Predicting City Government Structure

Abstract

This study employs discriminant function analysis to test the capacity of demographic and regional variables to predict city governmental structure-mayor versus manager government, partisan versus nonpartisan elections, and ward versus at-large council constituencies. Ambiguities in the existing literature regarding the relative influence of ethnicity versus regionalism on structural reform are clarified. Discriminant function analysis provides a better estimate of how well demographic and regional variables discriminate between cities with different structural forms, and a better understanding of which independent variables are the best estimators of each of these structural arrangements. There are some ambiguities in the existing literature about the strength of the relationships between socioeconomic composition of urban populations and the structure of municipal government. Over a decade ago, John Kessel (1962) employed simple contingency tables to relate govemmental form (mayor, manager, or commission) to city size, growth rate, proportions of foreign-bom, and economic base. His work was augmented by Robert R. Alford and Harry M. Scoble (1965), who used contingency tables to examine the relationships between govemmental form and size, growth rate, mobility, region, educational level, white-collar occupation, and ethnicity. Alford and Scoble reasoned that the choice of the more politicized mayor-council form of govemment versus the more businesslike council-manager form of government was a function of: (1) social heterogeneity (religion, race, and ethnicity); (2) class composition, white-collar occupations, and college education; and (3) population growth and mobility: