City assessments and local policy preferences: The relative roles of demographics, ideology, and civic culture

ABSTRACT

This research focuses on the relative impact of demographics, political ideology, and civic culture in the formation of local policy preferences and community assessments. The driving premise is that, whereas demographic and ideological culture variables will be correlated with policy preferences, local civic cultural variables will have independent effects and will enhance the explanatory power of models of local policy preference. Based on analysis of responses to citizen surveys, the findings strongly suggest that civic culture plays an important role not only in strengthening the explanatory power of models of policy preferences but in improving understanding of the statistical significance of ideological culture and demographics.

Notes

1. Charter township status is a special township classification created by the Michigan Legislature in 1947 to provide additional powers and streamlined administration for governing growing communities.

2. Canton Township is located on the far western edge of Wayne County, approximately 20 miles east of Ann Arbor and 35 miles west of Detroit. It was established in 1834 and has been one of the fastest growing communities in the state due in large part to its rural roots and the relative availability of undeveloped land; very little farmland remains in the community. Dearborn is an inner-ring suburb of Detroit with roots in the development of the Ford Motor Company; it was initially developed by Henry Ford to house his White employees and Ford retains substantial investments there. Southfield is a suburb of Detroit in Oakland County and developed beyond the inner ring of suburban communities. It is an edge city (Garreau, 1991) creating a secondary hub of financial, commercial, and retail headquarters beyond the limits of the city of Detroit.

3. Pew Research Center (2018). That the data are from 2010 is also not a particular problem given that the relationships explored in this research are not time bound and focus on general patterns of ideology and policy assessments.

4. Randomly selected telephone numbers for which a directory listing was available were sent an advance letter roughly 1 week prior to when an initial call attempt to contact the household would be made. When interviewers successfully contacted a household, the study procedures required them to randomly select an adult from among those residing in the household to be the respondent. The Trohldal-Carter technique was used as the mechanism for choosing a respondent within each household. Telephone numbers were called across times of the day and days of the week. In the case contact was established, the number was tried until a total of 12 attempts were made or the interview was completed, the interview was refused, or the case was determined to be ineligible or incapable. Interviews were conducted using the Computer Assisted Telephone Interviewing system.

5. The formula used here for response rate determination is as follows :

where I = complete interview; P = partial interview; R = refusal and break-off; NC = noncontact; 0 = other; UH = unknown if household occupied; UO = unknown other; e = estimated proportion of cases of unknown eligibility that are eligible (American Association for Public Opinion Research, 2016). Maximum response rates are as follows: overall, 39.7%; Canton, 42.2%; Dearborn, 37.8%, Southfield, 39.7%.

6. The three communities are significantly different (based on chi-square values) on almost all of the demographic indicators included on the survey (indicated by bold in Table 1). Canton Township has significantly more homeowners, but residents have more recent tenures. Canton also has more children and there are more married couple households. Respondents in Canton are significantly more likely to have a college degree, be employed full time, and have high incomes. Southfield has significantly more residents who identify as African American and Canton Township has the highest level of Asian respondents. The religious backgrounds of survey respondents are also significantly different by community. Dearborn has the most respondents who are Catholic and Muslim, and Southfield respondents are more likely to be Protestant and Jewish. Respondents in Canton were the most likely to indicate no religion. Respondents in Dearborn are significantly more likely to be foreign born, whereas those in Southfield are more likely to be native born.

7. The complete survey is available at http://gusp.msu.edu/?t=research.php.

8. It should be noted that a number of variables commonly used to measure demographics, political ideology, and civic culture were eliminated from the regression models due either (a) to multicollinearity (age/correlated with tenure; unemployment/correlated with median income); (b) because of significant missing data due to the survey being designed in a split-half format to pilot test questions; that is, only half of the respondent received each set of questions (political participation, community participation, community volunteerism); or (3) because they failed to show any explanatory power in any of the regressions (Christian/other religion, sexual preference, marital status). Regression results including these variables are available upon request.

9. Models were also run without the city dummy variables and the results did not change substantively. These results are presented in Appendix E.

10. There are some significant correlations among some of the ideological and civic culture variables, specifically between trust and community attachment. Regressions were run without the trust variables, with trust but without attachment, and without trust and attachment. In all cases, the adjusted R² was higher for regressions with the culture variables included. All regression results are available from the authors upon request. To test for the possibility that the number of ideological and civic culture indexes is inflating the R², models were also run with just those culture variables found to be significantly related to the dependent variables based on the analysis in Table 3. These results are presented in Appendix F. The findings with fewer ideological and civic culture indexes do not change the results substantially.

Additional information

Funding

This research has been supported by a Discretionary Research Fund Award from Michigan State University.

Notes on contributors

Laura A. Reese

Laura A. Reese is Professor of Urban and Regional Planning and the Director of the Global Urban Studies Program at Michigan State University. She is the editor of the Global Urban Book Series for Routledge Publications. Dr. Reese’s main research and teaching areas are in urban politics and public policy, economic development, local governance, and management in Canada and the United States and animal welfare policy. She has written 11 books and over 100 articles and book chapters in these areas as well as public personnel administration focusing on the implementation of sexual harassment policy.

Matthew Zalewski

Matthew Zalewski is an attorney in private practice and a doctoral candidate in Political Science at Michigan State University. He holds a BA in political science and history from Oakland University and a JD from Wayne State University. Matt’s current research interests fall into two broad categories: (a) state courts and judicial politics, judicial campaigns and states as participants in federal court litigation and (b) the legal environment of policymaking, specifically the effects of state–local government structures on local autonomy and the interaction of law, institutions, and policy processes in policymaking.