| Chapter 2: | Government Consolidation |
Due to a lack of information about many of these factors, we used two variables to measure the crisis climate: the population growth rate and the minority population growth rate in the ten years preceding the referendum. Each of these variables may indicate an increase in the demand for local government services that outstrips local governments’ capacity to supply services.
The number of general-purpose governments in the county may also influence the likelihood of consolidation. The usual case is that the government of the largest city in a county merges with the county government. Often, there are smaller cities that are not included in the consolidation; depending on the consolidation charter, these cities may or may not be able to vote on the consolidation. Larger cities are perceived as having more economic development prospects, particularly for industrial recruitment, and city-county consolidation immediately increases the size of a city. Voters in nonconsolidating cities within the county may support a consolidation referendum if they perceive that the consolidation would create spillover effects that would be beneficial to these smaller, nonconsolidated cities. Enhanced industrial recruitment is one potential benefit of consolidation. The ability to attract professional sports teams is another benefit that has been mentioned in the popular press.
The underlying model suggests that a voter referendum on consolidation will be called if the benefits of consolidation are perceived to be greater than their cost by at least some of the voting population. The costs and benefits of consolidation are not directly observable. The observed outcome is whether the referendum passes or not.
To implement the model empirically, a probit model is used. The variable Yi* is an unobserved variable that indicates the incremental benefits that will result from government consolidation. The referendum will be expected to pass if Yi* is positive (i.e., if the benefits of consolidation are perceived to be greater than the costs). Rather than observing Yi*, one observes Yi—that is, whether the referendum passes or not.
As suggested by the model, Yi is related to a set of characteristics.