Abstract
This article provides suburban scholars with a starting point for considering how municipal incorporation contributes to suburban inequality. I conduct an exploratory empirical analysis of incorporation in 2010 and find that incorporated suburbs are less racially diverse, Whiter, and have smaller shares of Black, Latinx, and Native American residents than unincorporated suburbs, suggesting that incorporation and its related municipal powers enable greater racial exclusion than strategies available to unincorporated suburbs. However, incorporated suburbs vary, and racial exclusion is most apparent in suburbs incorporated during and after the postwar suburban boom. Further, recently incorporated suburbs are more economically exclusive than unincorporated suburbs. I end by calling for greater integration of incorporation into suburban inequality research.
American suburbia is a racially and economically unequal landscape. What mechanisms have led to this? To answer this question, many scholars draw attention to the fact that suburbanization is characterized by “the fragmentation and proliferation of local governments” (Kruse and Sugrue 2006, 5). Although this is accurate, if we focus only on the number of local governments as the outcome of suburban political fragmentation, we miss another important layer of suburban political difference: whether a suburb is governed by a municipal government at all. In suburbia, a municipal government—the type of local governance often heralded as the hallmark of American democracy—is not a given. One in three Americans lives in an unincorporated community without a municipal government, and in some metropolitan areas, such as Atlanta and Philadelphia, the majority of suburbanites do (Cohen 2015).
Municipal incorporation—the creation of a new city, town, or village government—has critical implications for both our research analyses and theories about the mechanisms of suburban inequality. Perhaps the most common mechanism identified by a wide array of scholarship on suburban spatial inequality is zoning (Danielson 1976; Frug 2006; Rothstein 2017; Rothwell and Massey 2010). Zoning, however, is by and large a tool only available to incorporated suburbs. Incorporated communities differ fundamentally from unincorporated communities in the spatial tools and strategies they can harness and their level of political power and autonomy. These differences raise both theoretical and practical questions. Theoretically, how should we understand incorporation and its impact on suburban landscapes? Practically, how should we account for incorporation in our suburban studies?
I examine municipal incorporation as a possible mechanism of racial and economic inequality in suburbia. Previous scholarship on incorporation clearly demonstrates that racial and economic exclusion are primary motivations and goals of incorporation (Danielson 1976; Fischel 2015; Musso 2001; Rice, Waldner, and Smith 2014). However, although exclusion may motivate incorporation, is it in fact an effective strategy? We do not know the short- or long-term consequences of incorporation on racial and economic exclusion, and complicating the matter is that incorporation is just one strategy among many that suburban communities have at their disposal to accomplish exclusion. Is incorporation a more effective strategy than others? I conduct an exploratory empirical analysis to begin to provide clarity on these questions. The analysis sheds light on, first, the extent of incorporation and its geographical and temporal variation and, second, whether incorporation is an axis of racial and socioeconomic inequality between suburbs. To accomplish the latter, I examine the association of both incorporation status and timing of incorporation with the racial and socioeconomic compositions of suburbs, providing insight into the effectiveness of incorporation at accomplishing exclusion relative to strategies available to unincorporated suburbs.
I find evidence that incorporated suburbs are more racially exclusionary than unincorporated suburbs. They are less racially diverse, more White, and have smaller shares of Black, Latinx, and Native American residents. This exclusion is greatest among recently incorporated suburbs. Recently incorporated suburbs are also more economically exclusive, but suburbs that have long been incorporated are often less exclusive relative to unincorporated suburbs. Together, these findings suggest that—in line with the goals of the mostly White suburbanites who harness it—incorporation is an effective tool of suburban racial and economic exclusion. This work provides suburban scholars with a starting point for considering how incorporation may be integrated into their own work.
THEORIZING INCORPORATION
Municipal incorporation is not a common topic of study among social scientists (Leon-Moreta 2015a; Rigos and Spindler 1991; Smith 2018); however, when it is examined it is usually through the theoretical framework of the Tiebout public choice model. First articulated by the economist Charles Tiebout in 1956, the Tiebout hypothesis uses an economic lens to argue that municipal incorporation is a function of heterogeneity in preferences for public service provision in the population (Tiebout 1956). In the Tiebout model, residents must choose between a variety of local jurisdictions that vary in their levels of tax rates and service provision, and their preferences are revealed through their choice in living location. In other words, they “vote with their feet.” Local jurisdictions compete with one another for residents, and within this environment residents sort into groups that are relatively homogeneous in regard to service and tax preferences, creating an efficient allocation of public goods in which every resident maximizes their preferences through their living location and every jurisdiction only provides that which its residents want. The “best” number of jurisdictions is the number that allows for these efficient outcomes, so for more heterogeneous populations more jurisdictions will be needed (Dowding, John, and Biggs 1994; Musso 2001).
This theory’s predictions are complicated by decades of empirical scholarship that cast doubt on the accuracy of some of its core assumptions. First, the Tiebout model assumes absolute mobility of all residents; that is, any resident could move to any jurisdiction in their region (Dowding, John, and Biggs 1994). Empirical work on residential preferences and moves, however, show that the U.S. housing market is shaped by systems of capitalism and racism such that people of color and the poor may not be able to move to their preferred jurisdiction—or that their preferred jurisdiction has worse services and amenities than they would like (Dantzler, Korver-Glenn, and Howell 2022; Massey and Denton 1993; DeLuca and Jang-Trettien 2020). Second, the Tiebout model lacks a relational understanding of places. Though municipalities are theorized to compete for residents through service provision, the framework does not account for the relational nature of local suburban development. For instance, the creation of wealthy White suburbs and disadvantaged Black suburbs is not independent because “value for white spaces is predicated on the devaluation of Black places” (Purifoy and Seamster 2021, 6). According to the Tiebout model, these two communities represent “efficient” outcomes wherein White and Black residents have maximized realization of their “preferences” rather than an outcome in which municipal lines are drawn in the service of (re)producing White advantage.
To expand our knowledge, I argue that municipal incorporation is fruitfully understood through a theoretical framework grounded in an understanding of race and space as mutually constitutive and a recognition that spatial strategies and resource hoarding are core elements of the production of White advantage. Scholarship on the spatialized nature of race (Gilmore 2002; Pulido 2006; Cheng 2013; Lung-Amam 2017) provides a useful lens through which to understand municipal incorporation. Spatial strategies have always been critical to the production of racial categories and White domination (Loewen 2018; Massey and Denton 1993), and race structures the production of space (Gilmore 2002; Mills 1997). Space is never racially neutral but instead, as Laura Pulido argues, “a resource in the production of White privilege” (2000, 30). Through this frame, suburbanization is best conceptualized as a process of spatial inequality production (Douds 2021), and incorporation can be understood as one strategy among several for accomplishing that goal. However, domination is never complete (Jung 2015), and incorporation is also at times co-opted as a strategy of resistance, as discussed in the following section.
THE PURPOSES OF INCORPORATION
Here I provide an overview of previous scholarship on incorporation and argue that, despite ample work showing the exclusionary motivations behind incorporation, we have a surprising lack of empirical knowledge regarding its outcomes.
Municipal Incorporation
In the United States, municipal incorporation involves the creation of a new city, town, or village from a previously unincorporated area. The methods and requirements for a community to incorporate itself into a new government are set by states, and requirements vary widely (Oakerson 1987). Most states require minimum population sizes, often between 150 and five hundred; other common requirements are minimum population density, minimum distance from existing municipalities, and minimum tax base (Rigos and Spindler 1991). Once those requirements are met, a community petition, an election, and state certification are often required (Hill 1993). Incorporation is distinct from annexation, which occurs when a previously incorporated municipality adds new land area to its existing city limits. Secession, which occurs when part of an incorporated municipality separates from the municipality, may result in a new incorporation or the unincorporation of the seceding community (Smith 2018).
The powers of municipalities vary by state, but all incorporated municipalities in the United States have the powers of general government, economic development, physical environment, and culture and recreation. The majority of states also give municipalities powers over public safety, public works, and public health (Krane, Rigos, and Hill 2001). Counties serve all these functions for unincorporated areas. Thus, incorporation usually means that a community takes on responsibility for these functions itself, though in some cases, communities contract with the county to continue providing them (Smith 2018).
Most research to date on incorporation has focused on the causes of and motivations for incorporation. This research identifies several factors that predict incorporation—such as trying to avoid annexation by a central city and dissatisfaction with county services (Leon-Moreta 2015b, 2015a; Rice, Waldner, and Smith 2014; Rigos and Spindler 1991)—but one is identified repeatedly across studies ranging in time frame and geographic scope: in the majority of cases, incorporation is wielded by White suburbanites as a strategy for racial and economic exclusion and resource hoarding. In other but rarer cases, incorporation is also used by communities of color as a strategy of self-determination.
Incorporation as a Tool of Exclusion and Resource Hoarding
Research clearly establishes that, although many factors affect whether communities incorporate, the main motivation for suburban incorporation is the ability to control land use, including zoning, to enact exclusionary regulations (Danielson 1976; Fischel 2015; Musso 2001; Rice, Waldner, and Smith 2014). By being able to specify what can be built within their boundaries—including multifamily housing and minimum lot sizes—municipalities can effectively set a socioeconomic floor for who can access the community. Further, because incorporation also allows new municipalities to capture tax revenue that previously went to the county and had the potential for redistribution to other areas, it is also a mechanism of resource hoarding.
Examining postwar suburbanization, William Fischel argues that “zoning is an eminently political process, and it may be the most important municipal function in many communities. The most common reason for municipal incorporation of suburbs is to control zoning” (2015, 66). Rolf Pendall, Robert Puentes, and Jonathan Martin note that among suburbs is a “surprisingly standard template for land use planning” (2006, 2), and critical to these regulations is minimum lot sizes, a nearly ubiquitous practice (Gray and Furth 2019; Boudreaux 2016) that is a “workhorse of suburban zoning” (Fischel 2015, 30). Although counties can institute land use regulations in many states, including over unincorporated areas, their regulations are often much looser than those found in incorporated suburbs because their incentive structures favor denser development (Fischel 2001).
Zoning emerged in importance as other exclusionary tools became outlawed. Early suburbs used racially restrictive zoning and racial covenants to maintain the Whiteness of their communities (Wiese 2005). However, the Supreme Court struck down racially restrictive zoning in 1917 and racial covenants in 1948 (Fischel 2015). Nonracial zoning, however, had been legitimated by the Supreme Court in 1926, so from then on, enacting zoning regulations to exclude multifamily and smaller, more affordable housing became a direct way to exclude the poor and an indirect way to exclude people of color (Burns 1994; Schuetz 2008; Teaford 1979). The right of municipalities to enact zoning has rarely been successfully challenged (for an account of a rare legal decision to limit local zoning abilities in New Jersey, see Massey et al. 2013). Examination of which communities incorporate highlights their exclusionary aims: studying the thirty-two municipalities formed between 1950 and 1970 in California, Gary Miller (1981) finds that twenty-eight included Black populations of one or zero people. Further, only 10 percent of newly incorporated communities from 1990 to 2010 had minority White populations (Smith and Waldner 2018). Incorporation is also more likely in socioeconomically or racially diverse counties, suggesting its exclusionary motives to facilitate racial and economic segregation between municipalities (Alesina, Baqir, and Hoxby 2004; Lazega and Fletcher 1996; Leon-Moreta 2015b, 2015a; Musso 2001; Purcell 2001)
The process of racial exclusion through incorporation to enact zoning is apparent in the history of Black Jack, Missouri. Black Jack was a largely White, unincorporated community outside St. Louis. When a proposal to build racially integrated moderate- and low-cost housing in the community was unveiled in 1969, White residents responded by incorporating and then instituting a zoning ordinance prohibiting the construction of more than three homes per acre. This restriction made the proposed lower-cost housing impossible. Local pro-integration groups brought a lawsuit against the community and eventually won. The federal appeals court wrote, “The uncontradicted evidence indicates that, at all levels of opposition, race played a significant role, both in the drive to incorporate and the decision to rezone” (Rothstein 2017, 126). Despite winning the lawsuit, funding was no longer available by the time court battles ended and the housing was never built.
Through zoning, incorporation enables not only exclusion of those without but also hoarding of resources within. As Richard Hill argues, “in the context of a fragmented system of governments in the metropolis, municipal government becomes an institutional arrangement for promoting and protecting the unequal distribution of scarce resources” (1974, 1559). This occurs because incorporation allows municipalities to capture tax dollars that previously went to the county. When wealthy areas incorporate, they retain their tax dollars and keep them from being redistributed to poorer parts of the county. Although incorporated suburbs must provide their own services, such as police and water, incorporation usually results in lower tax rates as well as an increase in services (Frug 2006). As wealthy areas of counties politically separate themselves from poorer areas, this type of “revenue flight” can cause financial crises for the rest of the county (Lazega and Fletcher 1997).
Because explicit racial exclusion is illegal, incorporation-enabled zoning offers an indirect method based on socioeconomic resources. However, over the past several decades, American suburbs have been racially diversifying, driven in large part by middle-class and affluent people of color (Pattillo 2005; Lung-Amam 2017; Cheng 2013; Wiese 2005; Lichter, Thiede, and Brooks 2023). Given these new trends, zoning may not be as effective of a tool for racial exclusion as it once was, and increasing racial integration of many suburban communities provides suggestive evidence of this (Bader and Warkentien 2016; Zapatka and Tran 2023). That is, although zoning may continue to create economically exclusive communities, it may not effectively exclude people of color who are able to afford the housing in the community. If this is the case, then incorporation will not be associated with community racial composition or may only be weakly associated. On the other hand, the majority of poor Americans now live in suburbs (Kneebone and Berube 2013), and zoning regulations may effectively continue to segregate suburbanites across municipal boundaries along class lines.
Incorporation as a Tool of Autonomy and Empowerment
In contrast to the exclusionary goals of White communities—and, in many cases, as a response to them—some communities of color turn to incorporation as a tool of self-empowerment, autonomy, and place-making. Because these communities do not fit the dominant schema of White suburbanization, they have received much less scholarly attention, a feature of the suburban erasure L’Heureux Lewis-McCoy and his colleagues (2023) discuss in this issue. As White suburbanization increased in the 1940s and 1950s, Black communities increasingly turned to incorporation as a way to preserve existing communities or carve out new ones in areas beyond the central city (Wiese 2005). Residents of Lincoln Heights, Ohio, which began as a Black residential area for employees of a lumber company outside Cincinnati, took this approach, incorporating in 1946. Residents wanted the ability to self-govern and install needed infrastructure, including plumbing and paved roads. Despite gaining political independence, the suburb’s ability to thrive was stunted from the start because the county granted residents only 10 percent of the land originally proposed for incorporation, instead giving the local industrial tax base to a neighboring White suburb (Semuels 2015).
Many residents see these kinds of incorporations as a necessary response to the racist political economy of suburban landscapes. Ankur Goel and his colleagues write, “Neighborhood incorporation represents an opportunity for Black communities to exercise an amount of self-determination within American society… . The strategy seeks to undertake the unfinished business of the civil rights movement through group effort, community values and self-rule” (1988, 479). These communities are also sites of placemaking and “endurance, belonging, and resistance” (Hunter et al. 2016, 31). However, due to systemic racism embedded in the metropolitan spatial landscape, people of color do not reap the same benefits as White people from the same tools (see also Simms 2023, this issue). In her analysis of North Carolina, Danielle Purifoy (2019) finds that Black and Latinx residents do not receive the same benefits from incorporation—measured as proximity to amenities such as grocery stores and distance from disamenities such as landfills and brownfields—as Whites. Likewise, examining four cases of newly incorporated Black communities, Leora Waldner, Kristine Stilwell, and Russell Smith (2019) find that they faced problems not typically experienced by newly incorporated White communities, including contention with county governments, the need to raise taxes, and revenue shortfalls.
Despite their importance, incorporations by communities of color are much rarer than White community incorporations. Since 1990, only 10 percent of newly incorporated communities have been majority people of color (Smith and Waldner 2018). Incorporation remains a predominantly White tool.
Other Tools of Suburban Exclusion
Incorporating to enact zoning is not the only strategy for spatial exclusion available to suburbanites, and suburban residents acting through their elected representatives are not the only actors who may seek racially and economically segregated landscapes. Racial and class exclusion may be accomplished without incorporation through several other pathways. First, other forms of building restrictions, such as protective covenants and deed restrictions, can function similarly to zoning (Fischel 2015). Second, even though local homeowners are the actors of interest in theories regarding zoning and incorporation, other actors involved in suburbanization may profit from racial and economic segregation. Because U.S. housing values are affected by contextual factors, including proximity to the poor and people of color (Howell and Korver-Glenn 2020), developers have economic incentives to build segregated communities. New housing subdivisions often include homes within a narrow price range, and the “logic” of housing values incentivize developers to ensure that neighboring subdivisions are similar in price and racial composition (Hill 1974). Thus, even before residents move into a community, developers may build according to the segregated patterns that zoning eventually enshrines. Third, racism pervasive in all stages of the housing market also produces racial segregation across suburbs (Korver-Glenn 2018). Through these alternative pathways, it is possible that unincorporated suburbs accomplish the same levels of exclusion as incorporated suburbs.
RESEARCH QUESTIONS
Using current research, it is difficult to make clear predictions about the extent to which incorporation will associate with suburban racial and economic exclusion, for two reasons. First, surprisingly little research has been undertaken on the outcomes of incorporation, especially in the long term. Although exclusion may motivate incorporation, is the strategy effective and, if so, for how long? Black Jack, Missouri, the White St. Louis suburb that incorporated in 1970 to exclude Black residents, was 81 percent Black in 2019 (U.S. Census Bureau 2019). Is this kind of community trajectory common or an anomaly? Second, other strategies independent of incorporation enable communities to accomplish exclusion. There is little empirical research on these other strategies—particularly their scope and outcomes for racial and economic exclusion—leaving no basis on which to judge whether they are as effective or ubiquitous as those strategies available only to incorporated municipalities.
Is incorporation an axis of racial and economic exclusion in suburbs? Or, given other strategies, is it simply “an additional layer of already overdetermined racial inequity” (Purifoy 2019, 27)? My analysis is a starting point to answer these questions. I ask three questions to build our understanding of incorporation in suburbs and its role in inequality creation.
First, I examine the extent of municipal incorporation in suburbs geographically and temporally to establish the extent of suburban (un)incorporation.
Research Question 1: What is the rate of suburban incorporation, and do incorporation rates vary temporally and regionally?
Second, I examine incorporation as a source of inequality between suburbs. That is, I examine whether incorporation allows suburbs to exclude people of color and impoverished people more successfully than unincorporated suburbs.
Research Question 2: Is incorporation associated with the racial composition of suburban communities, and does this vary by date of incorporation?
Research Question 3: Is incorporation associated with the socioeconomic composition of suburban communities, and does this vary by date of incorporation?
In these questions, racial and socioeconomic exclusion is operationalized through composition measures. If incorporated suburbs are Whiter, have lower shares of people of color, or are socioeconomically advantaged relative to unincorporated suburbs, it suggests that incorporation and the strategies it enables are more effective tools of exclusion than those used by unincorporated suburbs. If there are no differences by incorporation status, either incorporated suburbs are not effective at accomplishing their goals or unincorporated suburbs are able to achieve similar outcomes in other ways. My analysis would not be able to distinguish between these two possibilities, but this information would inform future work. If incorporated suburbs have more people of color and fewer Whites or are socioeconomically disadvantaged relative to unincorporated suburbs, this would suggest that the exclusion strategies unincorporated communities use are more effective than those incorporated communities use.
DATA AND METHODS
Defining Suburbs
I define suburbs as metropolitan census places that are not the central city. Census places are the geography that best captures suburban communities as suburban residents understand them (Hall, Tach, and Lee 2016; Lichter, Parisi, and Taquino 2015). Census places include both incorporated places—cities, boroughs, towns, and villages—as well as census designated places, or unincorporated areas defined by the U.S. Census Bureau with input from local communities. Despite lacking a local government, unincorporated places “usually physically resemble incorporated places in that they contain a residential nucleus, have a closely spaced street pattern, and frequently have commercial or other urban types of land use” (U.S. Census Bureau 1994b, 1–2). For example, Sienna, Texas, is an unincorporated census designated place with twenty-two thousand residents (U.S. Census Bureau 2019). It is a master-planned housing development with shopping centers, schools, libraries, and other amenities similar to those of its neighboring incorporated suburbs. Census places may span across other geographies, including counties and county subdivisions.
Data
The primary analysis focuses on suburbs in 2010. Data on 2010 census place characteristics were obtained from decennial census tables provided by the IPUMS National Historical Geographic Information System (Manson et al. 2021). Characteristics not available through the 2010 Census short form were obtained from the 2008–2012 American Community Survey 5-year estimates.
Year of municipal incorporation was obtained from two data sources. First, I draw from the 1987 Census of Governments (COG), which asked all municipalities the date of their original incorporation (U.S. Census Bureau 1990). This was the only year that the COG included this question. I obtain information on municipalities that incorporated from 1987 to 2010 from the U.S. Census Bureau’s Boundary and Annexations Survey (BAS). Each year, every incorporated municipality is surveyed to obtain updated information on boundaries, names, and governmental status. BAS data tables are highly accurate for new incorporations since its inception in 1982 but do not include consistent information for municipalities incorporated before this. Thus, I use the BAS only to capture incorporations after the 1987 COG.1
Matching Places to Metropolitan Areas
To identify suburban census places, places must be matched to metropolitan areas. Metropolitan places were determined using a place-to-metropolitan-area crosswalk obtained from the Missouri Census Data Center Geocorr tool (MCDC Data Applications 2021). Among metropolitan places, suburban places are defined as places within a metropolitan area but not in the central city. Because some metropolitan areas contain more than one central city, such as the New York–Newark–Jersey City metropolitan statistical area (MSA), I define central cities as metropolitan places that are either the largest place within a metropolitan area (such as New York City) or have a population of more than two hundred thousand (such as Newark and Jersey City within the New York MSA). In 2010, 14,484 suburban census places were identified in all metropolitan areas.
Analytic Sample
I include all metropolitan areas in the analysis but exclude suburban places with populations of less than one hundred to ensure that the analysis is not unduly influenced by small places. Despite historical population requirements for unincorporated areas (typically at least one thousand), such thresholds were eliminated in 2000 (U.S. Census Bureau 2008). Further, all Hawaiian communities are dropped because Hawaii includes no incorporated communities (Cohen 2015). Finally, date of incorporation was not available for 1,199 incorporated municipalities; these places, which had a median population size of 978, were dropped from the analysis. These specifications lead to a final analytic sample for the main analysis of 12,709 suburban communities in 2010.
Measures
I describe creation of measures for the 2010 data, the primary data for my analysis. However, I also analyze data from 1970 to 2000 to assess temporal trends in incorporation. Descriptions of these data and measures are provided in the appendix.
Incorporated status is the primary independent variable in all analyses. In 2010, incorporated status was coded from the geographic field PLACECC. For incorporated municipalities, year of incorporation was recoded into eleven categories. Suburbs incorporated before 1910, the decade in which the first zoning ordinances were enacted (Fischel 2015), were coded together, then nine subsequent decades were coded, from 1920 to 1929 through 2000 to 2009. Unincorporated areas were coded as 0 for this variable and are treated as the reference category in regression models. In preliminary analyses, I used postestimation diagnostic tests to explore the possibility of combining decade categories to create a simpler measure, but, given the large number of outcomes examined, no single approach for collapsing was suggested by all models. Thus I have opted to use the noncombined measure for all models for consistency.
The outcomes for this analysis are racial-ethnic composition, racial-ethnic diversity, and socioeconomic composition. Racial-ethnic composition is measured as percentage Latinx, non-Latinx White, non-Latinx Black, non-Latinx Native American (combining Native American and Alaskan Native categories), non-Latinx Asian (combining both Asian and Native Hawaiian and Pacific Islander categories), and non-Latinx other race. Racial diversity is measured using the entropy index standardized to five groups: Latinx, non-Latinx White, non-Latinx Black, non-Latinx Asian, and non-Latinx Native American and other race. The index measures how evenly a suburb’s population is distributed across the five groups, ranging from 0, the population contains only one group, to 1, each group is 20 percent of the population (Hall, Tach, and Lee 2016). Suisun City, California, a Bay Area suburb, has a standardized entropy score of 0.96 and in 2010 was 24 percent White, 23 percent Black, 20 percent Asian, 25 percent Latinx, and 7 percent Native American and other race. Meanwhile, Huntington Park, California, a Los Angeles suburb, has a standardized entropy score of 0.07 and in 2010 was 97 percent Latinx, 2 percent White, 0.4 percent Black, 0.6 percent Asian, and 0.4 percent Native American and other race. The socioeconomic status of suburban places is captured through four measures: percentage of families in poverty, percentage of the population age twenty-five and older with a bachelor’s degree or higher, percentage homeownership, and median household income (in 2010 dollars).
In the multivariable models, population size and the percentage of housing built after 1970—the decade following the passage of civil rights legislation—are included as controls.
Analytic Plan
The analysis proceeds in the order of my research questions. To answer Research Question 1, I examine rates of incorporation by decade nationally and across time and regions to understand the extent of (un)incorporation. Research Questions 2 and 3 are concerned with incorporation as a source of inequality between suburbs and ask about the association of incorporation with the racial and socioeconomic compositions of suburbs. To answer these questions, I use multivariable regression models to delineate the relationship between incorporation and my outcomes of interest. In the first model, I use a dichotomous measure of incorporation:
where p indicates place and m indicates metropolitan area. Yp is the outcome for place p (such as percentage Black, percentage poverty), β0 is the intercept, and INCP is a suburb’s incorporated status. PP is the place population, and HP is the percentage of a suburb’s housing stock built after 1970. Metropolitan area fixed effects are indicated by δm. Metropolitan fixed effects account for the metropolitan scale of suburbanization and across-area differences. The error term is εp.
In the second model, I include the categorical measure of incorporation (INCYRP) as a set of dummy variables with unincorporated suburbs as the omitted reference group and the other groups indicating the decade that incorporated suburbs incorporated. This model provides insight into whether differences between incorporated and unincorporated suburbs vary by timing of incorporation:
RESULTS
Incorporation Nationwide and Regionally
I begin my examination of incorporation by assessing the phenomenon across time and regions to answer RQ 1. Figure 1 plots the number of all suburbs, incorporated suburbs, and unincorporated suburbs from 1970–2010 in the 100 largest metropolitan areas. I limit the visual to the 100 largest MSAs in each decade to provide a more standardized measure across decades, as the number of MSAs varies widely over time due to changing requirements for and definitions of metropolitan areas (U.S. Census Bureau 1994a). Though my primary analysis focuses on 2010, I include historical data here to understand temporal trends (for descriptions of data and measure construction for 1970 through 2000, see the appendix). Reflecting increasing suburbanization across this period, incorporated and unincorporated suburbs increase steadily until 2000, when both plateau. Incorporated suburbs are more numerous than unincorporated suburbs, but they compose a smaller portion of all suburbs over time. In 1970, 74 percent of all suburbs were incorporated, versus only 60 percent in 2010. In 2010, among all suburbs in all MSAs (not only the hundred largest) there were 7,494 incorporated suburbs in the primary analytic sample that were home to 82.9 million residents and 5,215 unincorporated suburbs home to 31.1 million residents. Nearly three in ten (27 percent) suburbanites live in unincorporated communities.
All, Incorporated, and Unincorporated Suburbs in Hundred Largest Metropolitan Areas, 1970–2010
Source: Author’s calculations.
I examine regional variation in suburban incorporation in 2010 in figure 2 (for tabular results, see table A.1). I use the eight regions defined by the Bureau of Economic Analysis (Bureau of Economic Analysis 2016). The rate of incorporation varies substantially across regions. Suburbs in the Plains and Great Lakes states have the highest rates of incorporation with the vast majority of suburbs incorporated (93 percent and 82 percent, respectively). Roughly six in ten suburbs are incorporated throughout the Rocky Mountains (56 percent), Southeast (57 percent), and Southwest (61 percent). In Mideast (Atlantic) states, roughly half (49 percent) of suburbs are incorporated; whereas the minority are in the Far West (39 percent) and New England (18 percent). The low rate of incorporation in New England is a result of their township system where towns, due to their largely rural nature, are not considered incorporated (Cohen 2015).
Percent of Suburbs Incorporated by Region and Year, 2010 (N = 12,709)
Source: Author’s calculations.
Year of Incorporation
So far, I have examined incorporation as a binary outcome—incorporated or not—but this masks variation within incorporated places based on timing of incorporation. Figure 3 presents the distribution of incorporated suburbs in 2010 by their year of incorporation. Incorporations began to grow in the early 1800s, remained high from the late 1800s until 1925, declined through the 1940s, peaked again in the 1950s, and have been declining in recent decades. The mean year of incorporation for suburbs in 2010 is 1911 (median=1907).
Distribution of Incorporated Suburbs by Year of Incorporation, 2010 (N = 7,494)
Source: Author’s calculations.
Note: Incorporations prior to 1700 reported at 1700 for visual ease.
Suburbanization did not, of course, occur evenly across place and time. Figure A.1 plots the timing of incorporation by region. Early suburbanization was concentrated in Southeast, Mideast (Atlantic), and Great Lakes states; more recent suburbanization is concentrated in Southwest and Southeast states. Once again, the Northeast displays low levels of incorporation relative to other areas because of the existing township system in which most towns are not considered incorporated (Cohen 2015).
Incorporation and Racial Composition
To answer Research Question 2, on the association between incorporation and racial composition, I estimate multivariable linear regression models predicting the racial composition outcomes. In the first model, I include a dichotomous measure for incorporated status as the key independent variable. In the second model, I use a categorical variable capturing incorporated status and decade of incorporation— unincorporated suburbs are the reference category. Logged population size and the percent of housing built after 1970 are included as controls, as are metropolitan area fixed effects. Table A.2 presents means and standard deviations of the racial composition outcomes by incorporation status.
Figure 4 displays regression coefficients with 95 percent confidence intervals for models with the dichotomous incorporation measure as well as for decade of incorporation for each of the seven outcomes. In all models, unincorporated suburbs are the reference category. For the dichotomous measure, the coefficients indicate the average difference between unincorporated and incorporated suburbs within the same metropolitan area while controlling for population size and housing stock age. For the categorical measure, the coefficients indicate the average difference between unincorporated suburbs and suburbs incorporated in the given decade within the same metropolitan area while controlling for populations size and housing stock age. Because of varying magnitudes of the associations across models, x-axis ranges differ slightly across models to conserve space.
Coefficients for Incorporation Status (Dichotomous and by Decade of Incorporation) from Linear Regressions Predicting Suburban Racial Composition, 2010 (N = 12,709)
Source: Author’s calculations.
Note: Diversity measured using five-group standardized entropy index. Models include controls for logged population size and percentage of housing built after 1970 as well as metropolitan area fixed effects.
Examining the model with the dichotomous measure of incorporation, incorporated suburbs are on average 2.8 percent Whiter than unincorporated suburbs. However, the dichotomous measure masks divergences by timing of incorporation. With the exception of suburbs incorporated in the 1930s and 1940s, incorporated suburbs are Whiter than unincorporated suburbs, but the size of the average difference varies by decade of incorporation. Suburbs incorporated prior to 1930 and in the 1950s are on average 2 to 3 percent Whiter, but after 1950 the size of the difference grows with each decade such that the most recently incorporated suburbs—those incorporated between 2000 and 2009—are 7.4 percent Whiter on average than unincorporated suburbs. Suburbs incorporated in the 1930s do not significantly differ from unincorporated suburbs in White composition, but, in contrast to the other decades, suburbs incorporated in the 1940s are on average 3.0 percent less White than unincorporated suburbs. The 1940s are exceptional across several of the models; I address this in the discussion.
The model results for the Black share of the population indicate that Black people make up higher shares of unincorporated communities on average, but differences are not present across all decades of incorporation. Using the dichotomous measure of incorporation, incorporated suburbs on average have slightly lower Black population shares—1.4 percent. The Black share of the population is significantly smaller in suburbs incorporated before 1910 (1.3 percent) and incorporated in the 1910s (2.1 percent). Suburbs incorporated in the 1920s and 1930s as well as the 1970s, 1980s, and 1990s do not significantly differ in Black composition. As in the White model, suburbs most recently incorporated in the 2000s feature the largest differences—4.4 percent lower Black share. Also as in the White model, suburbs incorporated in the 1940s buck the trend and on average have 2.0 percent higher Black shares than unincorporated suburbs.
Incorporated suburbs have lower shares of Native American residents on average (0.42 percent). However, when examining by decade of incorporation, differences are statistically significant only for the 1940s, 1970s, and 1980s. All coefficients, however, are negative, providing some evidence that Native American shares may be smaller in incorporated suburbs but fail to reach statistical significance due to their small average shares: they make up only 1.0 percent of suburban populations on average. In contrast, the share of the population categorized as other race does not vary significantly by incorporation status, either using the dichotomous measure or the measure including decade of incorporation.
Models predicting the Asian share of suburban communities show that, on the whole, differences between incorporated and unincorporated suburbs are not significant. Assessing by decade of incorporation reveals that Asians make up significantly higher shares in suburbs incorporated in the 1950s and lower shares in those incorporated before 1910. Unlike in the White and Black models, the most recently incorporated suburbs do not differ in their Asian population shares from unincorporated suburbs, suggesting that the Asian population is evenly distributed across unincorporated and newly incorporated suburbs.
The Latinx models reveal that on average Latinx population shares are 0.82 percent lower in incorporated suburbs than unincorporated suburbs, but examining differences by decade of incorporation reveals that differences are present primarily in postwar suburbs. Suburbs incorporated between 1950 and 1999 have lower Latinx shares than unincorporated suburbs, the size of the difference growing over time from 1.6 percent in the 1950s to 2.9 percent in the 1990s. Similar to the Asian models, suburbs incorporated in the 2000s do not differ from unincorporated suburbs in Latinx shares. Finally, as in the White and Black models, suburbs incorporated in the 1940s have larger Latinx population shares (2.0 percent) than unincorporated suburbs.
The final models examine multigroup racial diversity levels measured using a standardized entropy index. Model results indicate that, no matter how incorporation is measured, incorporated suburbs are significantly less racially diverse than unincorporated suburbs. Suburbs incorporated in the 2000s average 0.07 points lower on the diversity measure than unincorporated suburbs, though this is only roughly 40 percent of the standard deviation for the entropy measure. The diversity differences are consistent but small.
Together, these results provide evidence that incorporated communities are more racially exclusionary than unincorporated suburbs, as they have larger White shares and smaller Black, Native American, and Latinx shares and are less racially diverse. These differences are largest and most consistent in postwar suburbs, and the 1940s display an opposing trend of lower racial exclusivity. Shares of Asian and other race residents do not vary by incorporated status.
Incorporation and Socioeconomic Composition
I next turn to results from multivariable models predicting the socioeconomic composition of suburbs in figure 5. As in figure 4, x-axis ranges differ slightly across models to conserve space. Table A.3 presents means and standard deviations of the socioeconomic composition outcomes by incorporation status.
Coefficients for Incorporation Status (Dichotomous and by Decade of Incorporation) from Linear Regressions Predicting Suburban Socioeconomic Composition, 2010 (N = 12,709)
Source: Author’s calculations.
Note: Models include controls for logged population size and percentage of housing built after 1970 as well as metropolitan area fixed effects.
For most socioeconomic outcomes, suburbs that incorporated long ago are less economically exclusive than unincorporated suburbs, while suburbs incorporated since mid-century are more economically exclusive. However, at what decade of incorporation the association flips varies by outcome.
I first examine median household income. Comparing suburbs in the same metropolitan area and controlling for population size and housing age, incorporated suburbs on average have median household incomes $1,359 lower than unincorporated suburbs. However, the associations diverge when decade of incorporation is taken into account. Suburbs incorporated before and during the 1910s on average have median household incomes that are lower than unincorporated suburbs, whereas those incorporated after that—with the exception of the 1940s and 1960s—average higher median household incomes. The average difference grows across decades with the largest difference for most recently incorporated suburbs. For suburbs most recently incorporated in the 2000s, the average advantage is $7,428.
Poverty rates show a similar pattern. Overall, incorporated suburbs on average have higher poverty rates (0.43 percent), but this diverges across decade of incorporation. Suburbs incorporated before 1910 have higher average poverty rates than unincorporated suburbs (1.59 percent), but after that incorporated suburbs either average lower poverty rates or do not differ from unincorporated suburbs. Those most recently incorporated in the 1990s and 2000s have significantly lower poverty rates (–2.22 percent and –2.76 percent, respectively). Suburbs incorporated before 1910 have lower college degree rates than unincorporated suburbs, but those incorporated between 1920 and 1999 have a college degree advantage. The exact size of the advantage ranges from 2.5 percent for suburbs incorporated in the 1940s to 8.4 percent for suburbs incorporated in the 1930s.
Finally, incorporated suburbs have significantly lower rates of homeownership relative to unincorporated suburbs except for the most recently incorporated suburbs. Overall, incorporated suburbs average homeownership rates 4.6 percent lower than unincorporated suburbs. Suburbs incorporated after 2000 show a homeownership advantage of 3.9 percent. This pattern is quite striking, as one of the primary theorized mechanisms of exclusion through zoning is that suburban municipalities will limit the building of multifamily rental homes.
Supplementary Analyses
I also conducted a set of supplementary analyses to test the robustness of the multivariable model findings to additional specifications. First, I repeated the racial composition analyses with median household income added as a control. Results were substantively similar. Second, I reestimated the socioeconomic composition models with percent White, Black, Asian, and Latinx included as controls. Results were also substantively similar with the exception that differences in poverty rates between unincorporated suburbs and suburbs incorporated in the 1990s and 2000s were no longer statistically significant. This suggests a differential clustering of racial groups across suburbs with different poverty levels.
DISCUSSION
Municipal incorporation is a fundamental political division among American suburbs. In 2010, 73 percent of suburban residents had a local municipal government; 27 percent did not. Despite this great political divide, suburban scholarship has paid little attention to incorporation. Given that suburbanization is a process of spatial inequality creation (Douds 2021), it is critical that we understand what if any role incorporation plays as a mechanism in these inequality processes as well as how it relates to and interacts with other mechanisms. In this article, I provide a starting point for suburban scholars to integrate incorporation into their study of suburban inequality by offering an exploratory analysis of the connection between incorporation and racial and economic exclusion. Results indicate that incorporation is exclusionary not only in its motivations but also in its outcomes; it is a critical but understudied mechanism of racial and economic fragmentation in American suburbs.
I find that incorporated suburbs are Whiter and have smaller shares of Black, Latinx, and Native American residents than unincorporated suburbs. They are also less racially diverse. These results suggest that incorporation enables greater racial exclusion than strategies available to unincorporated suburbs. However, incorporated suburbs vary, and racial exclusion is most apparent in those suburbs that incorporated after the postwar suburban boom. During this period, as people of color began suburbanizing at greater rates (Frey 2014), Whites increasingly turned to incorporation and the zoning it enabled to maintain racial segregation. It is possible that postwar suburban incorporations were more exclusionary in design than previous incorporations, though historical accounts provide strong evidence for exclusionary aims of incorporations long before the war (Wiese 2005). It is also possible that exclusionary goals are attained in the short term but in the long term are harder to maintain. Or changes in regional or local populations, political leadership, or policies may change the goals of the community altogether. Longitudinal data on incorporation status and demographic composition are needed to help distinguish between these possibilities.
Similar to the racial composition results, I also find that more recently incorporated suburbs are more economically exclusionary than unincorporated suburbs, but suburbs incorporated long ago either do not differ or are less economically exclusionary. The exact timing of when this switch occurs varies across socioeconomic indicators. Once again, my analysis cannot tease out whether newly incorporated communities are more exclusionary because of temporal trends in exclusionary practices or waning efficacy over time. Research harnessing historical data could shed more light on this.
An exception to these general conclusions is that suburbs incorporated in the 1940s have higher shares of Black and Latinx residents and lower shares of White residents than unincorporated communities. My analysis cannot precisely parse the causes of this finding, but historical trends provide some context. The 1940s witnessed a wave of Black suburban incorporations as many Black people with rising incomes moved North in the Great Migration and as Whites began more aggressively carving out suburban space through incorporations of their own (Wiese 2005). When read with this history in mind, the analysis results may indicate that the Black suburbs incorporated in the 1940s have remained more welcoming of people of color than White-incorporated suburbs from that time. The legacy of these communities is recognizable today.
This analysis provides a foundational portrait of incorporation in the United States at a particular point in time—2010—and compares incorporated and unincorporated communities at an aggregate level. Examining incorporated suburbs by decade of incorporation provides greater understanding of variation within incorporated suburbs, but no corresponding measure exists for unincorporated communities. This category groups together long-standing communities that have had social identities for over a century with newly built suburban developments and many communities in between. Despite this aggregation, these communities have in common a political status—unincorporated—and this analysis provides an informative foundation for understanding the ways that incorporated and unincorporated suburbs differ to provide insight into processes of exclusion in suburbs.
Further, although the analysis conducted here indicates that, in the aggregate, incorporation has exclusionary outcomes, the study of incorporation must grapple with this political tool in its full context of a complex and overlapping system of local governance structures and political powers. The Woodlands, Texas, a majority-White suburb of Houston, highlights this complexity. It is a highly affluent, exclusive community but, unlike most other affluent White suburbs of the region, is not incorporated; instead it is classified as a special purpose district. While many suburbs incorporate to lower taxes and enact exclusionary zoning policies, The Woodlands has accomplished these results without incorporation. When incorporation was put on the ballot in November 2020, the community rejected the initiative largely on the grounds that it would raise local taxes (Vasquez 2021). That incorporation would raise taxes indicates that, rather than having its tax dollars redistributed to lower resourced parts of the county, the community has managed to capture more than its fair share of county tax dollars with its current arrangement. Further, despite lacking the power to zone, the community is governed by a tight web of strict deed restrictions that have a similar effect. Cases like The Woodlands highlight the complicated and creative use of local governance structures for exclusion and resource hoarding in suburbia.
Given my findings, I suggest that incorporation should be integrated into theoretical accounts of suburban inequality, particularly those related to spatial or political dynamics, and that scholars should account for incorporation in empirical analyses. To aid in these efforts, I have made a combined and cleaned dataset containing year of incorporation for all U.S. municipalities available through the Harvard Dataverse (Wyndham-Douds 2022). Because it is the background political terrain over which other suburban processes play out, scholars should consider how incorporation may affect the arenas that they study, such as residential segregation, schools, housing, or health.
Beyond integrating incorporation into studies of other suburban phenomenon, suburban scholars should also focus more on incorporation itself. We still have much to learn about incorporation processes and outcomes, as well as how incorporation operates in relation to other exclusionary measures and jurisdictions. Smaller-scale, in-depth qualitative investigations of communities could shed light on both how incorporations occur and precisely how mechanisms available to incorporated communities, including zoning, are leveraged (see Girouard 2023). Investigations on this scale could also reveal how local actors, including residents, local government officials, and developers, affect these processes. Finally, future research should examine the outcomes of strategies available to incorporated suburbs relative to unincorporated suburbs in more detail with a particular focus on jurisdictional arrangements that some consider alternatives to incorporation, such as special districts (Lazega and Fletcher 1997). Suburban inequality is accomplished through several overlapping, reinforcing, and potentially duplicate mechanisms, and identifying interventions to increase racial and economic equity requires understanding how all these elements work together.
APPENDIX: DATA SOURCES AND MEASURE CREATION FOR 1970–2000 DATA
The primary analysis focuses on 2010, and data sources and measures for that year are discussed in the main text. Figure 1 displays temporal trends that also include 1970 through 2000. I describe data sources and measure creation for those years in this section. Data on census place characteristics for the period were obtained from decennial census tables provided by the IPUMS National Historical Geographic Information System (Manson et al. 2021).
Matching Places to Metropolitan Areas
To identify suburban census places, places must be matched to metropolitan areas. Places in 1970 and 1980 were matched to metropolitan areas by intersecting 1970 and 1980 NHGIS place point GIS files with SMSA GIS boundaries for their respective decades (Manson et al. 2021). Place points that fell within the boundary of an SMSA were considered metropolitan. For 1990 through 2000, metropolitan places were determined using place-to-metropolitan-area crosswalks obtained from the Missouri Census Data Center Geocorr tool (MCDC Data Applications 2021). Among metropolitan places, suburban places are defined as places that are within a metropolitan area but are not the central city. Because some metropolitan areas contain more than one central city (such as the New York–Newark–Jersey City MSA), I define central cities as metropolitan places that are either the largest place within a metropolitan area (such as New York City) or have a population of greater than two hundred thousand (for example, Newark and Jersey City within the New York MSA). From 1970 to 2000, 35,947 suburban census places were identified in all metropolitan areas.
Analytic Sample
When examining temporal trends (see figure 1), I limit the analysis to the hundred largest metropolitan areas and suburban places with populations over one thousand. This population limit ensures that incorporated and unincorporated suburbs are comparable over time, as unincorporated areas were required to have a population of at least a thousand in the 1970 to 1990 period in most cases (U.S. Census Bureau 1994b). Population thresholds were eliminated in 2000 (U.S. Census Bureau 2008). Further, all Hawaiian communities are dropped because Hawaii contains no incorporated communities (Cohen 2015). Finally, date of incorporation was not available for 1,350 municipality-year observations; these places were dropped from the analysis. These specifications lead to a final analytic sample for the temporal trends of 26,231 suburban communities from 1970 to 2010.
Incorporated Status
Incorporated status is the primary independent variable in all analyses. In 1970, incorporation status was determined from the geographic field PLACEDES, which included a category of “unincorporated place.” In 1980, incorporation status was determined from the place name. Places with “CDP” (census designated place) in their names were coded as unincorporated; all other places were coded as incorporated. For 1990 through 2010, incorporation status was coded from the geographic field PLACECC.
Incorporated Status of Suburbs Across Decades, Nationally and by Region
Racial Composition and Diversity of Suburbs by Incorporation Status, 2010
Socioeconomic Composition of Suburbs by Incorporation Status, 2010
Distribution of Incorporated Suburbs by Year of Incorporation and Region, 2010 (N = 7,494)
Source: Author’s calculations.
FOOTNOTES
- © 2023 Russell Sage Foundation. Wyndham-Douds, Kiara. 2023. “Suburbs, Inc.: Exploring Municipal Incorporation as a Mechanism of Racial and Economic Exclusion in Suburban Communities.” RSF: The Russell Sage Foundation Journal of the Social Sciences 9(2): 226–48. DOI: 10.7758/RSF.2023.9.2.10. I thank the editors, co-contributors, and reviewers of this issue for valuable feedback and Michael Hout for providing advice on the analysis. Direct correspondence to: Kiara Wyndham-Douds, at wyndham-douds{at}wustl.edu, 1 Brookings Drive, Washington University in St. Louis, St. Louis, MO 63130, United States.
Open Access Policy: RSF: The Russell Sage Foundation Journal of the Social Sciences is an open access journal. This article is published under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.
REFERENCES
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.