In 2006, a revolution began in Sandy Springs, Georgia, located in Fulton County within the Atlanta metropolitan area. Long pining for their own city, Sandy Springs residents finally achieved their goal after decades of advocacy, becoming the first new metropolitan city Georgia witnessed in decades. The secession from Fulton County had far-reaching effects, sparking multiple other municipal incorporations in the area--Milton, Johns Creek, Chattahoochee Hills--even inspiring new cities to form outside the borders of Fulton County in adjacent counties.
What stimulates the formation of such clusters? Scholars have long noted that new municipalities--new cities, towns, and villages--incorporate in clusters; the pattern is not geographically random (Schmandt, 1961; Stauber, 1965; Smith, 2008; Smith & Debbage, 2006, 2011; Waldner, Rice, & Smith, 2013). Previous scholarly work suggested annexation might play a role (Stauber, 1965; Rigos & Spindler, 1997; Smith & Debbage, 2006), yet the cause of clustering remains unknown as few scholars have actually studied the phenomenon.
Stauber (1965) was among the first to confirm the existence of clusters. He demonstrated that new municipalities tend to occur in clusters, close together in time and space. He noted that "For over a generation, the phenomenon of multiple incorporations within a single county or contiguous counties and often within a brief time-span has been observed (especially in metropolitan areas) and remarked upon, usually in a critical way, by students of local government" (12)--yet the cause of such clustering was not known. Writing more recently, Smith and Debbage (2006) also identified distinct clustering patterns and similarly conclude that more detailed case studies of clustering are needed to further "shed light on the processes" (p. 110) that lead to the clustering. "Clearly," they conclude, "more research is needed to better understand the complex geographic clustering effects" (p. 118).
Why should scholars and practitioners care about the clustering of new municipalities? By creating new political entities (municipalities) in a restricted geographical landscape, NIMs (Newly Incorporated Municipalities) have profound impacts on governance structure, taxes, services, and elections (Miller, 1981; Jonas, 1991; Cox & Jonas, 1993; Burns, 1994; Foster, 1997; Musso, 2001; Rusk, 2003; Waldner, Rice, & Smith, 2013). NIMs can enhance local control, improve local government services, and provide valuable opportunities for new interagency alliances. Tieboutian supporters of public choice defend the right of residents to determine the governance that best suits their preferences, leading to new municipalities (Tiebout, 1956; Ostrom, Tiebout, & Warren, 1961).
NIMs can also dramatically fragment local political geography. As will be evident through the case studies, several new NIMs within a single county can profoundly alter the local and regional governance structure, re-wiring power structures, sharply reducing county revenues through lost taxes, and cutting the county out of land use decisions. Additionally, NIMs often adopt stricter land use regulations than the county, potentially exacerbating future class and racial segregation (Cox & Jonas, 1993). Metropolitan reformers further criticize the duplication of services and unnecessary competition for limited resources among a growing number of local government entities in a limited geographic area (Purcell, 2001).
For 50 years, scholars have known that new cities tend to form in clusters (Stauber, 1965; Schmandt, 1961), though they have been unable to explain the phenomenon. By pairing an empirical analysis with four rigorous case studies, we demonstrate that new city clusters are a primarily urban phenomenon and that counties with NIM clusters are markedly different from single-county NIMs (e.g., faster-growing, more densely populated, etc.). Contrary to all prior research, we also find that urban new city clusters do not form to ward off annexation attempts by existing cities. Rather, urban new city clusters form due to deep dissatisfaction with the county--a factor previously unidentified in the literature. Intense intergovernmental strife causes NIMs to form and secede from the county government. For the first time, we also identify cluster development mechanisms, finding clear evidence that pioneer NIMs--the first NIMs in the cluster--directly mentor and reduce transaction costs for subsequent incorporations.
Why new cities form in clusters proves to be a public policy matter of great import, rather than a simple mystery of administrative geography. In Fulton County and the other cases examined, mass defection from the county severely eroded county tax bases, created difficult-to-serve "orphans" (pockets of unincorporated land, generally low-income or minority), and radically reshaped the political and institutional landscape of the counties, among other effects. NIM clusters can have profound impacts on the counties in which they arise.
WHY NEW CITIES CLUSTER: PRIOR EXPLANATIONS FROM THE LITERATURE
Municipal incorporation creates a new city, town or village from a previously unincorporated community. In the United States, the unincorporated community is part of a county. Before incorporation, the county typically governs the community's affairs in many states and may provide services to the community. After incorporation, the new city assumes governance responsibilities such as land use planning, though the new city may contract with the county or private companies for services, such as water, sewer, fire, and police (Miller, 1981). Lands not part of the new city remain unincorporated and county-governed. In a few states, a system of townships prevails and no unincorporated land exists.
Rice, Smith, & Waldner (2014) found 22 reasons that stimulate new city formation, ranging from the desire to become eligible for local government grants to dissatisfaction with county governance. In order of frequency, the 5 most common reasons for municipal incorporation include: (1) to avoid annexation from a nearby city; (2) growth control and/or land use concerns; (3) desire to protect rural character or community identity; (4) desire to enhance services; and (5) revenue control. Other factors that played a significant role included dissatisfaction with the county government, the desire to gain eligibility for local governmental funds, economic development motives, racial factors, and the desire for enhanced political clout.
Local government boundary change scholars provide several explanations for why individual communities incorporate: (1) to avoid annexations (Stauber, 1965; Miller, 1981; Mumphrey, Wildgen, & Williams, 1990; Rigos & Spindler, 1991; Smith, 2011); (2) to increase or decrease service levels (Stauber, 1965; Teaford, 1979; Miller, 1981; Musso, 2001); (3) to stop land-use change (Fischel, 2001); (4) racial and/or socioeconomic exclusion (Danielson, 1976; Teaford, 1979; Miller, 1981; Weiher, 1991; Burns, 1994; Alesina, Baqir, & Hoxby, 2004; Smith & Debbage, 2011); (5) tax reduction (Miller, 1981; Burns, 1994); (6) tourism enhancement (Stauber, 1965); and (7) population growth or suburbanization (Schmandt, 1961; Wood, 1961; Stauber, 1965; Burns, 1994; Waldner, Rice, & Smith, 2013). Other explanatory factors include the regulatory rigor or lack thereof of state incorporation and/or annexation statutes (Martin & Wagner, 1978; Nelson, 1990; Rigos & Spindler, 1991; Liner & McGregor, 1996; Smith, 2011). Incorporation also results from political entrepreneurship by parties (such as chambers of commerce) benefitting from NIMs (Marando, 1974; Schneider & Teske, 1992; Burns, 1994; Feiock & Carr, 2001; Musso, 2001; Tkacheva, 2008).
Beyond the stimuli above, incorporations also may represent a deep-seated desire for democratic governance and direct representation. According to Alexis de Tocqueville (1835), a strong sense of place and civic responsibility are two hallmarks of a successful democracy. Likewise, the incorporation of a new municipality may also represent the truest form of Jeffersonian-style grass roots democracy. As Lazega and Fletcher (1997, p. 216) note, "Incorporation conjures images of Mayberry--local folks taking care of local problems. It is government on a first-name basis--close, personal, and responsive." As Jefferson surmised, decentralized local governments may better serve a homogeneous community's interests, as the elected officials are directly accountable to voters (Lazega & Fletcher, 1997). Viewed in this light, the advancement of democratic governance in the United States is alive and well, as communities incorporate behind a shared geographic identity. However, other political theorists have historically expressed concern that small municipalities may be too small to be truly democratic (Martin, 1957), or that "following the theories of James Madison, their populations tend to be too homogeneous and too intolerant of ... ideas which do not conform to the majority viewpoint, and hence are undemocratic" (Stauber, 1965, p. 29).
Yet why do clusters of new cities form? None of those explanations above account for the clustering of NIMs seen in the United States, only the formation of individual cities. Why would citizens in a given metropolitan area or county experience the creation of a multitude of new cities in geographic proximity within a short timeframe? Only one explanation has been proffered for clusters--annexation. Smith and Debbage (2006), examining the Southern United States, found that more than half of Southern NIMs form in a county where at least one other NIM formed. They hypothesized that "this clustering effect has emerged partly in response to the aggressive annexation tactics of nearby larger cities" (6). They note that "The clustering of NIMs in specific counties can be partially explained by a 'herd mentality' where a local political culture is...