Essays on aggregate labor supply.

Thesis (Ph. D.)--University of Rochester. Department of Economics, 2015.The theme of this thesis is to measure the aggregate labor supply elasticity both at the intensive and extensive margins. The first two chapters concern measuring the labor supply elasticity at the extensive margin in a manner robust to model specifications. The third chapter obtains an intensive margin elasticity of labor supply in an environment in which workers' hours are complements in production. The first chapter exploits micro data on the joint distribution of consumption and wages to measure the Frisch labor supply elasticity at the extensive margin. I derive the following reservation property of the working decision: in a class of models in which the wage process is exogenous (EWP models), given consumption, there exists a unique wage level above which individuals work and below which they do not. In particular, this property is robust to arbitrary heterogeneity in borrowing constraints, discount factors, and wage processes--intuitively, consumption summarizes these factors that affect individual labor supply. Those workers with low wages relative to consumption are inferred to be more marginally attached to the labor market. The number of such workers is key to the magnitude of the Frisch elasticity at the extensive margin. Using the joint distribution of consumption and wages observed from the PSID waves 1999-2011, I find that (i) the aggregate Frisch elasticity of labor supply at the extensive margin is 0.4, and that (ii) across various demographic groups, the elasticity ranges from 0.2 to 0.6. These estimates are similar to those of quasi-experimental studies, suggesting that the number of marginal workers implied by the data is relatively small. In the second chapter, I allow the wage process to be endogenous by writing a class of models in which individuals accumulate human capital through learning-by-doing (LBD). I again measure the labor supply elasticity at the extensive margin, but consider how the human capital accumulation affects the measured elasticity compared to the simpler environment in Chapter 1. I show that in this environment the reservation wage can be defined conditional on consumption and assets choices. Intuitively, if a worker with the same wage and assets with another individual consumes more, then this suggests that the worker has a higher shadow value of LBD. Thus, consumption and assets choices jointly reveal the willingness to work, or the reservation wage. Using the data of consumption, wages, and assets from the PSID waves 1999-2011, I find that the aggregate labor supply elasticity at the extensive margin under the human capital models is 0.36, while that under the EWP models is 0.4. The small elasticity gap is because individuals with low consumption are likely to have low assets as well, implying that understanding the relationship between consumption and wages remains key to predicting the employment responses to wage shocks. Second, for narrowly defined demographic groups, the measured elasticities range from 0.2 to 1. As with the EWP models, relatively elastic groups are those who are younger, single, nonwhite, female, or without college degree. Considering the human capital accumulation does not particularly change the demographic characteristics of more marginal workers. The third chapter is based on a paper coauthored with Michele Battisti of Ifo Institute, and Ryan Michaels of the Department of Economics at the University of Rochester. We study the labor supply elasticity at the intensive margin in an environment in which workers are complements in production. The complementarity of workers implies an incentive to coordinate labor supply within the firm, which compresses working-time adjustments across workers in response to purely idiosyncratic variation in their return from working. This places no restrictions, however, on the response of firm-wide working time to firm-wide shocks. We estimate a model in which heterogeneous firms and workers bargain on working time and earnings using the method of simulated moments. The target moments are from matched firm-worker data from North-East Italy. We revisit earlier findings of a small intertemporal elasticity of labor supply exploiting the model's prediction that this elasticity will be larger for firm-wide fluctuations than evaluated at the individual level. First, the model uncovers the Frisch labor supply elasticity at the intensive margin 0.53. This value is near the top end of the range of estimates found in earlier studies. Second, to study how ignoring the coordination of labor supply affects the implied elasticity, we simulate the model such that only 1/9 of a firm's workforce receives a lump-sum transfer, but the remainder of the firm's workers do not (The fraction of the workforce corresponds to one cohort of workers that shares the same productivity and preference in the model). If we use the treatment effect in this case to infer the Frisch elasticity, the implied elasticity is less than half the estimate 0.53 we uncover