The assumption that urban workers economize on commuting is implicit in urban economic theory. Yet it has been challenged by some recent studies. This paper estimates commute flows implied by three urban density functions: monocentric, polycentric, and dispersive. It finds that an urban density function better predicting the actual spatial patterns also better explains the actual commuting behavior. This finding helps us to preserve the assumption that urban workers make attempts to economize on commuting in their location choices.

This paper examines the population distribution in a rapid development area - Orange County in 1980. The population density gradients are estimated with a polycentric model as well as with a monocentric model. The paper shows that the polycentric model of population distribution fits Orange County better than the monocentric model, and that the population in Orange County is more dispersed than in Los Angeles County. It is found that Downtown Los Angeles has a significant influence on the population distribution in Orange County. The aggregate influences of centers on population distribution have a wide range which are affected by the inclusion of center Downtown Los Angeles.

This paper examines the relationship between urban structure and commuting behavior. Analyzing the 1980 journey-to-work data for the Los Angeles region, this paper has shown that polycentric density functions fit the actual urban structure better than the conventional monocentric model. This finding indicates the preeminence of accessibility to major employment centers in location choices.

This paper also estimates commute flows implied by the polycentric and monocentric functions. It finds the monocentric model very poor at explaining commuting behavior. The empirical results show that polycentric urban structure increases the urban commute. This finding helps to preserve the assumption that urban workers economize on commuting, and suggests that efforts to promote more efficient urban form, such as the jobs-housing balance policy, have the potential to succeed.

This paper examines the spatial pattern of worker residences with three different density functions: monocentric, polycentric, and dispersive. Analysis of the 1980 journey-to-work census data for the Los Angeles region reveals that the polycentric density function statistically explains the actual distribution better than the monocentric density function, but the dispersive density function fits best. These findings confirm a polycentric spatial pattern, and also imply that overall accessibility to employment opportunities is the primary determinant of residential location choices.

This paper examines the relationship between urban structure and commuting behavior. Analyzing the 1980 journey-to-work data for the Los Angeles region, this paper has shown that polycenteric density functions fit the actual urban structure better than the conventional monocentric model. This finding indicates the preeminence of accessibilty to major employment centers in location choices.

This paper also estimates commute flows implied by the polycentric and monocentric functions. It finds the monocentric model very poor at explaining commuting behavior. The empirical results show that polycentric urban structure increases the urban commute. This finding helps to preserve the assumption that urban workers economize on commuting, and suggests that efforts to promote more efficient urban form, such as the jobs-housing balance policy, have the potential to succeed.

We examine spatial patterns and their changes during the 1970s for the Los Angeles region, by estimating monocentric and polycentric density functions for employment and population. Downtown Los Angeles is clearly identified as the statistical monocentric center of the region, and it is the most consistently strong center in the polycentric patterns. Polycentric models fit statistically better than monocentric models, and there was some shift in employment distribution toward a more polycentric pattern. These findings verify the existence of polycentricity in Los Angeles and demonstrate for the first time that employment and especially population follow a polycentric pattern based on exogenously defined employment centers. The results confirm that both employment and population became more dispersed during the 1970s.

A debate over the empirical underpinnings of urban economic models is emerging under the unlikely rubric of "wasteful commuting." Hamilton (1982) shows that a commonly used monocentric model, in which employment and population densities decline exponentially from a center, greatly underpredicts actual commuting distances in typical U.S. and Japanese metropolitan areas. He concludes that the monocentric model is fundamentally flawed. This conclusion is challenged by White (1988b), who examines the cost-minimizing assignment of households to residential locations, taking density patterns as they are and measuring cost by travel time. White finds that for a sample of U.S. metropolitan areas, only 11 percent of actual commuting cost is in excess of the cost-minimizing amount, rather than the 87 percent found by Hamilton. Hamilton (1989) and Cropper and Gordon (1991), using variations of White's technique, obtain results intermediate between these extremes.