Abstract: Firms use internal promotion and external hires to fill managerial positions. However, the role of the internal labor market is typically ignored in the workhorse theory of worker mobility. To quantify the aggregate effects of internal promotion on labor market efficiency, we develop a novel job search model, which gives rise to an equilibrium sorting pattern of heterogeneous workers, firms, and jobs. We calibrate the model using administrative data from Denmark. The estimates reveal that external labor market friction and general human capital are crucial for understanding firm hiring strategies. The counterfactual experiment shows that banning internal promotion decreases aggregate production by about 6.8%, although promotion corrects the overinvestment of firm-specific human capital. Moreover, the counterfactual results imply that the general and firm-specific human capital investment would be 50% lower when shutting down promotion, which suggests the importance of promotion margin in understanding why firms provide training to workers.

Abstract: We show that an improvement to Google Maps enabled by real-time transit tracking increased public transit ridership, particularly on subways. Our findings are based on panel data combining information on public transit utilization with information on Google Maps’ staggered incorporation of real-time tracking across transit agencies in the United States. Ridership per capita is 16% larger 3 years after the rollout of real-time information in Google Maps.  Agency-specific applications that use real-time information have a smaller impact on ridership and do not reduce the estimated effect of Google adopting real-time. Using independent data, we verify that city residents reduce car trips and opt for transit during their daily commutes, indicating significant external benefits. We see suggestive evidence of improved air quality and estimate that travelers are willing-to-pay around $0.70 per trip for the improvement. These results suggest that increases in the salience and convenience of public transit can have substantial impacts on welfare.

Abstract: There is a long literature on the effect of minimum wages on employment outcomes, such as wages and employment. There is less attention paid to how minimum wages impact firm performance. In this paper, I focus on the effect of minimum wage reform in 2004 on firm capital and employment decisions in China. Using Chinese manufacturing firm-level data, I use a generalized difference-in-differences design to show that minimum wage increases wage, decreases employment. The minimum wage reform has a larger effect on low-wage firms that average wage and capital rose faster that capital per worker increased 0.95 percentage point per 1 percent increase in minimum wage level. I then build and calibrate a labor search model with a capital holdup problem: when firms invest more in capital, they owe their workers higher bargaining wages. The fitted model captures key features of Chinese capital and labor markets before the minimum wage reform. The model predicts that low-wage firms will increase capital relative to high-wage firms after a minimum wage increase. Furthermore, the minimum wage reform is predicted to increase the social surplus. A counterfactual experiment implies a 10% reduction in real interest rate generates similar results on average capital and average wage level as the 2004 reform.

Abstract: We study the implications of two-dimensional learning in labor market on earnings. We build a simple model in which worker observes output and updates beliefs about the worker type and the match qualities of all the firms that she has worked for. The work experience at a firm not only helps workers to learn more about the match quality with the current firm, but also about the match qualities for the previous firms and the worker type. This creates an incentive to return to one of the previous employers and help us explain a common pattern in labor market: about 5% of workers in Italy return to a previous employer at least once in their career. We then calibrate the model using Italian matched employer-employee data from 1975-2001. The fitted model captures key features of Italian labor market and suggests that the size of uncertainty on match-specific productivity is 10 times larger than that of the worker-level ability.

Abstract: In this paper, we investigate the role of the internal labor market over the business cycle in US. We empirically document that within-firm occupational mobility moves countercyclically, while job-to-job transition moves cyclically. To quantify how much insurance the internal labor market provides to workers, we build a multi-worker directed search model in which firms are heterogeneous in terms of their number of occupations/firm size. We show that large firms provide more insurance to their workers through the internal labor market. The results suggest a more efficient unemployment insurance (UI) design contingent on firm size. This paper also provides a novel insight on how firm size affects wage volatility.