Low-skill job accessibility: Earnings, employment and the urban poor

My dissertation models and tests a theory referred to as either the accessibility hypothesis or the spatial mismatch hypothesis. The hypothesis is one of several theories that attempt to explain depressed earnings and elevated joblessness among low-skilled inner city residents of major metropolitan areas in the United States. The theory addresses urban poverty in metropolitan areas possessing three structural properties: low-skill workers are highly centralized; capital, particularly manufacturing capital, has experienced decentralization over the past few decades; and public transportation is poorly suited for outward commuting. In essence, the hypothesis asserts that a spatial mismatch between low-skilled workers and low-skilled employment effectively shuts off the urban poor from suburban employment.;My dissertation develops two general equilibrium land use models for a system of cities possessing the structural properties asserted by the accessibility hypothesis. The first model, the equilibrium labor market model, is used to examine the impacts of job access factors and proposed policy prescriptions on the utility and wages of centralized, low-skill workers. The second model, the disequilibrium model, similarly examines the impacts of spatial mismatch on employment opportunities in central cities. The models generate clear theoretical predictions about spatial mismatch and its effects on earnings and employment, which I test empirically.;The equilibrium model is tested by estimating a reduced form equation for wages which specifies how equilibrium wages relate to the exogenous determinants of labor demand and supply in the urban labor market. Using 1980 Census data for the fifty largest metropolitan areas in the United States, ordinary least squares estimation provided little or no support for the detrimental effects of spatial mismatch on the earnings of low-skill, centralized, working males.;The disequilibrium model is tested by estimating a linear probability model that predicts the likelihood of employment for low-skill, centralized males based on their human capital and job access measures. Ordinary least squares estimation provided relatively strong evidence in support of the negative effects of all three structural components of accessibility on employment probability.