Research Projects
The Politics of Competence: Longitudinal and Comparative Analysis
Principal Investigator: Dr Jane Green, Institute for Social Change, University of Manchester
Co Investigator: Dr Will Jennings, University of Manchester
1 January 2012 to 31 December 2012
If a political party, or its leadership, is viewed as competent, the party benefits from an important electoral asset. This asset, or valence, relates to issues where policy agreement exists (Stokes 1963). Recent research reveals that valence issues are increasingly important for understanding vote choice in Britain (Denver 1994, 2003; Clarke et al. 2004; 2009). Issue competence ratings also explain party strategies to win votes (Budge and Farlie 1983; Petrocik 1993). Important questions are raised for long-term and comparative dynamics of issue competence. The proposed research will make available long-term cross-national indices of perceived political competence on issues, necessary for addressing these questions:
Is issue valence voting exhibited in electorates beyond Britain, and why?
What factors explain when issue competence judgments are more or less important for understanding elections, across time and context?
How should we understand the interaction of issue competence with other constructs, such as leader ratings, economic ratings, partisanship, events and left-right movements of electorates?
How may issue competence ratings account for variation in voter turnout?
The project will involve the collation of data over policy issues and spanning several decades, the estimation of valence indices in four countries (the UK, the U.S.A., Canada and Australia), initial analysis of these data, the sharing and dissemination of the indices with researchers and other beneficiaries, and the dissemination of preliminary findings in draft research papers and other research outputs. It will provide comparative measures over long time periods where none exist, to date.
Issue-competence indices will be estimated using Stimson's (1991) dyad ratios algorithm. This method extracts common variance across survey item and policy issue, controlling for potential bias from sampling error and variation. The approach has been used to generate aggregate estimates of public left-right policy preferences (Stimson 1991; Erikson et al 2002; Bartle et al. 2010) and common variance in other attitudes (e.g. Durr et al. 1997; 2000; Freeman et al. 1998; Chanley et al. 2000; Kellstedt 2000; Erikson et al. 2002; Keele 2005; 2007). It allows researchers to use existing data observations from different sources, overcoming missing values. The applicants developed such an index of perceived issue competence in the UK between 1950 and 2008 using 2,383 survey items from opinion polls and election studies. This work will be extended in the UK and for each party (and/or president) in the four countries. The countries are selected for comparability of clear party government and opposition, and for comparison with a presidential system of shared power in the U.S, and also for necessary English-language data and covariates. A feasibility study (£2,945) showed that the selected countries provide a good initial sample.
The following research questions will be explored.
1. Does issue competence aide our explanation of vote intention and turnout crossnationally, and does it improve on existing explanatory models?
2. What country-level characteristics account for variation observed in issue competence effects (e.g. presidential approval, strength of economic vote, partisanship etc.)?
3. Can macro-competence help explain variations in partisanship as a running tally? Are there common causal relationships across country?
4. How have key political events transformed macro-competence levels? Do they exhibit similar persistence across countries?
5. Do left-right shifts in public policy preferences (Erikson et al. 2002; Bartle et al. 2010) follow competence declines, separable from economic conditions and evaluations, etc.?