Moving beyond the urban/rural cleavage: Measuring values and policy preferences across residential zones in Canada

ABSTRACT

Despite urban forms becoming more varied, analysts continue to use a dichotomous urban/rural distinction when examining political attitudes. Using geocoding, our analysis of original survey data adopts a four-residential-zone approach to exploring Canadian political values and policy preferences: inner city, suburban, small urban, and rural. We find that an ideological polarization between inner city residents on the left and the rest of Canadians on the right emerges and is most pronounced when it comes to values related to new ideology and policy preferences concerning taxation, moral policies, and spending on social assistance. Within large urban centers specifically, suburbanites are more socially conservative than are inner city residents, while divisions between the two groups are smaller with respect to economic issues and the welfare state. We suggest that researchers should replace the binary category of urban/rural for political analyses with a residential zone approach in order to better capture the complexity of contemporary urbanity.

Notes

1. Canadian research has been particularly interested in the question of how place of residence correlates with citizens’ political values and policy preferences, and has considered urban/rural differences as well as differences with respect to province or region (i.e., blocs of geographically contiguous provinces). The appropriateness of including measurements relating to province and region when examining political values and policy preferences is beyond the scope of this study.

2. It is important to note that policy preferences and political values do not automatically align with vote choice. Voting behavior research (Canadian examples include Clarke et al., 2009 and Gidengil et al., 2012) has demonstrated that ideology (political values) and policy preferences are only one component in understanding why voters vote for particular parties, and that in some situations such considerations have only limited impact on vote choice.

3. Geocoding is the process of assigning an absolute geographic location (latitude and longitude) based on a complete or partial address.

4. Abacus Data Research administered the Comparative Provincial Election Project (CPEP) in the weeks immediately after each province’s provincial election: Newfoundland and Labrador 12–30 October 2011 (n = 851); Prince Edward Island 4–25 October 2011 (n = 509); Ontario 7–31 October 2011 (n = 1,044); Manitoba 5–31 October 2011 (n = 775); Saskatchewan 8–21 November 2011 (n = 821); Alberta 25 April–15 May 2012 (n = 897); Quebec 5–29 September 2012 (n = 1009); British Columbia May 15–29 2013 (n = 803); Nova Scotia 9–30 October 2013 (n = 797); New Brunswick 23 September–15 October 2014 (n = 657).

5. For a full definition of Statistics Canada’s concepts of Census Metropolitan Area and Census Ag- glomeration see Statistics Canada, “Census Metropolitan Area and Census Agglomeration definitions,” http://www.statcan.gc.ca/pub/93-600-x/2010000/definitions-eng.htm, accessed 18 September 2014.

6. OLS regression is used due to the ease of interpretation. Ordered logit regression analysis resulted in substantively similar findings. Results available from the authors upon request.

7. The coding for the sociodemographic variables: province (dummy variables for British Columbia, Alberta, Saskatchewan, Manitoba, Quebec, New Brunswick, Nova Scotia, Prince Edward Island, Newfoundland and Labrador with Ontario as the reference category); sex (female = 1, male = 0), age in years, years of educa- tion (0 = less than high school diploma, 8 = professional degree/doctorate), income (0 = less than $20,000, 10 = over $100,000); religious affiliation (Catholic, Protestant, other religion, with no affiliation/atheist as the reference category); visible minority (1 = visible minority, 0 = not visible minority); foreign-born (1 = foreign-born, 0 = Canada-born); union membership (1 = union member, 0 = non-union member); renter (1 = renter; 0 = owner).

8. Work on provincial political culture in Canada argues that province of residence is a key variable to understanding a citizen’s tendency to adopt left-wing and right-wing positions in terms of both old and new ideology (Anderson, 2011; Cochrane & Perrella, 2013; Schwartz, 1974; Simeon & Elkins, 1980). There have been discussions of the effect of gender and immigrant status on the values and policy positions of Canadians (Bilodeau et al., 2010; O’Neill, 2002). Various researchers also have pointed to clusters of Canadians sharing political values and policy positions based upon commonalities such as education, age, religion, and visible minority status (Cutler & Jenkins, 2002; Gidengil, 1990; Henderson, 2004; MacDermid, 1990). Researchers tending towards class analysis have focused on the effects of variables related to income and owning property when examining Canadians’ ideology (Ornstein et al., 1980; Ornstein & Stevenson, 1999).

9. In order to fully understand the relationships between the variables that we were examining, we also ran a model with a small number of sociodemographic variables (residential zone, province, gender, age, education, and income); results available from authors upon request. The key relationships of interest persist in the more parsimonious model.

Additional information

Notes on contributors

David McGrane

David McGrane is an Associate Professor of Political Studies at St. Thomas More College and the University of Saskatchewan. He has published in several academic journals and his most recent research is a book entitled Remaining Loyal: Social Democracy in Quebec and Saskatchewan (2014). His research interests include federal-provincial fiscal relations, multiculturalism, provincial elections, and childcare. He currently holds a Social Sciences and Humanities Research Council grant to write a book on the federal New Democratic Party of Canada.

Loleen Berdahl

Loleen Berdahl is a Professor of Political Studies at the University of Saskatchewan. Her cur- rent research explores science and expertise in policy making; additional research interests include Canadian federalism and regionalism, public policy, and public opinion. Dr. Berdahl is the Project Leader for the Survey and Group Analysis Laboratory (SGAL) at the University of Saskatchewan’s Social Sciences Research Laboratories.

Scott Bell

Scott Bell is a Professor of Geography and Planning at the University of Saskatchewan. His research and teaching interests are in the broad field of geographic information science, with specific interests in spatial cognition, human navigation, and application of GIS technology in the social sciences, health, and humanities. He also holds an academic appointment at Harvard University where he is the lead instructor of the GIS Institute within the Center for Geographic Analysis and Institute for Quantitative Social Sciences.