Table 1.
Derivation of the sprawl index.
| Attribute | Derivation |
|---|---|
| Centeredness | The centeredness variable is a measure of the degree of mono- or polycentrism within a metropolitan region and is based on three indicators: a density gradient, the percentage of the metropolitan population within a fixed radius of the central business district, and the number of population centers as defined by proximity of census tracts to regional density maxima. |
| Connectivity | The connectivity variable is a measure of the density of the street network and was based on the average block size and the percentage of blocks less than approximately 500 feet on a side (consistent with the dimension of a traditional urban block). As block size increases, the number of street intersections per unit of area decreases, which serves as an indicator of street network density. |
| Density | A composite density factor was derived through principal components analysis incorporating measures of gross population density, the proportion of metropolitan populations living at very low or very high densities, and the proximity of census tracts to urban centers. |
| Land-use mix | Three elements of land-use mix were integrated into a single, composite measure through principal components analysis. These elements include the ratio of jobs to population, the diversity of land uses, and the accessibility of residential uses to nonresidential uses at the level of the transportation analysis zone and within a 1-mile radius. |
| Sprawl index | A composite measure of urban compactness or sprawl was developed through an integration of these four urban form factors through principal components analysis. |
Each urban form attribute is reported on a scale with a mean value of 100 and an SD of 25 [across the 83 regions included in the Ewing et al. study (2003a)]. Higher values of the scores for centeredness, connectivity, density, and land-use mix reflect higher intensities of these attributes. Note that the Ewing et al. sprawl index scales in a negative direction (i.e., higher scores denote lower levels of sprawl) and has been modified in this study to scale in a positive direction (i.e., higher scores denote higher levels of sprawl) for ease of interpretation. The table is adapted from Stone (2008) and based on description of sprawl index from Ewing et al. (2003a).