¹ Robert Dahl, ‘The City in the Future of Democracy’, American Political Science Review, 61 (1967), 953–70; Richard Dagger, ‘Metropolis, Memory, and Citizenship’, American Journal of Political Science, 24 (1981), 715–37.

² Sidney Verba, Kay Schlozman and Henry Brady, Voice and Equality (Cambridge, Mass.: Harvard University Press, 1995). See also Henry Milner, Civic Literacy (Hanover, N.H.: Tufts University, 2002).

³ J. Eric Oliver, ‘City Size and Civic Involvement in Metropolitan America’, American Political Science Review, 94 (2000), 361–73, p. 362.

⁴ Dahl, ‘The City in the Future of Democracy’.

⁵ Oliver, ‘City Size and Civic Involvement in Metropolitan America’, p. 362.

⁶ Benjamin Barber, Strong Democracy (Berkeley: University of California Press, 1984); Stephen L. Elkin, City and Regime in the American Republic (Chicago: University of Chicago Press, 1987); Dagger, ‘Metropolis, Memory, and Citizenship’; Jeffrey Berry, Kent Portney and Ken Thomas, The Rebirth of Urban Democracy (Washington, D.C.: The Brookings Institution, 1993); Elinor Ostrom, ‘Metropolitan Reform: Propositions Derived from Two Traditions’, Social Science Quarterly, 53 (1972), 474–93.

⁷ H. G. Wells, ‘Administrative Areas’, reprinted in Oliver P. Williams and Charles Press, eds, Democracy in Urban America (Chicago: Rand McNally, 1961), pp. 145–53.

⁸ See, among many other works, Michael Keating, ‘Size, Efficiency and Democracy: Consolidation, Fragmentation and Public Choice’, in David Judge, Gerry Stoker and Harold Wolman, eds, Theories of Urban Government (London: Sage, 1995), pp. 117–34; Lawrence E. Rose, ‘Municipal Size and Local Nonelectoral Participation: Findings From Denmark, the Netherlands, and Norway’, Environment and Planning C, 20 (2002), 829–52; Bas Denters, ‘Size and Political Trust: Evidence from Denmark, the Netherlands, Norway and the United Kingdom’, Environment and Planning C, 20 (2002), 793–812; M. R. Martins, ‘Size of Municipalities, Efficiency and Citizen Participation: A Cross-European Perspective’, Environment and Planning C, 13 (1995), 441–58; Kenneth Newton, ‘Is Small Really So Beautiful? Is Big Really So Ugly? Size, Effectiveness and Democracy in Local Government’, Political Studies, 30 (1982), 190–206; K. Jones, R. J. Johnston, and C. J. Pattie, ‘People, Places, and Regions: Exploring the Use of Multi-Level Modelling in the Analysis of Electoral Data’, British Journal of Political Science, 22 (1992), 343–80.

⁹ Alan Di Gaetano and E. Strom, ‘Comparative Urban Governance: An Integrated Approach’, Urban Affairs Review, 38 (1999), 356–95; Bas Denters and Karen Mossberger, ‘Building Blocks for a Methodology for Comparative Urban Political Research’, Urban Affairs Review, 41 (2006), 550–71; David Judge, Gerry Stoker and Harold Wolman, Theories of Urban Government (London: Sage, 1995).

¹⁰ Christine Kelleher and David Lowery, ‘Political Participation and Metropolitan Institutional Contexts’, Urban Affairs Review, 39 (2004), 720–57.

¹¹ For example, Kelleher and Lowery’s 2004 study of twelve urban counties reported that correlations between the average size of the cities in a county and its levels of concentration and fragmentation were 0.873 and −0.748, respectively. The correlation between fragmentation and concentration was −0.679.

¹² Ostrom, ‘Metropolitan Reform: Propositions Derived from Two Traditions.’

¹³ Kelleher and Lowery, ‘Political Participation and Metropolitan Institutional Contexts;’ see also Newton, ‘Is Small Really So Beautiful?’

¹⁴ For example, public choice supporters of the Tiebout model often compare levels of political participation and government performance across large and small municipalities within metropolitan areas where the number of location choices provided by its level of jurisdictional fragmentation – the real engine underlying their hypothesis – is a constant.

¹⁵ When Kelleher and Lowery (‘Political Participation and Metropolitan Institutional Contexts’) looked at the same relationships using municipalities as the unit of analysis rather than urban counties, the strong correlations observed earlier are markedly attenuated. City size (rather than the average size of city in a county) was only weakly correlated with county concentration (r = 0.179) and county fragmentation (−0.058).

¹⁶ Verba, Schlozman and Brady, Voice and Equality.

¹⁷ Barber, Strong Democracy; Elkin, City and Regime in the American Republic; David Reisman, The Lonely Crowd: A Study of the Changing American Character (Garden City, N.Y.: Doubleday, 1953).

¹⁸ Mancur Olson, The Logic of Collective Action (Cambridge, Mass.: Harvard University Press, 1965); Steven Hansen, Thomas Palfrey and Howard Rosenthal, ‘The Downsian Model of Electoral Participation: Formal Theory and Empirical Analysis of Constituency Size Effect’, Public Choice, 52 (1987), 15–33.

¹⁹ David Lowery and Holly Brasher, Organized Interests and American Government (Boston, Mass.: McGraw Hill, 2004), pp. 37–45.

²⁰ Verba, Schlozman and Brady, Voice and Equality. See also Dennis Chong, Collective Action and the Civil Rights Movement (Chicago: University of Chicago Press, 1991); Jan Leighley, ‘Group Membership and the Mobilization of Political Participation’, Journal of Politics, 58 (1996), 447–63.

²¹ Virginia Gray and David Lowery, The Population Ecology of Interest Representation: Lobbying Communities in the American States (Ann Arbor: University of Michigan Press, 1996); see also Claude Fischer, ‘The Subcultural Theory of Urbanism: A Twenty-Five Year Reassessment’, American Journal of Sociology, 101 (1995), 543–77.

²² William A. Fischel, The Homevoter Hypothesis (Cambridge, Mass.: Harvard University Press, 2001), pp. 74–6; Mark Schneider, The Competitive City (Pittsburgh: University of Pittsburgh Press, 1989), pp. 8–9; Elaine Sharp, Citizen Demand Making in the Urban Context (University: University of Alabama Press, 1986), pp. 133–61; Ronald J. Oakerson and Roger B. Parks, ‘Citizen Voice and Public Entrepreneurship: The Organizational Dynamic of a Complex Metropolitan County’, Publius, 18 (1988), 91–112.

²³ William Blomquist and Roger B. Parks, ‘Fiscal, Service, and Political Impacts of Indianapolis-Marion County’s Unigov’, Publius, 25 (1988), 37–54.

²⁴ Charles Tiebout, ‘A Pure Theory of Local Expenditures’, Journal of Political Economy, 66 (1956), 416–24; Vincent Ostrom, Charles Tiebout and Robert Warren, ‘The Organization of Government in Metropolitan Areas’, American Political Science Review, 55 (1961), 831–42.

²⁵ Vincent Ostrom, Robert Bish and Elinor Ostrom, Local Government in the United States (San Francisco: Institute for Contemporary Studies, 1988), pp. 91–2.

²⁶ Ostrom, Bish and Ostrom, Local Government in the United States, pp. 93–5; Schneider, The Competitive City, pp. 8–9, 206–9.

²⁷ Fischel, The Homevoter Hypothesis; Schneider, The Competitive City. We should note the seeming incongruity of listing what appears so far as a pure public choice hypothesis in Table 1 under the large-is-lively school. The key to this paradox is that while some opponents of fragmentation recognize that reliance on traditional voice behaviours can be diminished in the face of a strong exit option, they do not view exit as an effective replacement for traditional modes of participation.

²⁸ For example, see Sharp, Citizen Demand Making in the Urban Context, pp. 156–9; Kelleher and Lowery, ‘Political Participation and Metropolitan Institutional Contexts’; William E. Lyons and David Lowery, ‘Citizen Responses to Dissatisfaction in Urban Communities: A Partial Test of a General Model’, Journal of Politics, 51 (1989), 842–68.

²⁹ Ada Finifter and Paul R. Abramson, ‘City Size and Feelings of Political Competence’, Public Opinion Quarterly, 39 (1975), 189–98.

³⁰ Oliver, ‘City Size and Civic Involvement in Metropolitan America’, p. 363.

³¹ Bernard H. Ross and Myron A. Levine, Urban Politics (Itasca, Ill.: F. E. Peacock, 2001), p. 307.

³² For example, see Myron Orfield, Metro Politics (Washington, D.C.: The Brookings Institution, 1997); Douglas S. Massey and Nancy A. Denton, American Apartheid (Cambridge, Mass.: Harvard University Press, 1993); David Rusk, Cities Without Suburbs (Baltimore, Md.: Woodrow Wilson Center, 1995); Anthony Downs, New Visions for Metropolitan America (Washington, D.C.: Brookings Institution, 1994); Robert Waste, Independent Cities (New York: Oxford University Press, 1998); Gregory R. Weiher, The Fractured Metropolis (Albany, N.Y.: State University of New York Press, 1991); Peter Drier, John H. Mollenkopf and Todd Swanstrom, Place Matters (Lawrence: University of Kansas Press, 2001).

³³ Nancy Burns, The Formation of American Local Governments (New York: Oxford University Press, 1994), p. 117.

³⁴ Peter Bachrach and Morton S. Baratz, Power and Poverty (New York: Oxford University Press, 1970).

³⁵ See Wells, ‘Administrative Areas’; David Lowery, Ruth Hoogland DeHoog and William E. Lyons, ‘Citizenship in the Empowered Locality: An Elaboration, a Critique, and a Partial Test’, Urban Affairs Quarterly, 28 (1992), 69–103; Fischel, The Homevoter Hypothesis, p. 12.

³⁶ Oliver, ‘City Size and Civic Involvement in Metropolitan America’, p. 362. Perhaps the one exception to this positive view is a pure Tiebout interpretation. Tiebout’s emphasis on exit essentially precludes viewing psychological attachment to local communities as a virtue (see Robert Salisbury, ‘The Local Community: An Arena for Citizenship in the Federal System’, in Stephen L. Schechter, ed., Teaching About American Federal Democracy (Philadelphia: Temple University Press, 1984), pp. 49–56); rather, it would be an extraneous impediment to the efficient operation of the market.

³⁷ See Wells, ‘Administrative Areas’; Robert C. Wood, 1400 Governments (Cambridge, Mass.: Harvard University Press, 1961), pp. 179–80; Norton E. Long, The Polity (Chicago: Rand McNally, 1962).

³⁸ For example, see John D. Kasarda and Morris Janowitz, ‘Community Attachment in Mass Society’, American Sociological Review, 39 (1974), 328–39; Robert J. Sampson, ‘Local Friendship Ties and Community Attachment in Mass Society: A Multilevel Systemic Model’, American Sociological Review, 53 (1988), 766–79. Oliver’s work does find that attachment is strongly related to length of residence at the individual level and residential mobility at the aggregate level.

³⁹ David Lowery and William E. Lyons, ‘The Impact of Jurisdictional Boundaries: An Individual-Level Test of the Tiebout Model’, Journal of Politics, 51 (1989), 73–97.

⁴⁰ David Lowery, William E. Lyons and Ruth Hoogland DeHoog, ‘Institutionally Induced Attribution Errors: Their Composition and Impact on Citizen Satisfaction with Local Government Services’, American Politics Quarterly, 18 (1992), 169–96.

⁴¹ Dahl, ‘The City in the Future of Democracy’.

⁴² Downs, New Visions for Metropolitan America.

⁴³ Oliver, ‘City Size and Civic Involvement in Metropolitan America’, p. 362.

⁴⁴ Fischel, The Homevoter Hypothesis.

⁴⁵ Anita Miller, ‘Rural Development: Considerations for Growth Management’, Natural Resources Journal, 43 (2003), 781–802.

⁴⁶ Kelleher and Lowery, ‘Political Participation and Metropolitan Institutional Contexts’.

⁴⁷ Dora L. Costa and Matthew E. Kahn, ‘Civic Engagement and Community Heterogeneity: An Economist’s Perspective’, Perspectives on Politics, 1 (2003), 103–12, p. 104.

⁴⁸ Oliver, ‘City Size and Civic Involvement in Metropolitan America’, p. 362. We shall see, however, that Oliver has come very close to arguing both sides of this issue.

⁴⁹ Costa and Kahn, ‘Civic Engagement and Community Heterogeneity’, p. 108.

⁵⁰ For example, see Rodney Hero, ‘Social Capital and Racial Equality in America’, Perspectives on Politics, 1 (2003), 113–22; Rodney Hero, ‘Multiple Theoretical Traditions in American Politics and Racial Policy Inequality’, Political Research Quarterly, 56 (2003), 401–8; Paula McClain, ‘Social Capital and Diversity: An Introduction’, Perspectives on Politics, 1 (2003), pp. 101–2.

⁵¹ Karl Deutsch, ‘Social Mobilization and Political Development’, American Political Science Review, 55 (1961), 493–514; Verba, Schlozman and Brady, Voice and Equality.

⁵² Daniel Rubenson, ‘Community Heterogeneity and Political Participation in American Cities’ (paper presented at the Annual Meeting of the American Political Science Association, Philadelphia, 2003).

⁵³ Oliver, ‘City Size and Civic Involvement in Metropolitan America’, p. 362.

⁵⁴ J. Eric Oliver, ‘The Effects of Metropolitan Economic Segregation on Local Civic Participation’, American Journal of Political Science, 43 (1999), 186–212, p. 191.

⁵⁵ Oliver, ‘The Effects of Metropolitan Economic Segregation on Local Civic Participation’. The previous scholarship referenced is Downs, New Visions for Metropolitan America; Rusk, Cities Without Suburbs; Massey and Denton, American Apartheid; Orfield, Metro Politics; Drier, Mollenkopf and Swanstrom, Place Matters; Waste, Independent Cities.

⁵⁶ Yvette Alex-Assenhoh, ‘Race, Concentrated Poverty, Social Isolation, and Political Behavior’, Urban Affairs Review, 33 (1997), 209–27.

⁵⁷ Oliver, ‘The Effects of Metropolitan Economic Segregation on Local Civic Participation,’ p. 204. Unfortunately, Oliver did not test this hypothesis directly by examining levels of fragmentation.

⁵⁸ Weiher, The Fractured Metropolis.

⁵⁹ Salisbury, ‘The Local Community’.

⁶⁰ Lowery and Lyons, ‘The Impact of Jurisdictional Boundaries; William E. Lyons and David Lowery, ‘The Organization of Political Space and Citizen Responses to Dissatisfaction in Urban Communities: An Integrated Model’, Journal of Politics, 48 (1986), 321–34.

⁶¹ Sharp,Citizen Demand Making in the Urban Context.

⁶² Tiebout, ‘A Pure Theory of Local Expenditures’; Oakerson and Parks, ‘Citizen Voice and Public Entrepreneurship’.

⁶³ Ostrom, Tiebout and Warren, ‘The Organization of Government in Metropolitan Areas’.

⁶⁴ Christine Kelleher and David Lowery, ‘Tiebout Sorting and Selective Satisfaction with Urban Public Services’, Urban Affairs Review, 37 (2002), 420–31.

⁶⁵ Robert Stein, ‘Tiebout’s Sorting Hypothesis’, Urban Affairs Quarterly, 23 (1987), 140–60.

⁶⁶ Kelleher and Lowery, ‘Tiebout Sorting and Selective Satisfaction with Urban Public Services’; Ruth Hoogland DeHoog, David Lowery and William E. Lyons, ‘Citizen Satisfaction with Local Governance: A Test of Individual, Jurisdictional, and City-Specific Explanations’, Journal of Politics, 52 (1990), 807–37.

⁶⁷ See Downs, New Visions for Metropolitan America; Richard Child Hill, ‘Separate and Unequal: Governmental Inequality in the Metropolis’, American Political Science Review, 68 (1974), 1557–68; Max Neiman, ‘Communication: Social Stratification and Government Inequality’, American Political Science Review, 70 (1976), 149–80; Weiher, The Fractured Metropolis.

⁶⁸ David Niven, ‘The Mobilization Solution? Face-to-Face Contact and Voter Turnout in a Municipal Election’, Journal of Politics, 66 (2004), 868–84.

⁶⁹ Scott McClurg, ‘Social Networks and Participation: The Role of Social Interaction in Explaining Political Participation’, Political Research Quarterly, 56 (2003), 449–65; Scott McClurg, ‘Indirect Mobilization’, American Politics Research, 32 (2004), 406–44.

⁷⁰ Oliver, ‘City Size and Civic Involvement in Metropolitan America’, p. 363; Claude Fischer, To Dwell Among Friends: Personal Networks in Town and City (Chicago: University of Chicago Press, 1982); Lyn Lofland, A World of Strangers: Order and Action in Urban Public Space (New York: Basic Books, 1973); Robert Huckfeldt and John T. Sprague, Citizens, Politics, and Social Communication: Information and Influence in an Election Campaign (New York: Cambridge University Press, 1995).

⁷¹ Oliver, ‘City Size and Civic Involvement in Metropolitan America’, pp. 370–1. Even if true, remember that research indicates that having friends and neighbours depends far more on residential mobility than population per se (Sampson, ‘Local Friendship Ties and Community Attachment in Mass Society’).

⁷² Steven J. Rosenstone and John Mark Hansen, Mobilization, Participation, and Democracy in America (New York: Macmillan, 1993).

⁷³ Roger B. Parks, Paula C. Baker, Larry L. Kiser, Ronald J. Oakerson, Elinor Ostrom, Vincent Ostrom, Stephen L. Percy, Martha Vandivort, Gordon P. Whitaker and Rick K. Wilson, ‘Consumers as Coproducers of Public Services: Some Economic and Institutional Considerations’, Policy Studies Journal, 9 (1981), 1001–1011.

⁷⁴ Robert Lineberry and E. P. Fowler, ‘Reforms and Public Policies in American Cities’, American Political Science Review, 61 (1967), 701–16; Robert Alford and Eugene C. Lee, ‘Voting Turnout in American Cities’, American Political Science Review, 62 (1968), 796–813; Albert K. Karnig and B. Oliver Walter, ‘Decline in Municipal Voter Turnout’, American Politics Quarterly, 11 (1983), 491–505.

⁷⁵ Curtis Wood, ‘Voter Turnout in City Elections’, Urban Affairs Review, 38 (2002), 209–31.

⁷⁶ Jeremy Weinstein, ‘Abandoning the Polity: Political Parties and Social Capital in American Politics’ (paper presented at the Annual Meeting of the American Political Science Association, Atlanta, 1999).

⁷⁷ We do not believe that the presence of a Knight newspaper in these communities significantly alters or taints our results or conclusions. Nearly all American cities have a newspaper of some sort targeting them. We have no reason to expect that the effect of such will be any different in Knight communities than in other American cities or towns. Generally, journalists abide by collective professional norms (or ‘pack’ journalism), and thus the types (and style) of coverage in different places is actually quite similar. Whether Knight papers (or communities in which they are located) are systematically different from other papers/communities is a question much larger than this article.

⁷⁸ For more information, see 〈http://www.knightfdn.org〉 (retrieved 5 June 2007).

⁷⁹ These percentages vary because the random samples were based on varied county sizes, and were focused on establishing appropriate regional variation. The cities in which larger samples are from the central city are, generally, cities in which the central city occupies a larger proportion of the county population. In the final analyses, cases without full data are excluded by listwise deletion. This brings the final number of observations in all models to approximately 7,500. The number of individuals in a city ranges from 100 in Myrtle Beach to approximately 530 in Philadelphia.

⁸⁰ See Michael N. Danielson and Paul G. Lewis, ‘City Bound: Political Science and the American Metropolis’, Political Research Quarterly, 49 (1996), 203–20, and Jered Carr, ‘Perspectives on City County Consolidation and its Alternatives’, in Jered B. Carr and Richard C. Feiock, eds, Reshaping the Local Landscape (Armonk, N.Y.: M.E. Sharpe, 2004), pp. 3–24.

⁸¹ Philip Meyer, ‘Introduction’ to Listening and Learning: Community Indicator Profiles of Knight Foundation Communities and the Nation (Report prepared by the John S. and James L. Knight Foundation, Princeton Survey Research Associates, American Institutes for Research, and the Urban Institute, 2004).

⁸² Nationally, about half of all cities have a mayor–council structure (see Ann O’M. Bowman and Richard C. Kearney, State and Local Government (Boston, Mass.: Houghton Mifflin, 2002)).

⁸³ Nationally, approximately 60 per cent of cities have pure at-large elections (see Virginia Gray and Peter Eisinger, American States and Cities (New York: HarperCollins, 1997).

⁸⁴ Nationally, only a quarter or so of all municipalities conduct partisan elections (Gray, and Eisinger, American States and Cities).

⁸⁵ Wendy Rahn, Kwang Suk Yoon, Michael Garet, Stephen Lipson and Katherine Loflin, ‘Geographies of Trust: Explaining Inter-Community Variation in General Social Trust Using Hierarchical Linear Modeling (HLM)’ (paper presented at the Annual Conference of the American Association of Public Opinion Research, Nashville, Tenn., 2003); Wendy Rahn and Thomas J. Rudolph, ‘A Tale of Trust in American Cities’, Public Opinion Quarterly, 69 (2005), 530–60.

⁸⁶ Oliver, ‘The Effects of Metropolitan Economic Segregation on Local Civic Participation’; Oliver, ‘City Size and Civic Involvement in Metropolitan America’.

⁸⁷ With eighty-eight municipalities, Long Beach is an outlier with respect to fragmentation. When this variable is excluded, the mean level of fragmentation only drops to 10.46. In supplemental analyses not reported, we exclude Long Beach, and the results are identical to those presented here, if not slightly stronger.

⁸⁸ Oliver, ‘City Size and Civic Involvement in Metropolitan America’, pp. 369–70. These intervening relationships might be examined using simultaneous equation models. However, we will see that the hypotheses associated with these relationships fare quite poorly. We opted, therefore, for Oliver’s (‘City Size and Civic Involvement in Metropolitan America’, pp. 369–70) simpler approach of seeing how the inclusion of theoretically more proximate variables attenuates the estimates of variables thought to operate further back in the causal process.

⁸⁹ We also examined Oliver’s (‘The Effects of Metropolitan Economic Segregation on Local Civic Participation’) more proximate indicator of city economic heterogeneity – logged median income and its squared value in a polynomial specification where the former is expected to have a positive and the latter a negative association with participation. Unfortunately, inclusion of these variables generated substantial collinearity problems. Indeed, in the fuller specifications presented below as well as less specified models, collinearity was so extreme that both variables could not be included in the models. Thus, we opted to exclude Oliver’s measures of economic heterogeneity and focus squarely on the direct impact of institutional fragmentation on participation.

⁹⁰ The measures were as follows: Ideology: ‘In general, would you describe your political views as very conservative, conservative, moderate, liberal, or very liberal?’ Age: ‘What is your age?’ Answers are in years. Education: ‘What is the last grade or class you completed in school?’ Responses include (1) none to 8th grade, (2) incomplete high school, (3) high school graduate, (4) business, technical or vocational school after high school, (5) some college, (6) college graduate, and (7) post-graduate work. Non-white: ‘What is your race? Are you white, black, Asian, American Indian or some other race?’ Responses were collapsed into 0 = White; 1 = Black, Asian or Pacific Islander, American Indian or Alaskan Native, Mixed Race, Other. Home Ownership: ‘Do you rent or own your home?’ Income: ‘Approximately what is your total family income before taxes?’ Answers: less than $10,000 (1); $10,000–19,999 (2); $20,000–29,999 (3); $30,000–39,999 (4); $40,000–49,999 (5); $60,000–99,999 (6); $100,000 or more (7).

⁹¹ For example, see Wendy Rahn et al., ‘Geographies of Trust’.

⁹² For example, Kelleher and Lowery, ‘Political Participation and Metropolitan Institutional Contexts’. And some types of research designs are inattentive to both individual and institutional sources of variation in political participation (Costa and Kahn, ‘Civic Engagement and Community Heterogeneity’).

⁹³ Stephen W. Raudenbush and Anthony S. Bryk. Hierarchical Linear Models: Applications and Data Analysis Methods (London: Sage Publications, 2002); Marco S. Steenbergen and Bradford S. Jones, ‘Modeling Multilevel Data Structures’, American Journal of Political Science, 46 (2002), 218–37; Rahn and Rudolph, ‘A Tale of Trust in American Cities’.

⁹⁴ Stewart J. D’Alessio and Lisa Stolzenberg, ‘A Multilevel Analysis of Relationship between Labor Surplus and Pretrial Incarceration’, Social Problems, 49 (2002), 178–93.

⁹⁵ Steenbergen and Jones, ‘Modeling Multilevel Data Structures’.

⁹⁶ Steenbergen and Jones, ‘Modeling Multilevel Data Structures’; Raudenbush and Bryk, Hierarchical Linear Models.

⁹⁷ Raudenbush and Bryk, Hierarchical Linear Models, p. 24; Steenbergen and Jones, ‘Modeling Multilevel Data Structures’, p. 224.

⁹⁸ This is essentially a one-way ANOVA with random effects.

⁹⁹ The values for the constant tell us the average log-odds of voting and joining a civic organization. To understand them more succinctly, we transform them to odds and probabilities. The odds of voter registration are 3.25 (probability of 0.77), and the odds of membership in a civic organization are 0.31 (probability of 0.24).

¹⁰⁰ The ICC is calculated according to the following formula: τ ₀₀/(τ ₀₀ + π ²/3), where τ ₀₀ is the city variance component. This is slightly different than what is used to calculate the ICC in linear models, due to problems with heteroscedasticity (Raudenbush and Bryk, Hierarchical Linear Models, p. 334; Tom Snijders and Roel Bosker, Multilevel Analysis (London: Sage Publications, 1999); Daniel Rubenson, ‘Can Social Capital Account for Differences in Political Participation Across American Cities?’ (paper prepared for the Annual Meeting of the American Political Science Association, Washington, D.C., 2005)).

¹⁰¹ Rahn and Rudolph, ‘A Tale of Trust in American Cities’.

¹⁰² Rubenson, ‘Can Social Capital Account for Differences in Political Participation Across American Cities?’

¹⁰³ Steenbergen and Jones, ‘Modeling Multilevel Data Structures’.

¹⁰⁴ Raudenbush and Bryk, Hierarchical Linear Models, pp. 130, 256–7.

¹⁰⁵ To determine the appropriate specification and the necessity of fixed or varying effects, a first option is to estimate a ‘saturated’ model in which all Level 1 predictors are included and specified as random in their corresponding Level 2 models. Significant effects are retained in subsequent estimations, while insignificant ones are dropped. However, Raudenbush and Bryk (Hierarchical Linear Models, p. 256) recommend against doing this unless the sample sizes within each individual unit are relatively high (in other words, large enough to conduct ordinary least squares (OLS) regression in each unit). An additional strategy they suggest is to ‘step up’ (p. 257). In other words, they advise using OLS regression within units or theoretical considerations to isolate a small subset of variables. Subsequently, then, they suggest building the model up to investigate whether each new specification is an improvement over prior estimations. One can then compare the intercept variance components to see if the introduction of these parameters improves the amount of variance still unexplained (Raudenbush and Bryk, Hierarchical Linear Models, p. 309).The decisions concerning random versus fixed effects are further complicated in non-linear models because standard tests and comparison mechanisms between different specifications (such as likelihood ratio tests) cannot be used. As a result, in deciding about random versus non-random Level 1 variables in our estimations, we pursued multiple strategies. This was possible because each of the cities has an average of 297 respondents for the dependent variable of Vote, ranging from 98 to 533 individuals in each city, and an average of 298 respondents for Civic, ranging from 98 to 530 within cities. First, for Vote, we ran a saturated model. In this model, in addition to the random effect for the slope, the random effects associated with the following variables were also significant (p <0.05): School Satisfaction, City Government Satisfaction, Education and Race. Additionally, we ran regression analyses within each individual city. For School Satisfaction, City Government Satisfaction and Race, we observed significant effects with both positive and negative signs in multiple cities, providing evidence that these variables have different effects in different cities. For Education, although the effects were always positive, the magnitude of the significant coefficients across cities was quite varied and, thus, again offers evidence of differing effects of this characteristic in different cities. Thus, in all subsequent hierarchical models, we include these four random effects. For Civic, the only significant random coefficient was for the slope (p <0.05).

¹⁰⁶ However, because explaining the variation in these variables is not a goal of this research, we do not discuss or interpret their magnitude or significance.

¹⁰⁷ This result is only modestly surprising, however. Others have found similar results, which are often attributed to greater and more sustained mobilization efforts among minority citizens.

¹⁰⁸ Oliver, ‘The Effects of Metropolitan Economic Segregation on Local Civic Participation’; Oliver, ‘City Size and Civic Involvement in Metropolitan America’.

¹⁰⁹ Three cities are clearly outliers with respect to their population – San Jose, Detroit and Philadelphia. Therefore, in models not reported here, we used a conservative modelling strategy to test the sensitivity of our results to these cities. We estimated the base models, excluding these three cities and with no random effects, and the results were nearly identical to those reported here. The only major difference was that the sign for mayor-council institutions (in both models) was positive and highly significant. Interestingly, these alternative specifications actually offer more highly significant results for the jurisdiction-level variables of size, fragmentation and concentration.

¹¹⁰ Oliver, ‘The Effects of Metropolitan Economic Segregation on Local Civic Participation’; Oliver, ‘City Size and Civic Involvement in Metropolitan America’.

¹¹¹ Oliver, ‘The Effects of Metropolitan Economic Segregation on Local Civic Participation’.

¹¹² See Rusk, Cities Without Suburbs; Weiher, The Fractured Metropolis; Downs, New Visions for Metropolitan America.

¹¹³ Oliver, ‘City Size and Civic Involvement in Metropolitan America’. On the surface, our results also seem to be inconsistent with those reported by Kelleher and Lowery (‘Political Participation and Metropolitan Institutional Contexts’). The biggest real difference concerns the significance of the size variables. While the estimate of the size variables in the Kelleher and Lowery models were always positive, none were significant. We do not have a ready explanation for this except to note that their model, aside from working with aggregate rather than survey data, was characterized by significant collinearity issues. Less problematic is their finding of a powerful interaction between fragmentation and concentration. We could not test for this interaction in the present analysis given very high levels of collinearity in the models with the full array of dummy controls for the Knight cities. Given this problem we opted for the simpler specification of testing for simple effects for these two variables. However, when Model 3 in Tables 4 and 5 was estimated dropping concentration and including an interaction between concentration and fragmentation, strong evidence for the interaction was evident. In the voting registration model in particular, this pattern of estimates suggests that the modestly positive impact of concentration rapid erodes in the face of any concentration of an urban-county’s population with its central city. In the analyses actually presented in Tables 4 and 5, these interactions are almost certainly lost in the coefficients of the city dummy variables. In the broad sense of providing much strong support for the large-is-lively school, then, the current findings are consistent with those earlier reported by Kelleher and Lowery.

¹¹⁴ Oliver, ‘City Size and Civic Involvement in Metropolitan America’.

¹¹⁵ Oliver, ‘City Size and Civic Involvement in Metropolitan America’.

¹¹⁶ Oliver, ‘City Size and Civic Involvement in Metropolitan America’, p. 371.

¹¹⁷ Importantly, our results do not indicate that the variables we have examined which tap these hypotheses are unimportant in other ways or that they are not a function of fragmentation. Rather, our results suggest that there is no simple link from fragmentation per se through those intervening variables to participation.