Essays on the macroeconomics of labor markets

Abstract

This thesis is composed of three chapters that study the behavior of economies where trade in both labor and goods market is subject to search frictions. It analyzes the interactions between frictional labor and goods markets and examines technology and preference shocks as alternative sources of business cycle fluctuations in unemployment, hours worked and inventories. In Chapter 1, the focus is on the Diamond-Mortensen-Pissarides model with Nash wage bargaining. This model provides a qualitatively appealing theory of unemployment, but its ability to explain the observed magnitude of fluctuations in unemployment remains debated. I add goods market frictions to this model, and show that they affect workers' bargaining position, provide a rationale for a high value of non-market activity and also affect its cyclical properties. These frictions can thus amplify the response of unemployment and vacancies to changes in the measured labor productivity caused by either technology or preference shocks. The response of the vacancy-unemployment ratio in the extended model is about twice as large as in the model with labor search only if either (1) goods and search effort are substitutes in the goods market matching function and fluctuations are a result of a technology shocks, or (2) when goods and search effort are complements in the goods market matching function and the driving force of fluctuations are preference shocks. Finally, I show that if preferences are additively separable and goods market matching function has unit elasticity of substitution, preference and technology shocks are observationally equivalent and can not be separately identified by an economist who would analyze data on labor productivity, output, employment and wages.Chapter 2 shows that introducing goods market frictions into an otherwise standard model provides a simple but attractive framework to analyze the behavior of inventories over the business cycle. It also shows that the behavior of sales and inventories over the business cycle contains information that allows to identify the contribution of technology and preference shocks to fluctuations in unemployment. I employ Bayesian methods to estimate a model with goods and labor search frictions using U.S. data on labor productivity and inventory-sales ratio, and find that the implied response of vacancies and unemployment to changes in measured labor productivity is about twice as large as in the model with labor search only. Goods market frictions also allow the model to account for the main facts on inventories - procyclical inventory investment, countercyclical inventories-sales ratio, and sales which are more volatile than production.In Chapter 3, I examine another shortcoming of the labor search model identified by Shimer (2010) and related to the so called labor wedge - the gap between the firm's marginal product of labor and the household's marginal rate of substitution between consumption and leisure. I show that under the business cycle accounting approach proposed by Chari, Kehoe, and McGrattan (2007) goods market frictions in the model manifest themselves as a labor wedge: In an expansion, firms find it easier to sell goods, and consumers benefit from higher availability of goods and smaller disutility from search effort required per unit of consumption purchased; this encourages larger response of the intensive margin of labor supply than in the standard frictionless model. It thus also alleviates the issue arising in model with frictional labor markets, where search frictions act as adjustment costs, and thus result in a labor wedge that resembles a counterfactually procyclical tax on labor income.