Access to jobs : transportation barriers faced by low-skilled autoless workers in U.S. metropolitan areas

Abstract

One of the major concerns in today's urban labor market is spatial mismatch, the geographic separation between jobs and workers. Although numerous studies examine spatial mismatch, most of them focus on inner-city minorities, and the spatial mismatch problem for all autoless workers in a metropolitan area as a whole has not been well explored. Focusing on low-skilled workers and welfare recipients, this dissertation explores and quantifies the importance of job accessibility in employment outcomes for disadvantaged workers without autos in U.S. metropolitan areas. Metropolitan areas studied are Boston, San Francisco, and Los Angeles for low-skilled workers and Los Angeles for welfare recipients. An essential component of the analysis is the calculation of improved job-access measures that take into account supply and demand sides of the labor market and travel modes. The resulting measures indicate that, contrary to the perception of many spatial mismatch studies, central-city areas still offer more of a geographical advantage in accessing employment opportunities than suburban areas, despite the substantial suburbanization of employment. In other words, spatial mismatch is greater in suburban areas than in central-city areas. The measures also indicate that the levels of spatially accessible job opportunities are considerably lower for transit users than for auto users. In other words, spatial mismatch is much greater for transit users than for auto users. This transit/auto disparity is much greater than the central-city/suburb disparity, suggesting that the mode of travel has greater importance in determining job accessibility than location.

(cont.) These findings suggest that spatial mismatch may pose a serious problem for autoless workers, particularly for those who live in suburban areas, although it may not be a problem for workers with autos. By incorporating the improved job-access measures into multinomial logit (MNL) models and regression models with Heckman correction, I find that improving job accessibility for transit users significantly augments the employment probability and the probability of working full-time for low-skilled autoless workers in San Francisco and Los Angeles. Further, in all three areas the job-access effect is greater for low-skilled autoless workers than for low-skilled auto-owning workers. Applying the same analytical framework for welfare recipients in Los Angeles, I find consistent results. I also find that job accessibility for transit users plays a more important role in employment outcomes in San Francisco and Los Angeles, more highly auto-dependent areas, than in Boston, a more compact area with relatively well-developed transit systems. The empirical findings together suggest that spatial mismatch is in fact the problem for autoless workers in suburban areas where jobs are dispersed and public transportation is poorly developed. The findings also suggest that spatial mismatch is more likely to be an employment barrier for those who live in suburban areas than for those who live in central-city areas, which contradicts the dominant view among spatial mismatch researchers. The empirical findings hold important policy implications...