Commuting Patterns and Firm Decentralization

Abstract

Since most employed individuals must travel between home and work on a daily basis, work place commuting has provided the point of departure for the monocentric model of residential location (Alonso 1964; Mills 1972; Muth 1969). One of the major behavioral assumptions of the monocentric model is that consumers prefer less commuting to more, all else equal. Hence, the model predicts that workers will take advantage of any opportunity to reduce their commutes. In the monocentric situation, in which all employment is located at the central business district (CBD), no such opportunity arises. However, decentralized' firm location enables workers to shorten their commutes as firms have left the CBD to move into the residential area. The monocentric model thus predicts that, if firms have decentralized, workers should have very short commutes in longrun equilibrium. However, significant commuting persists in the face of large-scale firm decentralization. Such commuting leads to the belief that frictions, with respect to residential and job mobility, exist which prevent workers from achieving the long-run equilibrium solution predicted by the monocentric model. In this paper, I hypothesize that individuals with greater job and residential mobility will have shorter commutes, given decentralized firm location. Residential and job mobility are further hypothesized to be functions of socio-economic and employment characteristics. The results show sex, income, race, household responsibilities, and job mobility to be important in determining commuting behavior. The monocentric model is based on a separation of the house and work place. Theoretically, this separation can be measured either in terms of commuting time or distance, however, most tests of the monocentric model have used commuting distance (Hamilton 1982; Blackley and Follain 1987). The time vs. distance issue is explicitly addressed in this paper by using both measures in the estimations. The results