The recovery after the great recession has renewed interest in the functioning of labor markets in the US. For example, the 2016 Economic Report of the President mentions ’jobs’ 100 times. The 2011 and 2006 reports mention ’jobs’ only 86 and 17 times respectively. Major financial media outlets now hold monthly contests urging viewers to ’guess the number’ prior to the release of BLS non-farm payrolls data. Further, labor income has become a widely discussed issue. Given the increased focus on employment, any model used for policy analysis needs to carefully consider the assumptions underlying the labor market.
All labor market models face tradeoffs between accurate prediction, plausible assumptions, and complexity. When applying these models for policy analysis, one needs to carefully consider the implications of the assumptions underlying the labor market. For example, a model that incorporates a textbook search model of unemployment will not generate sufficient fluctuations in unemployment. Some large scale DSGE models make predictions about hours worked, however they are unable to make predictions about unemployment. Few models consider the asymmetry of unemployment over the business cycle. In the US, unemployment rises further above trend during recessions than it falls below trend during expansions. Models producing symmetric data systematically under-predict the depth of recessions for shocks of a given size.
This dissertation fills a gap in the literature by emphasizing the importance of higher order moments in labor markets. The current literature largely ignores the importance of higher order moments in labor markets. I consider these moments in three ways. First, I use the observed higher order moments of the US data as an empirical target. Second, I use models with heterogeneous agents. In Chapter 2, endogenous variables vary by wealth. Therefore the higher order moments of the wealth distribution are important in determining aggregate variables. In Chapter 3, identical agents may receive different wages. The endogenous distribution of wages determines the dynamics of the aggregate wage. Finally I use solution methods that preserve higher order moments of simulated data. Standard practice is to use log linearization, however this approximation loses any higher order characteristics of the model in question. In Chapter 2, I develop a nonlinear solution method that preserves the higher order moments of simulated data.
In Chapter 2 I consider the relationship between inequality and income. This relationship has been the subject of much research since Kuznets  hypothesized that inequality first rises with income and then falls. This relationship was originally the result of migration between agricultural and industrial sectors of the economy; however this explanation has since been discredited. This chapter adds to the literature by developing a new mechanism linking inequality and income. The mechanism works via labor markets. The incentive for households to supply labor varies with the initial level of wealth. Therefore, the second moment of the wealth distribution becomes important for determining aggregate variables. The model implies that there is a non-monotone relationship between income and inequality. However, it is rotated relative to the Kuznets curve. Finally, I find empirical support by examining cross country data to evaluate this new mechanism relative to other solutions proposed in the literature.
I propose solutions to several issues regarding the predictions of frictional labor market models in Chapter 3. I develop a new wage setting mechanism that generates realistic moments of the aggregate wage. Secondly, I show that an alternate calibration of the model generates realistic volatility in the job finding rate as well as unemployment skewness. Finally, I link unemployment asymmetry to investment asymmetry though a household savings problem. This allows me to produce realistic second and third moments of both unemployment, the average wage, and investment. This realistic replication of the labor market is important because unemployment and the average wage are important components of aggregate household income. Therefore, producing realistic moments and co-movements of unemployment and the average wage is crucial to generating accurate dynamics in household income.
Finally I document a new empirical fact regarding recent changes in the cyclicality of the real wage in the U.S. I show that the wage has changed from procyclical to countercyclical in recent years. There are currently no papers that acknowledge or address this change. I hypothesize that this change may be due to rising wage inequality. If low wage jobs are more sensitive to the business cycle, then rising wage inequality strengthens compositional effects on the average wage. Using a structural VAR approach, I construct a counterfactual in which inequality is no longer rising and is not subject to further shocks. In this counterfactual, the average real wage remains procyclical. This indicates that rising wage inequality is able to account for a large portion of the change in real wage cyclicality.
|Commitee:||Driskill, Robert, Huang, Kevin, Huffman, Gregory, Oh, Hyunseung, Palacios, Miguel|
|School Location:||United States -- Tennessee|
|Source:||DAI-A 80/11(E), Dissertation Abstracts International|
|Keywords:||Business cycle asymmetry, Inequality, Unemployment volatility|
Copyright in each Dissertation and Thesis is retained by the author. All Rights Reserved
The supplemental file or files you are about to download were provided to ProQuest by the author as part of a
dissertation or thesis. The supplemental files are provided "AS IS" without warranty. ProQuest is not responsible for the
content, format or impact on the supplemental file(s) on our system. in some cases, the file type may be unknown or
may be a .exe file. We recommend caution as you open such files.
Copyright of the original materials contained in the supplemental file is retained by the author and your access to the
supplemental files is subject to the ProQuest Terms and Conditions of use.
Depending on the size of the file(s) you are downloading, the system may take some time to download them. Please be