Research Papers
Unlucky college undergraduates entering the labor market in a recession suffer a persistent loss in their earnings in the medium- to long-term. Due to this ``scarring effect," the opportunity cost for graduate school attendance decreases when an individual is exposed to a recession. This paper examines whether staying in school can help the unlucky cohort in terms of future labor market outcomes. There are two channels: delaying the time to enter the labor force and accumulating human capital. Using data from the National Survey of College Graduates (NSCG), I find that graduating during a recession will increase the probability of pursuing a graduate degree by 4 percentage points, and the return for the induced graduate degree is about 23% in annual salary. At the same time, there is no statistically significant effect on the employment probability for those graduate degree holders induced by the recession. These findings provide evidence that the benefit those induced gain to graduate degree holders is from the additional accumulated human capital; the effect of delayed entering into the labor market is negligible. I also find younger non-white females in non-STEM majors from non-research universities are more sensitive to the recession when making their graduate school decision.
Gender Wage Differentials in China from 1995 to 2018: Distributional Evidence Accounting for Employment Composition using Partial Identification (with Xintong Wang and Alfonso Flores-Lagunes ) (New Draft Coming Soon)
This paper aims to examine changes in the gender wage distribution differentials in China from 1995 to 2018. We use data from the China Household Income Survey (CHIP) 1995-2013 and the China Family Panel Studies (CFPS) 2014 and 2018. To effectively account for changes in employment, we employ nonparametric bounds. To also account for the labor supply's intensive margin, we compute workers' working hours and hourly wage using available information in CHIP and CFPS. Our methodology adopts a weak median dominance assumption and a monotone instrumental variable to tighten the bounds. We find statistically significant evidence that over the years from 1995 to 2018, the median gender wage gap for the young workers (age 25-45) who are non-college-educated has increased by 0.18 - 0.62 log points. By splitting the study period, in the period between 1995-2007, we document a significant increase in the median gender wage differentials among both the young workers who are college-educated and not college-educated. When looking at the 25th wage percentile, the bounds estimates also show a significant increase in the gender wage gap for the young college-educated in 1995-2007, by 0.09 - 0.32 log points. When looking at the 75th wage percentile, the bounds suggest a significant gender gap increase for the college-educated and the young non-college-educated, by 0.16 - 0.40 log point. For the 2007-2018 period, we do not find evidence that the gender wage gap has increased further, and our bounds for the young college-educated suggest weak evidence of a decrease of the gender wage gap by -0.06 to -0.16 log points at the median and by -0.03 to -0.17 at the 75th percentile. These results suggest that the narrowing of the gender wage gap might be potentially larger than what was previously documented by Song (2019).
Gender Wage Gap in Urban China: Change and Decomposition (New Draft Coming Soon)