Remaking Local Government : Success and Failure Under Indiana ’ s Government Modernization Act

During the period of 2008 to 2012, four consolidation attempts occurred under Indiana’s Government Modernization Act: two proposing city–county consolidation and two proposing town–township consolidation. The town–township mergers received overwhelming support, with more than 70% of voters approving consolidation in each case. The city–county mergers, on the other hand, were each defeated by a margin of approximately two to one. In this paper, we conduct a comparative case study of the four consolidation proposals using Leland and Thurmaier’s (2004) C3 model as an analytical framework. The results of our study indicate that greater demographic diversity and the potential for large cost shifts from urban to rural customers stimulated opposition to the city–county consolidations. The successful town– township consolidations were characterized by more homogeneous populations and modest and predictable fiscal impacts. We find no evidence that pre-existing functional consolidation or strong elite advocacy on behalf of consolidation leads to greater support for the referendum.

consolidation efforts.In conformance with Leland and Thurmaier's C 3 model as our analytical framework, our thesis is that important features of the institutional context, crisis climate, and consolidation charter facilitated support for consolidation in the town-township cases but stimulated opposition in the city-county cases.

Literature
Much of the literature on local government consolidation has focused on city-county consolidation, although only a small number of city-county consolidation attempts have been successful at reducing costs and enhancing revenue, and many other types of consolidation occur.For example, the consolidation of intra-county government units involving cities, villages, towns, and townships also have been considered.Gaffney and Marlowe (2014) examined changes in taxes and spending for six city-city consolidations and found that total expenditures increased in three cities and decreased in three after consolidation.The expenditure decreases were due to changes in debt service.Tax revenue increased in five of the six cities.Bunch and Strauss (1992) examined the potential consolidation of nine small municipalities in western Pennsylvania and found that consolidation may provide the realization of economies of scale and reduce costs and associated budget deficits in these communities.In his examination of the short-term effects of consolidation in five Canadian municipalities, Vojnovic (2000) found that only one city, Abbotsford, limited salary increases to an average of 1.5%, while the other cities had considerably larger increases.The ability of Abbotsford to contain salary increases was due to explicit agreements regarding salary changes before the consolidation.Overall costs increased (or were expected to increase) in three municipalities and decreased in two.In her analysis of consolidation in Ottawa, Canada, Reese (2004) noted that pay and benefit levels tended to increase as differing pay systems and union contracts were renegotiated.Because wages and salaries are usually the largest component of local government budgets, increases in pay and benefit levels are likely to be associated with increases in overall expenditures.
Much of the academic literature on consolidation has focused on the process of city-county consolidation.Prior analyses consist of case studies of specific communities (Durning, 1995;Leland & Thurmaier, 2000;Savitch & Vogel, 2000, 2004), with much of this literature describing successful cases rather than analyzing why the consolidation attempts succeeded (Leland & Thurmaier, 2004).Recent comparative case studies examining the process of consolidation within a causal model (Leland & Thurmaier, 2004, 2005, 2006) bridge this gap.These comparative case studies show that consolidation attempts linked to economic development are more likely to be successful, while attempts linked to equity via redistribution of income from suburbs to city and efficiency gains from reducing duplication and achieving economies of scale are likely to fail.
Local government consolidation attempts may be unsuccessful for a variety of reasons.In his analysis of transaction costs associated with the merger of Norwegian municipalities, Sorenson (2006) found that revenue disparities and differences in party preferences among municipalities are impediments to voluntary merger.Large efficiency gains also make merger more likely to be supported by elected and administrative leaders.Leland andThurmaier (2005, 2006) found that consolidation referenda that are approved by voters usually focus on the economic development aspects of consolidation rather than on increased equity or efficiency.
A broader literature has examined how population heterogeneity affects the number of political jurisdictions, which ultimately affects consolidation.Alesina, Baqir, and Hoxby (2004) examined factors that influence the number of political jurisdictions in a county and found that both racial and income heterogeneity positively influence the number of jurisdictions (school districts and municipalities) with race having a stronger influence.As the probability of interacting with a person of another race or income level increases, there are more jurisdictions in a county.They also note that lack of consolidation even in the face of economies of scale is how jurisdictions respond to heterogeneity.In their analysis of the amalgamation of Swedish municipalities, Hanes, Wikstrom, and Wangmar (2012) found that income differences affected municipalities' willingness to voluntarily merge.Bates, Lafrancois, and Santerre (2011) found that income differences between Connecticut public health districts negatively influence the consolidation of these districts.
To explain successful consolidation referenda, Rosenbaum and Kammerer (1974) proposed a model (R&K model), which has since been widely applied.The model has three basic elements: 1.A crisis climate develops.It may result from demographic shifts, changes in the quality of government services, or other factors.Citizens demand a government response.2. If citizens are not satisfied with the government's response, they lose confidence in the structure of the local government and support for consolidation develops, usually bolstered by the support of civic organizations and the local media.3. Finally, the initial support for consolidation is strengthened by accelerator events, such as a scandal or the loss of a leader.Ultimately, the effort coalesces in a consolidation referendum being on the ballot.Leland and Thurmaier (2004) augmented the Rosenbaum and Kammerer (1974) model incorporating the work of Johnson and Feiock (1999) and Feiock and Carr (2000) to develop a causal model of the consolidation process called the C3 model of local government consolidation with testable hypotheses.The C3 model adds the referendum campaign as the final stage of the model, incorporates institutional framework, particularly the legal framework in which the consolidation attempt occurs, considers voter alienation, includes the impact of specific types of charter provisions, and expands the role that civic elites play in the process (Leland & Thurmaier, 2004).
This paper extends the literature in two primary ways.First, we consider and compare two types of local government consolidations city-county and town-township.Previous research has focused primarily on city-county consolidations with limited study of other types.Second, our consolidation cases occurred within a single state over a limited time period and governed by the same state enabling legislation, allowing us to focus on local differences.Other comparative case studies have examined cases from different states, governed by different legislation, and further apart in time, making it more difficult to make inferences.

Local Government in Indiana
Local government in Indiana consists of counties, municipalities (cities and towns), townships, school districts, and special districts (see table 1).Indiana cities are divided into three classes depending on population: first-class cities have 600,000 inhabitants or more; second-class cities have between 35,000 and 599,999 inhabitants; and third-class cities have fewer than 35,000 inhabitants.Other municipalities, regardless of population, are towns.Second-and third-class cities are governed by a mayor and city council.Towns are governed by a town council (and appointed town manager).Towns with a population of 2,000 or more may become cities through a petition of voters and a referendum.Likewise, cities may become towns through a petition of two-thirds of taxpayers to the circuit court.Indianapolis is the only first-class city

Local Government Consolidation in Indiana
There is one consolidated city-county government in Indiana, UNIGOV, which consolidated Indianapolis and Marion County in 1970 through a 1969 Act of the General Assembly.Blomquist andParks (1993, 1995) provide information on the background and structure of UNIGOV.Segedy and Lyons (2001) and Rosentraub (2000) evaluate various aspects of UNIGOV.
A consolidation referendum was held in Evansville and Vanderburgh County in 1974, but Buschmann (2007), the two paths of this act seek to achieve balance between competing interests by equally including all affected units of government, preserving their governing autonomy, as well as embracing citizen participation in the reorganization process.
Prior to the GMA, consolidation attempts in Indiana required special legislation.Schaal, Taylor, and Faulk (2017) analyze southern states for the presence or absence of general law enabling consolidation and find that those states without general law tend to have no consolidation referenda.Only two states without general law enabling consolidation, Georgia and Florida, run counter to the trend and have held many referenda as a result of special legislation.Since passage of the GMA, several more consolidation attempts for a variety of types of governments have occurred in Indiana, providing further evidence that authorization for consolidation under general law facilitates consolidation efforts.

Data and Method
Following Leland, and Thurmaier (2004), we utilize the comparative case study method to analyze these consolidation attempts, using their C3 model as an analytical framework.Not only do case studies provide rich details, but the comparative research design allows us to "measure each case against testable (i.e., falsifiable) hypotheses" (Leland & Thurmaier, 2004).
We compare case studies of four consolidation attempts in Indiana, which occurred from 2008 to 2012: Town of Zionsville-Eagle Township-Union Township in 2008, Town of Yorktown-Mt.Pleasant Township in 2010, City of Muncie-Delaware County in 2012, and Evansville-Vanderburgh County in 2012 (see figure 1).Focusing on consolidation attempts occurring within one state and within a limited time period allows us to hold constant key variables related to the institutional context while examining the effect of variation in other characteristics across each case.
In each case study, we draw much of our data from the plans of reorganization, which are similar to a consolidation charter, and contemporaneous news reports about the development of

Hypotheses
Drawing on Leland and Thurmaier's (2004) C3 model, we focus on key characteristics of the institutional context, charter provisions, and referendum campaign to explain the disparate outcomes of the four consolidation referenda analyzed in this study.Each of these characteristics and our hypotheses are presented below.
Institutional Context Leland and Thurmaier (2004) suggest that the legal framework governing consolidation attempts is a key variable affecting the likelihood that a consolidation attempt reaches a referendum.In our study, the legal framework is held constant across all cases, as all four consolidation attempts were governed by the Government Modernization Act, which allows locally initiated consolidation attempts to reach referenda.By holding this aspect of the study constant, we are able to focus on other variables and their impact on consolidation outcomes.
A community's social and economic characteristics represent other important aspects of the institutional context within which a consolidation attempt takes place.Leland and Thurmaier (2004) suggest that, when the urban and non-urban interests in a community have diverging interests, then it will be more difficult to gain approval for consolidation than in a more homogenous community.Our first hypothesis is related to these characteristics: In communities in which the urban and non-urban populations exhibit diverse social and economic characteristics, there will be lower support for political consolidation than in communities in which these populations are homogenous.
Another potentially influential factor within the institutional context is the existing state of functional consolidation in a community.Leland and Thurmaier (2004) explain that the existence of interlocal agreements for the provision of shared services may reflect a high level of trust between the governmental units.This trust may make it easier for the affected governments to consider political consolidation.In addition, existing functional consolidation decreases the complexity of the consolidation process, which may lead to greater support for consolidation.On the other hand, a high level of functional consolidation may decrease support for political consolidation because many of the benefits of shared service provision have already been gained.For the purposes of analysis, we state our second hypothesis in a form in which functional consolidation is expected to lend support to political consolidation.
H 2: In communities with a high level of functional consolidation, there will be greater support for political consolidation than in communities with little functional consolidation.
Charter Provisions Leland and Thurmaier (2004) contend that specific provisions of the consolidation proposal may help or hinder the campaign in support of consolidation.Key provisions include those that affect taxes and the number of elected officials in the consolidated government.
Proposals that avoid large changes in tax burdens are likely to receive greater support than plans that dramatically increase taxes (Leland & Thurmaier, 2004).Many city-county consolidation charters utilize provisions intended to minimize the impact on tax burdens, such as the creation of urban and non-urban tax districts with different levels of service and tax rates.Despite these provisions, the merging of city and county departments providing similar services may result in cost shifting between groups of taxpayers.In contrast, town-township consolidations encompassing a relatively smaller scope of services and with less duplication of services are likely to result in less cost shifting.The cost shifting inherent in city-county consolidation may stimulate opposition by groups of taxpayers who expect an increased tax burden as a result of consolidation, leading to our third hypothesis: H3: In communities in which political consolidation is expected to result in significant tax increases for some taxpayers, there will be lower support for political consolidation than in communities in which taxes are unchanged or reduced for all.
Another important charter provision is the size and structure of the unified council for the consolidated government.Leland and Thurmaier (2004) suggest that a council reduced in size compared the combined city and county councils and consisting of a mix of district and at-large seats will be more likely to garner support than other configurations.Smaller councils will be perceived as more economical, and the mix of representation forms will allow the representation of geographically based interests and countywide perspectives.
All four consolidation charters analyzed in this study provided for a mix of district and at-large representation on the consolidated council.They varied, however, in the extent to which they reduced the number of elected officials in the post-consolidated government.Because the number of elected officials provides an easy to understand -if rough -measure of how much "streamlining" of government will result from consolidation, we suggest our fourth hypothesis: In communities in which political consolidation will result in a large reduction in the number of elected officials, there will be greater support for political consolidation than in communities in which the reduction is small.

Referendum Campaign
The referendum campaign is the final stage of the consolidation process.Leland and Thurmaier (2004) suggest that the role of consolidation entrepreneurs and community elites in the campaign will be a key determinant of the success or failure of the referendum.We explore this aspect of the model by focusing on the role of community elites in our final hypothesis.In particular, we examine the role of the sheriff.Leland and Thurmaier (2004) argue that the county sheriff is a power factor in other consolidation attempts.
H 5: In communities in which community elites are strongly supportive of political consolidation, there will be greater support for political consolidation than in communities in which community elites are largely opposed or divided.
In the next section, we examine the history and characteristics of the four consolidation efforts in order to test these hypotheses.

Analysis and Discussion
Indiana's Government Modernization Act (GMA) governed all four consolidation attempts and provided the means by which consolidation proponents were able to place consolidation on the agenda.The GMA provides two paths for forming a reorganization committee charged with developing a plan of reorganization (the consolidation charter).Under one alternative, the governing bodies of the local government units considering reorganization create the reorganization committee by enacting identical resolutions.This was the path followed in both town-township consolidations.In the Zionsville-Union-Eagle and Yorktown-Mount Pleasant consolidations, the town council and township board(s) were the parties that initiated the consolidation effort by enacting resolutions and appointing the committee members.
The other alternative allows reorganization proponents to file petitions supporting the creation of a reorganization committee.A petition signed by at least 5% of the voters in the subdivision must be filed with each political subdivision to be included in the reorganization proposal. 3This alternative was exercised in both city-county consolidations.In the case of the Evansville-Vanderburgh consolidation, the League of Women Voters of Southwestern Indiana organized the petition drive (Langhorne, 2009a).In the Muncie-Delaware consolidation, a local taxpayer group, Citizens of Delaware County for Property Tax Repeal (CDCPTR), was the petition drive organizer (Werner, 2010).Once the petitions were filed and verified by the county clerk, the GMA required the governing bodies of the subject political subdivisions to enact the resolutions necessary for forming a reorganization committee and then appoint members to the committee.

Crisis Climate and Power Deflation
Although we make no predictions related to the crisis climate or power deflation in these cases, a review of events leading up to the consolidation proposals provides a fuller picture of the context in which each consolidation proposal was advanced.Some elements of the crisis climates were a result of events affecting all Indiana local governments, such as the recently enacted property tax caps and the attention that the Kernan-Shepard report had attracted to local government reorganization.Other elements of the four communities' crisis climates were more local in nature.
In 2008, the Indiana General Assembly had enacted property tax caps that were partially implemented in 2009 and fully implemented in 2010 [see Faulk (2013) for a description of Indiana's property tax caps and revenue impacts].The result of the tax caps was that nearly every local government unit in the state saw a reduction in property tax revenues, some fairly modest, others quite severe.In Muncie and Delaware County, the impact of the tax caps was large.Muncie saw a greater than 30% reduction in property tax revenues; Delaware County government saw a reduction of more than 17%.The impacts in Evansville and Vanderburgh County were more modest, less than 5% Evansville and less than 4% for Vanderburgh County government.The tax cap impacts in the towns and townships varied.In Zionsville, tax caps reduced revenues by around 6% and less than 2% in the townships.Impacts in Yorktown-Mount Pleasant were larger, around 10% for Yorktown and 8% for Mount Pleasant Township.
The Kernan-Shepard report, released in 2007, had drawn a great deal of attention to the topic of government reorganization and restructuring (Langhorne, 2009b).The report (Indiana Commission on Local Government Reform, 2007) recommended sweeping changes to Indiana local governments, including eliminating townships, replacing each county's three-member county commission with a single elected county executive, and consolidation of schools with less than 2,000 students.Although almost none of the recommendations were enacted by the General Assembly, during the years immediately after the report was released many local governments were concerned that the General Assembly would make large changes to local government in Indiana.In Muncie-Delaware and Evansville-Vanderburgh, fiscal conditions heightened the crisis climate and led to controversies that could be considered to constitute power deflation.
Responding to the large reduction in property tax revenue, Muncie Mayor Sharon McShurley laid off 32 firefighters and five police officers (Smith, 2009).This controversial decision contributed to an environment in which the potential for consolidation-related efficiencies was attractive to local activists, such as the CDCPTR.
In Vanderburgh County, uncertainty about the eventual impact of the tax caps resulted in a decision by the county to not renew a homestead property tax exemption (Langhorne, 2009c).
The homestead credit was later reinstated, which, along with the impact of the tax caps, resulted in a $15 million combined budget shortfall for the Evansville and the county, leading the local newspaper's editorial board to call for consideration of local government consolidation (Evansville, Vanderburgh County facing $15 million fiscal crisis, 2009).The local League of Women Voters' president referred to a need for greater government efficiency when explaining the reason for the petition drive (Langhorne, 2009a).
In Zionsville and Yorktown, property development and annexation issues contributed to the crisis climate.Zionsville officials had expressed concern about rapid development just outside town limits, which was governed by Boone County's more permissive development standards (Woodson, 2008).There also was concern among town and township officials that aggressive annexation by the nearby community of Whitestown would limit Zionsville's ability to expand (Annis, 2007;Olson, 2007).
In Delaware County, Yorktown and Muncie had been engaged in an annexation battle dating back to 2005 (Slabaugh, 2005;Yencer, 2005).This territorial conflict intensified when the Muncie-Delaware consolidation effort got underway.Yorktown and Mount Pleasant officials initiated their own consolidation effort, in part, to minimize the impact on Yorktown and Mount Pleasant in the event that the Muncie-Delaware effort was successful (Yorktown-Mt.Pleasant Township Reorganization Committee, 2011).In both the Yorktown and Zionsville cases, it is difficult to identify any instances of power deflation.The consolidation efforts served as the responses to the crises and in both cases were ultimately successful.

Institutional Context
Once set in motion in each community, the consolidation efforts proceeded under a common legal framework, but other aspects of the institutional context varied across the four communities.Our first hypothesis suggested that the level of support for consolidation would depend, in large part, on the level of socioeconomic diversity in the community.Specifically, we expect that greater support will be exhibited in communities where the urban and non-urban populations are more similar.Table 2 compares the urban and non-urban populations in each case on three dimensions: racial diversity, income, and poverty.In the city-county consolidation cases, "non-urban" refers to the portion of the county outside the city limits.In the towntownship cases, "non-urban" refers to the portion of the township(s) outside the town limits.
A review of the data presented in table 2 reveals that the urban and non-urban areas in the two city-county consolidation cases were very different.In both Muncie and Evansville, the urban population was poorer, had much lower earnings, and consisted of a greater proportion of nonwhite residents than in the non-urban areas of the county.The communities in which the towntownship consolidations occurred were more homogenous.Zionsville and the surrounding townships were similar.Yorktown was different from the other communities in that the urban population was less diverse and wealthier than the surrounding area, but the urban and nonurban areas were still more similar than in either of the city-county consolidation cases.These data indicate that the more homogeneous communities experienced greater support for consolidation than the more diverse communities, as predicted by our first hypothesis.
Our second hypothesis addressed the effect of functional consolidation on support for political consolidation.We hypothesized those communities in which a high level of functional consolidation would exhibit greater support for political consolidation.Evansville-Vanderburgh was the only community in which a large number of local government services were functionally consolidated.After the 1974 consolidation referendum was defeated, city and county officials proceeded to implement many of the consolidated functions that had been proposed in the consolidation plan.As a result of this functional consolidation, by the time of the 2012 referendum, Evansville and Vanderburgh County had 12 jointly funded local government departments: traffic engineering, central dispatch, animal control, human relations commission, purchasing, computer services, solid waste management, emergency management, building commission, area plan commission, and legal aid society (Langhorne, 2010a).
In the other communities, fewer services had been functionally consolidated.Muncie and Delaware County's only jointly funded department was central dispatch.In Zionsville-Eagle-Union and Yorktown-Mount Pleasant, only the fire departments were jointly funded.In the case of the towns and townships, however, the fire department is the only government service for which both towns and townships are responsible, so one might claim that the towns and their surrounding townships had already implemented all the functional consolidation that was possible.In any event, within our four cases, there doesn't appear to be any systematic relationship between functional consolidation and support for political consolidation; thus, our second hypothesis is not supported.

Charter Provisions
The consolidation charters describe how the consolidated government will be structured, and the specific provisions of the charter will influence the level of support it receives (Leland & Thurmaier, 2004).In all four cases, the reorganization committee included common charter provisions intended to garner support for consolidation.All four reorganization plans provided for at least two service districts with different levels of service and tax rates that would enable non-urban residents to avoid receiving -and paying for -the more intensive urban services provided to urban residents.All four plans also preserved, at least initially, the less-restrictive county land use ordinances that were presently in effect in the non-urban areas of the county or township (City of Evansville-Vanderburgh County Reorganization Committee, 2011;Communities of Zionsville Area for Better Government, 2008;Muncie-Delaware County Government Reorganization Committee, 2010;Yorktown-Mt. Pleasant Township Reorganization Committee, 2011).These provisions were intended to minimize the fiscal and regulatory impact of consolidation on the non-urban populations affected by the consolidation plans.
In the two town-township consolidations, only a very small fiscal impact was projected in either case.The only high-cost common service between the towns and townships was fire protection; in both cases the communities were already jointly funding their local fire departments under longstanding agreements.The remaining township services, such as park and cemetery upkeep and poor relief, would be transferred to the town government with no expectation that service costs were likely to change.The urban service districts consisting of the area within the preconsolidation town limits would continue receiving and paying for the level of service provided prior to consolidation.The newly consolidated governments would enter into memorandums of understanding (MOUs) with the county governments to continue providing the preconsolidation level of sheriff patrol and road maintenance (Communities of Zionsville Area for Better Government, 2008; Yorktown-Mt.Pleasant Township Reorganization Committee, 2011).As a result, the town-township consolidations were projected to produce small cost reductions for both urban and non-urban residents, primarily due to the elimination of the town trustee's salary (Crowe Horwath, 2011a;Woodson, 2007).
In the two city-county consolidations, the fiscal impact of the consolidation was a matter of contention between consolidation supporters and opponents.In both cases, one item of concern was the shifting of sheriff patrol costs from urban to non-urban taxpayers.In both reorganization plans, the police patrol function was to remain unconsolidated.The urban areas would continue to be served by the existing police department, and the sheriff would continue patrolling the non-urban areas (City of Evansville-Vanderburgh County Reorganization Committee, 2011; Muncie-Delaware County Government Reorganization Committee, 2010).In the pre-consolidation situations, city property owners, as county taxpayers, help fund the sheriff's patrol.In both cases, however, financial analyses produced by Crowe Horwath projected a substantial post-consolidation increase in the tax rate for non-urban taxpayers and decrease for urban taxpayers as a result of allocating all sheriff patrol costs to the non-urban service district (Crowe Horwath, 2010, 2011b;Gootee, 2011a).
In our third hypothesis, we predicted that consolidations in which some taxpayers were expected to experience large post-consolidation tax increases would receive less support than consolidations in which taxes are unchanged or reduced for all.In both city-county consolidations, fiscal analyses projected a potentially large shift in costs from urban to rural taxpayers.Both of these proposals were defeated soundly in the referendum.In the towntownship consolidations, however, fiscal projections indicated that there would be a modest decrease in spending and little overall change in tax rates.Both of these proposals were strongly supported at the polls.Thus, our third hypothesis is supported.
Another potentially important charter provision is in regards to the number of elected officials in the consolidated government.In our fourth hypothesis we suggest that comparing the number of elected officials in the pre-and post-consolidation governments will provide observers with an easy to understand -if rough -measure of how much the consolidation will "streamline" local government.In all four cases, the reorganization plan called for a smaller number of elected officials in the consolidated government than in the separate governments.
The two city-county consolidations include a number of constitutional offices, such as sheriff, auditor, and treasurer, which could not be eliminated, although they could potentially be stripped of their powers (and salary) in the consolidated government.There is potential, however, for a reduction in the executive and legislative officials.In Evansville-Vanderburgh and Muncie-Delaware, the reorganization plan called for a reduction in executive and legislative elected officials from 20 (one city mayor, nine city council members, three county commissioners, and seven county council members) to 16 (one mayor or county executive and 15 council members), a 20% reduction (City of Evansville-Vanderburgh County Reorganization Committee, 2011;Muncie-Delaware County Government Reorganization Committee, 2010).
In Yorktown-Mount Pleasant, the consolidation reduced the executive and legislative officials from nine (five town council members, three township board members, and one township trustee) to seven (all town council members), a 22% reduction.In Zionsville-Eagle-Union the reduction was larger because of the elimination of two townships instead of one.In this case, elected officials were reduced from 13 (five town council members, six township board members, and two trustees) to seven (all town council members), a 46% reduction (Communities of Zionsville Area for Better Government, 2008; Yorktown-Mt.Pleasant Township Reorganization Committee, 2011).
Although the Zionsville consolidation resulted in a relatively large reduction in elected officials, the reductions in the other three cases were roughly the same.These results provide no support for our fourth hypothesis.

Referendum Campaign
In our fifth and final hypothesis, we predicted that campaigns in which community elites are strongly supportive would have greater success than those in which elites are divided or opposed.In the Zionsville-Eagle-Union and Yorktown-Mount Pleasant consolidations, community elites in the form of town council members, township board members, and township trustees were responsible for initiating the consolidation efforts and were supportive of the efforts of the reorganization committees in their respective communities.We are not aware of any organized opposition to either of these consolidation referenda.In these two cases, community elites were strongly supportive.
In Muncie-Delaware and Evansville-Vanderburgh elected officials did not take the initiative in placing consolidation upon the public agenda.In both cases, it was left to other groups to circulate petitions to require the elected officials to act.The Evansville-Vanderburgh effort was initiated by the League of Women Voters of Southwestern Indiana, who conducted the petition drive needed to initiate the consolidation process (Evansville-Vanderburgh County Consolidation, 2009).The Muncie-Delaware consolidation effort was initiated by the Citizens of Delaware County for Property Tax Repeal (Werner, 2010).
In Evansville-Vanderburgh, however, there were many supportive community elites.The mayor and at least one county commissioner expressed support for the effort early in the petition stage (Langhorne, 2009a(Langhorne, , 2009b)).The editorial board of the local newspaper was a strong advocate in support of the consolidation referendum (Vote 'yes' to bring community into the 21 st century, 2012).The sheriff supported the consolidation effort during the campaign, but his support came only after the resolution of a long, drawn-out controversy over whether to merge the Evansville Police Department and Vanderburgh County Sheriff Office (Langhorne, 2012a).Leland and Thurmaier (2004) argued that, among elites, the county sheriff is a particularly influential force in city-county consolidation attempts.In the Evansville-Vanderburgh effort, conflict between the sheriff and police chief during the drafting of the consolidation plan highlighted the sheriff's strong influence.Early in the process, the Evansville police chief and Vanderburgh County sheriff submitted competing proposals for consideration by the public safety subcommittee.Under the chief's proposal, the police department and sheriff's office would remain separate, but the police department would take on responsibility for law enforcement operations on a countywide basis.The sheriff's department would be responsible only for operation of the county jail, court security, and serving court papers.Under the sheriff's proposal, the two law enforcement agencies would be merged into one, under the supervision of the sheriff (Langhorne, 2010b).The public safety subcommittee voted 3-1 to recommend adoption of the sheriff's proposal by the full reorganization committee.The sheriff and other supporters of his proposal argued that law enforcement operations should be led by an elected official (Langhorne, 2010c).
As the full reorganization committee drafted the reorganization plan, the police chief and local chapter of the Fraternal Order of Police (FOP) continued to argue against consolidation of the two agencies.The FOP argued that consolidation would lead to increased costs and reduced efficiency (Langhorne, 2010c).The police chief raised concerns about the potential for conflict between the mayor and sheriff that could negatively impact law enforcement.The sheriff continued to argue that accountability for public safety properly belonged to an elected sheriffnot to an appointed police chief (Langhorne, 2010d).
At one point, the reorganization committee rejected both proposals and developed a novel proposal of its own.Under this proposal, the two agencies would remain separate, but the sheriff -not the mayor -would appoint the police chief.This plan pleased neither the chief nor the city police officers, and the sheriff found it merely "palatable" (Langhorne, 2010e).Ultimately, however, the sheriff's original proposal was included in the plan of reorganization submitted by the committee (City of Evansville-Vanderburgh County Reorganization Committee, 2011).
Under the Government Modernization Act, the governing bodies that appointed the reorganization committee, in this case the Evansville City Council and Vanderburgh County Commission, have the power to amend the plan of reorganization prior to submitting it to the voters.Once this process began, some city council members expressed concern that the law enforcement merger might be a "killer" because of opposition among their constituents (Gootee, 2011a).Officials decided to remove the law enforcement merger from the plan.Discussion then turned to whether to include provisions in the plan for future reconsideration of merging the agencies (Gootee, 2011b).Although the sheriff continued lobbying for a merger, the final plan prohibited any consideration of a law enforcement merger until after the 2024 election, if approved by voters in a referendum (Gootee, 2011c(Gootee, , 2011d)).Despite this change to the plan, the sheriff supported consolidation during the campaign, while the Fraternal Order of Police opposed it (Langhorne, 2012a).
Over $400,000 was spent by supporters and opponents in the referendum campaign with approximately three quarters of the total being spent by the supporters.Pro-consolidation contributors included the local chamber of commerce, a number of individual businesses, and the mayor's political action fund.Opponents included the Farm Bureau and many individual contributors (Langhorne, 2012b).
In the Muncie-Delaware consolidation, the proposal had little support among community elites.While one city council member expressed support for the consolidation proposal (Gregory, 2012), the mayor, two of three county commissioners, including the commission president, and the local leaders of both political parties all opposed the consolidation effort (Roysdon & Walker, 2012a).Nothing in the public record indicates whether the sheriff took a position.In this case, unlike in Evansville-Vanderburgh, there had never been any serious discussion about merging the city police and county sheriff departments.The chair of the reorganization committee indicated that they decided very early in the process to leave the two departments separate in an attempt to avoid controversy (E.Kelly, personal communication, September 26, 2016).Unlike in Evansville-Vanderburgh, where the city council and county commission amended the plan in a manner intended to head off opposition, the Muncie City Council and Delaware County Commission used their authority to make the proposal unlikely to pass by including a double supermajority requirement for passage (Roysdon & Walker, 2012a).The local newspaper, while supportive of the goals of consolidation, urged voters to reject the proposal because it was "not ready" (Muncie StarPress, 2012).
Campaign contributions in support of and opposition to the consolidation proposal were much lower than in Evansville-Vanderburgh (less than $15,000 total), but most of the campaign support (almost $13,000) went to the opposition committee (Roysdon & Walker, 2012b).In both the Evansville-Vanderburgh and Muncie-Delaware consolidation efforts, members of the minority community expressed concerns about representation under a consolidated government, with the local NAACP president urging a "no" vote on the Muncie-Delaware proposal (Harley, 2012;Langhorne, 2012c).
In any event, these results do not support our hypothesis that elite support influences the referendum results.In Evansville-Vanderburgh there were many community elites, including the sheriff, who devoted a high level of resources to a well-funded campaign in support of the consolidation proposal, and they lost by a two-to-one margin, nearly the same as in Muncie-Delaware, where elites were strongly opposed to consolidation, and the supporters were far outspent by the opponents.

Conclusions and Extensions
Our paper enhances the consolidation literature first by examining town-township consolidations in addition to city-county consolidations.Second, it focuses on the single state of Indiana within a limited time frame guided by the same enabling legislation, thus allowing us to carefully discern local governmental differences.Comparing the institutional frameworks, charter provisions, and referendum campaigns for four consolidations involving two towntownship cases and two city-county cases, our findings indicate that homogeneous communities are more likely to support consolidation efforts than diverse communities.Perhaps the more homogeneous demographics in the town-township consolidations led to more willingness for residents to trust that post-consolidation issues could be resolved.Such trust and reciprocity often develop when communities are socially engaged in networks (see Putnam, 2000).
We also find that projected fiscal impacts in the four consolidation cases substantiate our hypothesis stating that there will be less support for political consolidation if such consolidation is expected to yield significant tax increases for some taxpayers.This comparison was made possible by the inclusion of the relatively uncomplicated town-township consolidations.
Because the towns and townships had few overlapping services, except fire protection, which they were already providing jointly, there was little risk of cost shifting.Neither was there much opportunity for cost savings.In this simplified problem, environment voters could make a decision about consolidation in terms of how well it addressed the issues in the crisis climatethreats of outside development in these cases.If these town-township consolidations had been affected by threats of cost shifting similar to those in the city-county consolidations, then nonurban voters would have been faced with a more difficult cost-benefit decision in deciding if protection from unwanted encroachment was worth the increased cost of services.But in these cases, they could obtain the benefits of consolidation essentially for free.Thus, uncertainty about cost shifting between urban and rural taxpayers seems to stir up opposition to consolidation.
Our analysis fails to demonstrate, however, that communities with existing high levels of functional consolidation will have greater support for political consolidation.Duplicative services create the opportunity for cost-savings but also creates the possibility of this cost shifting.The town-township consolidations had little overlap in services and, therefore, little opportunity for savings, but this situation creates more certainty about post-consolidation cost of services.Neither do we find that large reductions in the number of elected officials has any discernable impact on the level of support for political consolidation.
Last, this paper does not find that high levels of support by community elites will necessarily translate into greater support for political consolidation.In the face of opposition, elite support and a well-funded campaign are not sufficient to achieve passage of a city-county consolidation in all cases.
For future research, we can compare these successful town-township consolidations to three proposed consolidations that were initiated during the same time frame but failed to reach referendum in the Towns of Avon and Brownsburg and the City of Greenwood with their surrounding townships.This future analysis will allow us to examine if the factors that seemed to lead to success in Zionsville and Yorktown were missing in these other three cases, leading to their failure.
Examining these unsuccessful cases also may allow us to estimate the importance of the annexation wars that occurred in the Zionsville and Yorktown areas during the period prior to consolidation.In both cases, rural residents faced the prospect of annexation by what many may have perceived as a less desirable municipality.These annexation prospects may have generated support for consolidation in the rural areas, but we cannot draw any inferences about the impact because the condition was present in both town-township consolidations.

Table 2 .
Socioeconomic Characteristics and Referendum Support