Determinants of Urban Sprawl in France

This paper studies the determinants of urban sprawl in France using panel datasets for the four largest metropolitan areas (Paris, Lyon, Marseille, Lille) over the period 1985–98. A measure of urban sprawl is proposed at municipality level. Due to the huge heterogeneity of the panels, it seems difficult to make the fundamental homogeneity assumption underlying pooled models. Thus, random coefficient models under heteroscedasticity of the disturbances are estimated for each metropolitan area using a hierarchical Bayes approach based on the Markov chain Monte Carlo simulation method. It is found that urban sprawl is positively related to the income growth of the tax payers’ fiscal households for the period of rapid growth in the late 1980s. At the opposite extreme, the income effects are negative for non-tax payers. During the recession, income effects are significant neither for tax payers’ nor for non-tax payers’ fiscal households and are significantly positive for tax payers and negative for non-tax payers over the recovery. Finally, on average, the inequality index—difference of average net income between tax payers and exempted fiscal households—has a lower impact on urban sprawl than the income effect.