Macroeconomic analysis has traditionally focused on accounting for regularities in highly aggregated economic time series. In the context of the labor market, for example, it is common practice to seek to account for movements in aggregate employment or unemployment. It is well known that highly aggregated data mask a great deal of economic activity; even in periods when aggregate employment changes very little, there are very large changes in the employment status of individual workers and in the distribution of employment across individual establishments. Standard macroeconomic models abstract from these changes. For some issues, this abstraction from underlying labor market flows may be innocuous, but in other contexts it is surely a serious shortcoming. The analysis of many labor market policies are clearly best carried out in aggregate models which do account for labor market flows. Indeed, there are growing empirical and theoretical literatures that focus on these underlying labor market flows, and these models have been used to begin addressing the effects of various policies. Despite the advances that have been made, however, much remains to be done. The goal of this research is to sharpen the connection between the theoretical and empirical branches of this literature so as to better understand what model features are quantitatively important in accounting for the key regularities in the data. This effort should also sharpen both empirical and theoretical issues individually, since the process of confronting theory with data will isolate those empirical relationships which have a lot of power to discriminate between model features, and hence are worthy of closer scrutiny, as well as isolate those model features which are most significant in accounting for empirical regularities and hence guide future theoretical extensions. Loosely speaking, one may identify two processes that give rise to labor market flows. One is the rearranging of employment positions across establishments, and the other is the rearrangement of workers across existing employment positions. Empirical work has shown that both of these processes are important in accounting for labor market flows. Though the former has been studied, it remains true that not that much is known of what model features are important quantitatively to match the cyclical properties of job reallocation. The second process has been subject to little analytical study. This project brings this feature into the class of matching models commonly used to address labor market flows. Labor market regulations which differ between Europe and the US are likely to distort this second process and this project will also investigate the allocational and welfare consequences of policies which restrict wage differentials across workers.
