Department of Economics
This paper evaluates the long-term welfare consequences from experiencing a recession as youths, considering the impact on schooling, wages and labor supply. The paper also explores the mechanisms that account for lifetime wage changes, including changes from schooling, from work experience, and from job mobility. I develop and estimate a directed search equilibrium model with heterogeneous agents and aggregate shocks. Experiencing the 1981-1982 recession at age 16-22 causes a 2.2% to 3.0% lifetime welfare loss. Endogenizing schooling decision avoids overestimation of the welfare loss. The loss in experience and tenure explains the persistent wage loss after the recession.
Using data from the restricted-access National Longitudinal Survey of Youth 1997, we estimate the long-term impact of an early career recession on various health outcomes and health-related behaviors up to age 30 for males and females by education groups. The early career recession is measured by the unemployment rate of the graduation state in the year when the individual enters the labor market upon receiving the highest degree. Because the timing of labor market entry could potentially be affected by economic conditions, we instrument the unemployment rate when entering labor market using the state unemployment rates at age 18 and age 22. Our main findings are that first, an early career recession has an adverse impact on health outcomes and health-related behaviors in general, and second, these adverse effects are especially pronounced among lower educated individuals. In particular, an early career recession increases the probability of bad health status for high school graduates, but has no effect on college graduates; it also leads to more depression for high school graduates than college graduates among males. We also find that a higher unemployment rate at early career significantly increases adverse health behaviors such as smoking, heavy drinking, and illicit drug use among high school graduates, but there is no statistically significant impact on college graduates. In addition, different gender-education groups respond differently in time use, such as time spent on exercise, sleep, and watching TV. An early career recession also reduces daily fruit intake for all males and unskilled females.
We study the role of state sector employment on overall flexibility of labor market, using China as an example. Using unique matched monthly Urban Household Survey data that are representative of the Chinese urban households for the period of 2004-2006, we show that the Chinese urban labor market is characterized by very low dynamics and high long term unemployment. In particular, state-sector employments are much less dynamic compared to non-state sector. We extend the Mortensen and Pissarides equilibrium search and matching model to explain these patterns. The model incorporates two sectors, state and non-state, and considers three major differences between the two sectors: state sector has lower labor productivity, higher labor adjustment costs, and extra labor costs. The calibrated model is able to predict the simultaneous decline of state sector and increase in unemployment rate that has happened in the 1990s and early 2000s. Simulation also implies that further decline of state sector would improve labor market dynamics significantly but only have very small effect on the levels of unemployment. We also find that the extra labor cost of state-sector firms plays a major role in explaining differences in labor dynamics between state and non-state sectors.
How do adult children and their parents decide to live together? While previous studies usually model co-residence as a non-cooperative or Nash bargaining game between one parent and one child, this paper models living arrangement as a two-sided matching game that captures not only the competition between multiple children, but also the competition between parents and in-laws. This is the first application of Shapley-Shubik-Becker framework in the co-residence context. Different from the marriage matching, family network plays an important role in the co-residence matching by restricting the possible co-residence arrangements. Using data from China Family Panel Study, we develop a novel way to simulate family network to fill in the missing marriage links in the data. We then estimate the model and use the empirical results to quantify the impact of tastes, housing costs, elderly care, and child care on family living arrangement. Our model and estimations also allow us to predict the effect of various policy changes on co-residence pattern and labor supply in China.