Urban Sprawl Analysis

Choosing the variables to capture the magnitude of the urban sprawl is the main challenge for the evaluation of the consequences of urban sprawl. One of the most well-known measures for assessing the urban sprawl employs variants of population density or developed areas as a proxy. However, the use of this kind of variables has been criticized for two main reasons. First, as pointed out by Hortas-Rico and Solé-Ollé (2010), there is no agreement regarding the right variables to capture density (density of housing units, population or employment), the extent of the space over which density should be characterized (total or urbanized area), and the scale at which density should be measured (metropolitan or municipality). Secondly, the density does not describe the urban areas properly, although it indicates the presence of scale of certain urban services, it fails showing the distribution of the

Both of these variables were largely discussed in the first essay of this dissertation, therefore, in the current analysis, we are only limited by using its results as a measure of urban sprawl. On the question of GI, the results from the fist essay indicated that the urban sprawl, measured by this index, prevails mainly in cities from high urban concentration. Due to it, we multiply the GI for a dummy variable that equals to 1 if the municipality belongs to high urban concentration, and 0 if the municipality belongs to medium urban concentration.

It is expected that sprawl raises the cost of providing public services because it fails to capitalize on economies of scale and/or optimize on facility location. Keeping the control variables equal, per capita spending on public services will be negatively influenced by population density and by GI, and positively influenced by the percentage of county urban land and by the