We specify and implement a test for the presence and importance of labor market networks based on residential proximity, in determining the establishments at which people work. Using matched employer-employee data at the establishment level, we measure the importance of these network effects for groups broken out by race, ethnicity, and measures of skill. The evidence indicates that these types of labor market networks do exist and play an important role in determining the establishments where workers work; that they are more important for minorities and the less skilled, especially among Hispanics; and that they appear to be race based.

JEL No. J20,L25,L53 We use a new database, the National Establishment Time Series (NETS), to revisit the debate about the role of small businesses in job creation. Birch (e.g., 1987) argued that small firms are the most important source of job creation in the U.S. economy, but Davis et al. (1996a) argued that this conclusion was flawed, and based on improved methods and using data for the manufacturing sector they concluded that there was no relationship between establishment size and net job creation. Using the NETS data, we examine evidence for the overall economy, as well as for different sectors. The results indicate that small establishments and small firms create more jobs, on net, although the difference is much smaller than what is suggested by Birch's methods. However, the negative relationship between establishment size and job creation is much less clear for the manufacturing sector, which may explain some of the