. In this paper, I address that question empirically by studying marijuana possession arrests in cities from 1988 to 2008. I estimate fixed effects models with city-specific time trends that can condition on unobserved heterogeneities across cities in both their levels and trends. I find that these laws increase marijuana arrests among adult males by about 15–20%. These results are further validated by findings from data on treatment admissions to rehabilitation facilities: marijuana treatments among adult males increased by 10–20% after the passage of medical marijuana laws. 


Non-cognitive Traits

Gender Gap in Upward Mobility: What is the Role of Non-cognitive Traits?  (with Susan J. Linz)

International Journal of Manpower, 38(6), 2017, pp. 835-853. Working paper version

Do non-cognitive traits contribute to the gender gap in supervisory status and promotion? We use a large linked employer-employee dataset collected from six former socialist countries to assess the link between non-cognitive traits and upward mobility. Controlling for workplace heterogeneity, we find that gender differences in locus of control, the preference for challenge versus affiliation, and adherence to work ethic together can explain about 7–18% of the gender gap in supervisory status and promotions Overall, non-cognitive traits provide an important, though incomplete, explanation for the gender gap in upward mobility.


Work Ethic in Formerly Socialist Economies (with Susan J. Linz)

Journal of Economic Psychology, 39, December 2013, pp. 185–203.

Do younger workers in transition economies have a different work ethic from those who were trained and employed in the former socialist economy? Is there a positive link between work ethic and earnings among workers in transition economies? We address these questions using data collected from employee surveys conducted in Armenia, Azerbaijan, Kazakhstan, Kyrgyzstan, Russia, and Serbia. Employing a composite measure, we find that younger workers tend to adhere more strongly, and older workers less strongly. This result is obtained in the majority of cases for the individual work ethic components, as well. We also find work ethic adherence is stronger among men than women, among supervisors, and among participants who exhibit an internal locus of control. The link between work ethic and earnings is positive: participants who scored highest on the work ethic measure earn 15% more than those who scored lowest. Commonalities across these six culturally and economically diverse countries provide a foundation for developing a more global perspective of work ethic and worker performance.


Weber, Marx, and work values: Evidence from transition economies (with Susan J. Linz)

Economics System, 37 (3), September 2013, pp. 431-448.

Are work values a cause (Weber) or consequence (Marx) of the economic environment? The collapse of the Soviet Union at the end of 1991 provides a unique opportunity to investigate this link. Using data collected from an employee survey conducted in over 340 workplaces in Armenia, Azerbaijan and Russia, we investigate generational differences in adherence to the Protestant work ethic (PWE). Our results indicate that Marx was ‘right’ about the link between work values and economic environment. That is, despite economic and cultural differences emerging during the transformation process, in all three countries, participating workers born after 1981 adhere more strongly to PWE than workers born before 1977. Moreover, the estimate magnitudes are very similar across these economically and culturally diverse countries. More generally, PWE adherence is stronger among participating workers with an internal locus of control and among supervisors. PWE adherence also tends to be stronger among participants with high relative earnings, as well as among those working in organizations that reward hard work with the chance to develop new skills or learn new things.