Determinants of Urban Sprawl: A Panel Data Approach

This paper applies fixed effects (within-groups) and between-groups estimations to panel data to test hypotheses of the monocentric urban model with urbanized area data for the period 1990�2010. The paper examines the impact of population, household income, transportation cost, and land rent at the urban fringe on urbanized area spatial size. The fixed effects regression finds that a 1- percent increase in population and a 1- percent decrease in travel cost causes an urbanized area to expand by 1.087 percent and 0.127 percent, respectively. The impact of household income is non-linear. The regression results from the between-groups estimation indicate that geographic and political factors help explain the spatial size differences across urbanized areas. The spatial size of an urbanized area is larger with a higher percentage of the urban fringe overlying aquifers, a higher percentage of local revenues from intergovernmental transfers, a higher percentage of urban fringe incorporated in 1980, and a lower elevation range in the urban fringe.