In the preface to his ground-breaking City Limits, Paul Peterson declared that “urban political analysis has been removed once again to the periphery of the discipline.”Footnote 1 At the time Peterson wrote, many political scientists could recall a not-too-distant past when urban scholars and their work represented mainstream political science. Peterson expressed hope that it would “be possible for the urbanists of the eighties to address the central concerns of political life with the vigor and conviction with which it was possible to address them two decades earlier.” Nearly 25 years later, Dennis Judd evaluated the field's success in this endeavor in a 2005 Urban Affairs Review article. Judd disappointingly concluded that “urban politics has continued to occupy an uncomfortable space at the margins of the discipline.”Footnote 2 Its status remains the same in 2009.
Some urban scholars suggest that the field's disconnection from mainstream political science is a good thing, allowing urbanists the freedom to pursue more meaningful questions and nuanced answers.Footnote 3 But other scholars, intent on rectifying the field's peripheral position, have proposed a variety of reasons for their maligned status. Perhaps most importantly for explaining the ghettoization of the field is the failure of many urban scholars to connect their questions and findings to more general political science phenomena or to utilize theoretical developments from mainstream political science.Footnote 4 Scholars like Harold Gosnell, Robert Dahl, Edward Banfield, James Wilson, and Theodore Lowi once successfully engaged the broader political science community through their city politics research; today a great deal of urban politics work focuses on debates that are simply not central to the wider discipline.
Another reason that urban politics remains estranged is a lack of attention to social scientific methods. Over the last half-decade, political science moved towards a more quantitative discipline while the field of urban politics remained largely qualitatively oriented and focused on single cases. In part as a result of this methodological approach, many scholars have been unable (or unwilling) to adequately address counterclaims and alternative explanations in their work.Footnote 5 Worse, as Judd argues, some urban politics scholarship can be characterized as ideologically intolerant, hyperbolic, and overly normative.Footnote 6
A final reason that urban politics has become less central to political science is that cities themselves no longer seem as important to political life. Political affairs and media attention have become increasing nationalized and there is no widespread sense of an impending “urban crisis.” Today most of the population resides outside of the central city and many cities look demographically and politically distinct from the nation as a whole.Footnote 7 Furthermore, some political science scholars have come to believe that no important or interesting politics happens in cities—the federal system severely constrains city decision making; whatever freedom cities are granted by states and the national government is hampered by the drive to compete for development and population with other localities; turnout is shockingly low as are interest and knowledge of local politics; and numerically speaking individual local policy decisions affect an insignificant number of people relative to state or national policies.
Although each of these explanations probably has some truth to it, I do not attempt to adjudicate among these propositions. Instead I lay out a rationale for studying local government from a number of different perspectives. Scholars of city politics have offered political science an enormous amount of insightful and careful work. But the reality is that today, many scholars are discouraged from studying the local level or are insufficiently trained in the local context to effectively use the data. It is quite possible to get an undergraduate degree (or a Ph.D.) from many political science programs without knowing the first thing about city politics. Indeed, many political science departments do not have a faculty member who is regularly engaged in research on local governments, nor is this viewed as a particular deficit. So I attempt here to persuade scholars (both established and emerging) of the value in studying what happens in America's smallest political arenas.Footnote 8
Before doing so, I need to clarify what I mean by the study of local politics. I include in this category analysis of all sub-state units (e.g., cities, suburbs, rural areas, counties, school districts, etc.). While I use the terms “urban” and “city” throughout this essay, the term “local politics” is probably the most accurate. Further, I think of local politics as both a dependent and an independent variable. For instance one might study political behavior directed at local policy (city politics as an outcome) or alternatively analyze the effect of local institutions on political outcomes like presidential turnout (city politics as a causal force). Similarly, some local politics work analyzes the ways in which cities affect the world beyond their borders, while other work studies how the world outside of the city affects the development and practice of politics within city limits.Footnote 9
So then, what makes these local factors worth studying? Fundamentally, local politics is similar to and different from politics at other levels of American government. For both reasons scholars should be concerned with reinserting the subfield into mainstream political science. Because politics in cities is at once distinctive and analogous to politics more generally, studying local politics can provide insights for scholars into the functioning of our political world. I offer three related arguments in support of these claims. First, the local level is the source of numerous political outcomes that matter because they represent a large proportion of political events in the United States. Second, there are methodological advantages to studying local politics. Finally, analyzing politics at the sub-state level can generate thoroughly different kinds of questions than a purely national-level focus and can offer quite different answers to questions that apply more generally. I address each of these arguments in turn.
Local Politics Are American Politics
Studying local politics ought to be integral to the study of political science both because local politics is, in and of itself, important and because local contexts shape state and national politics. In the United States a large proportion of political activity occurs at the sub-state level. The vast majority of elected officials are local legislators. In 1992, the most recent year that the Census Bureau collected data on local legislators, the Census counted 342,812 members of local governing boards representing 84,955 different local governmental units, compared to the federal level where there were 535 legislators serving in one governmental unit, and 7,382 state legislators in the 50 states.Footnote 10 In many places local jurisdictions also hold elections more frequently than states or the federal government. This means that most elections in the United States are local elections, most campaigns are local campaigns, and in some cases, most votes are local votes.Footnote 11 One implication of this structure is that understanding the quality and functioning of American democracy is impossible without attention to local elections and legislatures.Footnote 12
Local decisions and policies also account for a large and growing proportion of total government activity. Of the 21,039 public employees counted by the Census in 2002 approximately 13 percent were employed by the federal government, 24 percent by the states, and 63 percent by local governments.Footnote 13 Approximately one-quarter of all governmental revenues and expenditures were local revenues and expenditures in 2001.Footnote 14 The tremendous variation in institutions across local governments means that policy outcomes can be highly varied in different contexts.Footnote 15 Additionally, as a result of the decentralized structure of American government, the local level is charged with implementing many federal and state policies. There can be substantial variation in enactment and ultimately the success or failure of such policies.Footnote 16 The trend toward devolution of policy-making authority over the last 20 years has meant that local governments have come to play an increasingly large role in the creation of public policy.
In general, residents tend to care deeply about the outcomes produced by local governments—from schools, to public safety, to land use decisions. Since the 1960s the American National Election Study (ANES) has asked respondents to state what they think is the most important problem (or problems) facing the nation. Starting in the 1980s social welfare issues like housing, unemployment, and education have ranked among respondents' top three concerns, and since the 1990s, public order issues like crime, drugs, and gun control have ranked among the top three.Footnote 17 Although not exclusively the province of local governments, these issues represent a significant proportion of the political activity at the sub-state level. Additionally, about 85 percent of respondents agree that local elections are important.Footnote 18 When asked about the performance of different levels of government, about 31 percent of respondents stated that their local government was doing a good or very good job, compared to 26 percent of respondents who rated state government that way, and only 15 percent of respondents who felt the same about the federal government.Footnote 19 Yet, despite the importance of local government, we lack a comprehensive body of theoretically driven research that explains variation in policy outcomes and public opinion at the local level.
On another note, many non-local topics cannot be adequately studied without attention to local politics. In a large number of policy arenas, national-level debates and federal policy have grown out of local level activity, which in turn affects local politics. For example, scholars have argued that elements of the modern conservative movement first developed in cities as a response to particularly urban challenges relating to race, redistribution, taxation, jobs, education, and the provision of services.Footnote 20 Similarly, Prohibition-era restrictions on the sale, manufacture, and transport of alcohol were first passed and fought in cities, counties, and states, and were motivated by a desire to assimilate immigrants and Catholics who represented significant demographic groups in early cities. The Constitutional amendment enacting Prohibition nationwide in turn had an enormous effect on city governance as local politicians became enmeshed with underworld activity and city police forces became the enforcers of various facets of the federal law.Footnote 21
Federal housing policy also shows evidence of a cyclical relationship between local politics and national policy. The racially discriminatory regulations of the federal Home Owners Loan Corporation, the Federal Housing Administration, and the Veteran's Administration mortgage programs institutionalized practices that had been developed by local-level entrepreneurs to keep blacks segregated in northern cities.Footnote 22 The federal programs in turn reinforced racial divisions in cities throughout the twentieth century.Footnote 23 Truly understanding federal practices like these requires an understanding of the local context that gave rise to the policies in the first place and the effects that the federal policies have in cities in an ongoing way.
Scholars of democracy and inequality have also come to realize that unpacking puzzles (and discovering solutions) in these realms requires a complex understanding of space and place that necessitates a focus on local factors.Footnote 24 Socio-demographic and political contexts matter for determining whether people participate in politics, how they participate in politics, and what the effects of their participation are likely to be. For instance, recent work by Oliver and Wong shows that racial attitudes are significantly affected by the level of integration in respondents' neighborhoods; more diversity leads to more positive attitudes towards out-groups.Footnote 25 Understanding the relationship between segregation and racial attitudes can help us predict and explain political patterns over time and across places. On another topic, Gimpel, Dyck, and Shaw find that Republicans are less likely to vote when they live in Democratic neighborhoods even when controlling for individual socioeconomic characteristics.Footnote 26 Thus, the local context is likely to shape political outcomes in a variety of ways.
But scholars must be attentive not only to the effects of local context on political actors and outcomes but also to those factors that shape the context itself, which means attending to local political processes. Local policy variation inevitably leads to residential and business location decisions, which generate segregation along various dimensions (race, class, educational attainment, etc.). In turn, this sorting affects important political outcomes like citizen participation, polarization of opinions, and inequality.Footnote 27 Swanstrom explains that “a special strength of [the urban politics field] is its ability to expose and contest the background conditions that shape metropolitan development and disable American democracy.”Footnote 28
Methods Issues
Regardless of the arguments for studying cities substantively, many scholars will be more interested in questions that are not unique to local politics. But there are still methodological reasons to invoke the local level in the study of political science because there are many generalizable lessons that can be gleaned from cities. In the past scholars fruitfully turned to cities to answer a range of important political science questions. Dahl drew far reaching conclusions about the nature of power and democracy by analyzing urban redevelopment, public education, and partisan politics in New Haven, Connecticut.Footnote 29 Sayre and Kaufman studied delegation, bureaucratic control, and effective governance using the unique case of New York City.Footnote 30 Hunter, Bachrach and Baratz, and Gaventa all relied on city settings to explain forms of influence and agenda control.Footnote 31 Browning, Marshall, and Tabb investigated the processes of coalition building and political incorporation by analyzing a sample of California cities.Footnote 32 Other scholars have used larger data sets to study local politics, like Aiken and Alford, who analyzed the distribution of federal funds in more than 600 cities, or Welch and Bledsoe, who studied the effect of institutions on representation in hundreds of communities.Footnote 33 These classic works were able to draw insightful conclusions because there are political lessons that transcend jurisdictional boundaries. However, the integration between the study of local phenomena and larger political science questions has been weak over the last two decades. Today, tremendous opportunities exist for scholars to reconnect local politics to the mainstream.
Many institutional and behavioral factors vary substantially at the sub-state level. This variation is particularly important in the study of institutions that tend to vary slowly or not at all at the national level and in only a limited way at the state level. For instance, whereas all members of Congress and most members of state legislatures are elected in single-member districts, most cities elect some or all of their members in multi-member (at-large) elections and they change the number and proportions of councilors elected in each type of system over time. Scholars have analyzed how the selection method affects the election of certain types of candidates (such as racial and ethnic minorities and women), but the reasons that these institutions have differential effects remain understudied. For instance, my article with Melody Valdini finds that white women are more likely to be represented on at-large (versus districted) councils but additional research is needed to explain why.Footnote 34 Perhaps voters are more willing to diversify in at-large settings when they can choose women and men in the same election or, alternatively, that multi-member elections change the strategies of elites (or a combination of both). Knowing which argument has more explanatory power could help us to learn more about why women remain underrepresented in Congress as well as in cities and better understand how voters use gender as a heuristic in voting.
The role of political parties also varies in local elections. Most locally-elected officials run on nonpartisan ballots where their party label is not printed next to their name, but about 23 percent are officially partisan.Footnote 35 Even in nonpartisan systems, the degree to which parties actually participate varies widely. In some cases both parties are active and well represented in elected offices; in others, one party accounts for all (or nearly all) of the activity and representation. In still other cases, parties are relatively absent from the political scene. Such differences offer scholars the opportunity to analyze election outcomes and policy-making in different partisan environments.
The variance across cases and sheer size of the N allows for better analysis in part because of the presence of control cases. Matching is much easier in this context. If we think that party labels matter when voters select candidates, it is useful to also include cases in which party labels are not available. Similarly, if one argues that partisan control of the legislature matters more than ideological distribution, it might be helpful to compare cases in which parties officially organize legislators and where they do not. And because the size, institutionalization, and professionalization of city councils varies extensively across time and place, scholars have the opportunity to study how changes in these factors affect everything from elections to representation.
Diversity at the local level is also useful for scholars invoking qualitative methods. Peter John explains that because scholars have access to large numbers of cases with a high degree of natural variation, “researcher[s] can select cases on variations in the independent variable as well as ensure that the dependent variable varies too.”Footnote 36 Particularly for scholars interested in difficult-to-study topics, like understanding the ways in which access to power holders is important for achieving policy outcomes, or why party activists become involved in politics, the smallness and variety of contexts at the local level can be beneficial.
Different Answers at Different Levels
Theories do not sufficiently explain a large amount of local phenomena at this point, and significant findings at the local level have remained unincorporated into political science more generally. In my view, the most promising urban scholarship strives to do both of these things. That is, we can study particularly interesting local-level topics and use the methodological advantages that come from variation to address general questions in political science. A focus on the uniqueness of cities is one way to invoke both benefits. For example, cities have a different set of public responsibilities than states or the federal government, many of which have clear distributive implications. Scholars interested in determining whether or not the government works, and for whom it works, may find more precise (and collectable) measures at the local level than at state or national levels. One might investigate which neighborhoods have well-maintained parks, schools, and libraries and which do not, or analyze where crime, sewer overflows, and fires are most likely to occur. More frequently than is true at the federal level, residents tend to share views on the desirability of these kinds of outcomes and the connection between political decisions and outcomes can be relatively clear. Scholars might then use these data to explain the factors that enhance representation in a democratic system more generally.
The urban politics field has also traditionally researched questions that have been, at best, peripheral to national level research agendas, but that should be relevant to the broader political universe. A large body of urban scholarship pioneered by Clarence Stone, referred to as regime theory, analyzes how informal power plays a role in the development and implementation of policy.Footnote 37 Regime theory integrates political and economic forces in order to evaluate the governance of cities and explores the role of voters in varied political and economic contexts. Clearly, public and private power intersect at other levels of government and scholars might use the insights of regime theory to build theories elsewhere. For example, recent work on the relationship between federal policy and citizens' preferences has offered substantial evidence that what voters (or some subsets of voters) want has little relevance to major government initiatives.Footnote 38 Drawing on regime theory could help scholars to derive expectations about the conditions under which we might expect participation to be more or less significant.
The effect of patronage was once a dominant focus of local-politics literature. Since the demise of the classic political machine, very little has been written on patronage. At the national level, scholarship on the bureaucracy has tended to focus on understanding executive and Congressional control with scant attention to the electoral benefits of patronage appointments. A return to local level analysis could contribute to our understanding of the bureaucracy more generally by taking advantage of the large numbers of municipal employees (and wide variation across cities) and the historical linkage between reelection and patronage. By doing so scholars might gain insight into the effect of public employment on elections and the tradeoffs between bureaucratic performance and electoral rationales.
Other scholars have found that inattention to the local level has led previous analysts to draw, at best, incomplete pictures of the political world. One example comes from the field of American political development. Contrary to the received wisdom among many scholars of American political development, William Novak (Reference Novak1996) argues that the nineteenth-century American state was not weak, nor governed by an extreme Lockean liberalism.Footnote 39 He comes to understand this by looking to state and local laws where he finds substantial regulation of social life in early America. A second example comes from the domain of voting. Zoltan Hajnal and I take advantage of the varying demographic composition across cities to analyze the effect on election outcomes of differential levels of turnout among minority groups.Footnote 40 We find that contrary to evidence at the national level, turnout imbalances can make a difference to political results particularly where minorities make up substantial proportions of the population.
Hurdles
Clearly then, studying local politics offers benefits to scholars interested in many different kinds of questions. However, there are serious challenges at this point to embarking on such a path. Lack of centralized data presents one of the biggest hurdles. The dearth of quantitative studies of local politics is both a cause and effect of this shortage. Students searching for large data sets to mine have few options at the local level. This is made even more complicated by difficulty in determining exactly how to measure the proper political community (e.g. incorporated cities vs. metropolitan areas). In the past it was particularly troublesome to collect election returns at the local level. The emergence of the Internet has helped substantially. But it continues to be true that answering an interesting question frequently requires collecting the data oneself. In some cases, e.g., conducting a survey across multiple locales, the collection can be extremely expensive. Additionally, there are unique methodological challenges in studying cities, given their subordinate status in the political system and the strong influences of neighboring places. These hurdles will become less onerous as more scholars enter the local politics field and share their data and methods, but for now new local-politics scholars must confront these challenges.
Second, some scholars will view the uniqueness of the local context and local politics as a liability. It is true that the collective action problems on city councils will never rival the problems faced by Congress and that most local elections will be far less visible than state or national elections. Effectively using the distinctive qualities of local politics to inform larger questions in political science requires that scholars be attentive to the lessons that cannot be applied at other levels of government as well. Cities are different—they have diverse demographic distributions, elite networks are relatively small, conflict over space is paramount, they compete for population and businesses, and they have unique functional responsibilities. This inevitably means that some aspects of local-politics scholarship will never be interesting to scholars outside of the subfield. This does not mean that studying local politics for its own sake is unimportant. On the contrary, I believe that our knowledge of American politics remains incomplete without a thorough understanding of local political phenomena; but it does mean that there are some debates that will remain confined to the subfield. For example, one of the arguments that I make elsewhere is that machine and reform political coalitions are more similar than the classic urban literature has suggested.Footnote 41 While I believe this to be an important contribution to a central urban politics debate, this particular facet of my book is not likely to transcend the urban field (or be interesting to scholars who have no interest in local politics per se). However, pieces of the book relate to larger debates, including discussions of the mechanisms that coalitions use to capture and keep power and analyses of turnout and policy responsiveness when coalitions dominate for long periods of time. I see both kinds of contributions as important when studying politics at the city level.
Because it is a mistake to assume that the constraints, patterns, and processes that have been explained well at the national level will directly translate at the local level, pursing local research can be difficult and time consuming—as if starting from scratch. It is not enough to just “use” cities as cases, because cities may well be different. The flip side is that there are tremendous opportunities for theory building, description, and explanation at the local level using a wide range of methods.
In sum, even recognizing these hurdles, there are good reasons to study local politics. Doing so can offer both substantive and methodological benefits, and ultimately advance our knowledge of political phenomena. Maybe someday it will be the case that when classes are taught on elections, at least one segment is dedicated to understanding elections in cities, both because these elections are like other kinds of elections and because they are not. Research on local politics can and should contribute to broader debates in political science and ensure that we understand both how and why cities are unique.
