Policy Moods and Institutions – A Comparative Approach

Abstract

Our aim is to bring together scholars who have been working on ‘policy mood’ to: 1. Agree standards relating to the collection of data and discuss the creation of a new data resource similar to the Manifest Project Database. 2. Compare methods of estimating the policy mood. 3. Assess the impact of political institutions (inter alia political parties, electoral systems, constitutions, welfare systems etc.) on the policy mood. 4. Assess the impact of the policy mood on institutions (especially governments, political parties and elections). 5. Compare findings with MARPOR and Policy Agendas Projects. The aim of this research session is to establish a comparative framework for several stand-alone ‘policy mood’ research projects. We will also produce a book on Moods and institutions –a comparative approach to be submitted either to ECPR Press or the OUP Comparative Politics series. The concept of policy moods Public preferences over controversial issues have an observable tendency to move together over time. This empirical observation naturally leads to the inference that there are one (or more) unobservable ‘factors’, ‘dimensions’ or `policy moods’ that account for this common longitudinal variation. US studies suggest that by far the most important single factor relates to disagreements about government activity. This dimension is labelled `the policy mood’ or simply ‘the mood’ (Stimson, 1999). In the US, the structure of aggregate public preferences broadly reflects elite level disagreements between ‘liberals’ (who advocate greater activity) and `conservatives’ (who advocate less government activity) (Stimson, 1999). Since parties tend toward consistency, these ideological differences are transformed into institutional disagreements between parties, with Democrats advocating more and the Republicans advocating less government (Carmines and Stimson, 1989). In the US, at least, there is little evidence of a clear secondary dimension. And, whatever dimension there is, does not appear to be the assumed ‘social’ or ‘post material’ dimension’ (Stimson, 2004). Studies have used estimates of the policy mood and candidates’ positions to test spatial models of presidential elections (Erikson et al., 2002). The ‘distance’ between presidential candidates and the electorate exerts a powerful impact on the outcomes of presidential elections, even controlling for macropartisanship. The same studies demonstrate that the policy mood has a ‘thermostatic’ relationship with policy, moving in the opposite direction (Soroka and Wlezien, 2009). Policymakers, furthermore, both respond to and anticipate changes in public preferences (Erikson et al. 2002). The mood is, therefore, central to studies of party competition and political representation. The US findings have been broadly replicated in studies of the British (Bartle et al., 2011), French (Stimson et al., 2012), German (Weiss, 2011) and Spanish publics (Bartle et al. 2014), though scholars have used the terms of ‘left’ and ‘right’ to label the policy mood. The average preference on the policy mood is as an indicator of the ‘political centre’. The political centre appears to move in the opposite direction to government policy and have an association with election outcomes (Bartle et al, 2011, Bartle et al, 2014). Surprisingly, in all three cases, there is little evidence yet of a distinct ‘social dimension’. Another aim of this session will be to develop strategies for further assessing the dimensionality of public preferences and make these estimates available. Policy mood and institutions The ‘minimal model’ of public opinion suggests that individuals know little about the detail of politics and have little incentive to acquire information. If this characterisation of the typical voter is correct, it is difficult to understand how (and indeed why) their decisions do (or should) guide policy. One response is to assume that voters use various heuristics to make reasonable choices (Popkin, 1994). Another response is to suggest that institutions – particularly the political parties – simplify things for the electorate (Sniderman, 2000, p. 75). Parties – pursuing their own interests of winning votes – take positions on seemingly unrelated issues and into coherent policy packages, determining which positions ‘hang together’, can be coded as ‘left’ or ‘right’ and subsequently averaged to produce an indicator of the political centre. As new issues emerge, the parties take different positions. In this way, new issues are folded into the existing ‘liberal-conservative’ or ‘left-right framework’ (Carmines and Stimson, 1989). The relationship between policy moods, dimensions or factors and institutions raises some classic ‘chicken and egg’ issues about causation. ‘[I]nstitutions may be thought to embed history and political thought’ (Rhodes et al., 2006, p. xiii). Accordingly, accounts of institutions should also begin with a theory of preferences. The formation of socialist parties that advocated a massive expansion in government activity in the late nineteenth century, for example, compelled all other parties to take a position on this ‘super issue’ (Bara and Weale, 2006). The institutions of public schools, national health services, welfare and regulatory states all reflect the preferences of past generations. In short, preferences preceded the creation of institutions and were subsequently embedded within the institutions of the political parties (Mair, 1997). The MARPOR studies of party programmes have suggested that political parties can be located on a single left-right dimension using ‘RILE’ scores (Budge et al., 2001; Klingemann et al., 2006). These studies have massively increased our understanding of party competition, representation and elections can be learned by adopting a comparative approach (Budge, 1994). Yet they have also been hampered by the fact that there have been no public opinion-based analogues of RILE scores that would enable analysts to assess the parties (and governments) responsiveness and the responsiveness of preferences to policy. This workshop will seek to close the gap by assessing the functional equivalence of the policy mood and RILE indicators.   It is proposed that the workshop will address the following issues: 1. Standards of collation and coding protocols Estimation of the mood requires repeated measurement of preferences in nationally representative surveys in order to assess longitudinal change. This requirement is simply stated but raises a number of questions that will be addressed at the workshop: A. What to do about seemingly trivial variations in question wording: ignore and preserve individual series or start new series and lose continuity? B. What to do about identical questions posed by different organisations, using different samples and survey modes? C. How to deal with non-substantive responses? D. Whether the middling or seemingly neutral responses are really positions? E. The inferential consequences of errors about the polarity (coding left as right etc.)? The workshop can also consider the requirements for making bi-annual, quarterly or monthly estimates of the policy mood in order to tie in with conventional studies of vote intentions and/or government approval. 2. Methods of estimation Until recently, the only way of estimating the policy mood (or moods) was the dyads ratio algorithm (Stimson: 1999). More recently, Anthony McGann has recently developed an alternative estimator based on item response theory (IRT) that in large measure validates the dyads ratio algorithm estimates (McGann, 2014). The IRT estimator has several desirable properties – not least the fact that it is based on a clear micro-level theory of survey response. Nevertheless, there is still some work required in order to explain the IRT approach to a wider set of scholars, establish the equivalence between these methods and also between these methods and more widely used methodologies, such as principal components and factor analyses (which are appropriate where there is either no or very little missing data). We will also address whether the methods should be viewed as ‘exploratory’ in the sense of making no a priori assumptions about the dimensionality of public preferences or should as ‘confirmatory’ and assume a dimensional solution. The workshop will also compare estimates of the policy mood using both the dyads ratio algorithm and IRT estimator the estimates of the ‘median voter’ that are based on vote shares and RILE scores (Kim and Fording, 1998). 3. The impact of institutions on the policy mood The relationship between institutions and policy moods is a reciprocal one. The workshop will examine both causal arrows. The order that they are dealt with in this proposal does not pre-suppose the greater importance of either causal arrow: (a) Electoral systems The electoral system is important because – together with the social structure – it influences the number of political parties within a system. In general, the more proportional systems tend to produce more parties. Nevertheless, theories of issue evolution (Carmines and Stimson, 1989) and agent-based models (McGann, 2008) suggest that parties reduce the dimensionality of politics. To date, the policy mood has been estimated mostly for majority systems. In countries where there is not regular and complete alternation of governments, the dynamics of the policy mood may differ from the expectations of the thermostatic models. This is something that needs to be tested in a genuinely multi-party system, such as the Netherlands. There is a literature that compares the responsiveness to public opinion of various electoral systems (Powell 2000; McDonald and Budge 2005). McGann and Latner (2013) argue that institutions cannot cause outcomes, but can only channel public opinion. They provide a model of how different institutions do this, but cannot fully test it due to the fact that they do not have a consistent measure of public opinion over time. Indeed, nearly all the literature on the effect of institutions faces the same problem. The policy mood provides a measure of public opinion that can be used to assess the (sometimes contradictory) claims of these literatures. (b) Political parties As we have noted, political parties simplify politics by developing coherent policy programmes. They also incorporate new issues by taking opposing ‘positions’. The workshop will consider examples of how parties responded to ‘new issues’ (e.g., membership of the EU), incorporated them in their programmes and thereby allow the electorate and analysts to code new issues within familiar frameworks. The workshop may consider why members of party families take similar positions across a large number of countries and, in those cases where this does not happen, the implications for comparative indicators of the policy mood. (c) Sub-national policy moods To date, studies of the policy mood have focussed on the national mood, relying on preference data from nationally representative studies. We will give some consideration to estimating the policy mood(s) in regions that enjoy a significant degree of autonomy (e.g., Scotland and Catalonia) in order to summarise intra-national differences and assess the influence of constitutional characteristics (e.g., federal or unitary systems) on the policy mood (Soroka and Wlezien, 2009; Dellepiane et al., 2014). 4. The impact of mood on institutions, election outcomes and public policy Existing studies strongly suggest that changes in the policy mood are either responded to or actually anticipated by policy-makers. Nevertheless, political institutions modify the impact that the policy mood has on policy, since institutions articulate or frustrate demands from the public and political parties. (a) Electoral systems There is a very considerable and influential literature that argues that different institutions lead to certain political economic outcomes. For example, it is commonly argued that PR electoral systems lead to higher government spending (Persson and Tabellini 2003; Iversen and Soskice 2006; Crepaz 2001, amongst others). There is also a literature which suggests that institutions that produce multiparty government leads to veto players and policy immobilism (Tsebelis 2002). This can be tested once we have measured public preferences in a range of countries. (b) Political parties To date, the absence of a measure of the electorate’s preferences that is the equivalent of RILE scores has made it difficult to assess the responsiveness of parties and electorate to each other. The policy mood provides exactly the sort of indicator that allows us to test whether party positions and policy respond to the electorate or whether the electorate follow the parties as is implied by ‘preference shaping’ models (Dunleavy and Ward, 1981). 5. The relation between mood and MARPOR/Political Agendas Projects This workshop will recognise that there are already two major programmes that deal with some of the relevant issues and will draw comparisons with the approaches and findings. In particular, the workshop will consider: A. The relationship between party positions as indicated by MARPOR analyses of party programmes (especially the RILE scores) and the policy mood. In particular: (a) Whether the estimates of policy mood can be made directly comparable with the RILE scores by including only those measures of preferences that reflect the topics included in RILE scores (Budge et al., 2001). (b) Methods of recalibrating mood and RILE scores to produce indicators of the spatial distance between parties (candidates) and the electorate (Downs, 1957). B. Consider the relationship between the policy mood, RILE scores and the various measures of the ‘political agenda’ of governments, parties and the electorate (Jennings and Wlezien, 2014). Participants This proposal falls in ‘group b’ category of the established proposals for the research sessions, since the policy mood research agenda has been operational for some time now. It is also worth saying that it builds on previous ECPR events since it was envisaged in a panel at the Glasgow ECPR General Conference (P343 The Comparative Macro Polity: Policy Agendas and Policy Moods). A more distant meeting also took place during the Pisa ECPR General Conference 2007 (Panel PN197, Macro polity-I). The number of participants in the research session is expected to reach the predetermined maximum of six. Prospective participants come from a range of countries including: UK (Three from) John Bartle (Essex), Anthony McGann (Strathclyde), Sebastian Dellepiane-Avellaneda (Strathclyde), Will Jennings (Southampton) US James Stimson (North Carolina, Chapel Hill)* Italy Paolo Bellucci (Siena) Spain Agustí Bosch (UAB) or LLuis Orriols (Madrid) * Stimson has published research on the US, UK and France. Those named who cannot attend have asked the participants to provide a report on discussions. We will establish a discussion group with a notice board. All prospective participants come from ECPR member institutions.   References Bartle, John, Agustí Bosch, and Lluís Orriols. 2014. The Spanish policy mood, 1978-2012. Paper presented to the 8th ECPR General Conference. University of Glasgow 3-6 September. Bartle, John, Sebastian Dellepiane-Avellaneda, and James A. Stimson. 2011. The moving centre: preferences for government activity in Britain, 1950-2005, British Journal of Political Science, 41: 259-85. Budge, Ian, HD Klingemann, Andrea Volkens, Judith Bara, Eric Tanenbaum. 2001. Mapping policy preferences: Estimates for parties, electors & governments 1945–1998. Oxford: Oxford University Press. Carmines, Edward and James A. Stimson. 1989. Issue evolution: Race and the transformation of American politics. Cambridge: Cambridge University Press. Crepaz, Markus. 2001. Veto players, globalization and the redistributive capacity of the state: A panel study of 15 OECD Countries. Journal of Public Policy 2: 1-22. Dunleavy, Patrick and Hugh Ward. 1981. Exogenous Voter Preferences and Parties with State Power: Some Internal Problems of Economic Theories of Party Competition. British Journal of Political Science 11: 351-80. Erikson, Robert S., Michael MacKuen, and James A. Stimson. 2002. The macro polity. New York: Cambridge University Press. Iversen, Torben, and David Soskice. 2006. Electoral institutions and the politics of coalitions: Why some democracies redistribute more than others. American Political Science Review 100: 165-181. Kim, Heemin and Richard C. Fording, Voter ideology in western democracies, 1964-89. 1998. European Journal of Political Research, 33: 73-97. Klingemann, Hans Dieter, Andrea Volkens, Judith Bara, Ian Budge, Michael D. MacDonald. 2006. Mapping policy preferences II: Estimates for parties, electors, and governments in Eastern Europe, European Union, and OECD 1990-2003. Oxford: Oxford University Press. Mair, Peter. 1997. 1997. Party System Change: Approaches and interpretations. Oxford: Oxford University Press. McDonald, Michael, and Ian Budge. 2005. Elections, parties, democracy: Conferring the median mandate. Oxford: Oxford University Press. McGann, Anthony. 2008. Why uidimensionality? Party Systems as Constraints on Choice. Paper presented to the MPSA Conference. Chicago, 2-6 April. McGann, Anthony J. 2014. Estimating the political center from aggregate data: An item response theory alternative to the Stimson dyad ratios algorithm. Political Analysis 22: 115-129. McGann, Anthony, and Michael Latner. 2013. The calculus of consensual democracy: Rethinking patterns of democracy without veto players. Comparative Political Studies 46: 823-50. Persson, Torsten, and Guido Enrico Tabellini. 2003. The economic effects of constitutions, Munich lectures in economics. Cambridge, Mass.: MIT Press. Powell, G. Bingham Jr. 2000. Elections as instruments of democracy: Majoritarian and proportional visions. New Haven: Yale University Press. Sniderman, Paul M. 2000. Taking sides: a fixed choice theory of political reasoning. In Elements of reason: Cognition, choice and the bounds of rationality, Arthur Lupia, Matthew D. McCubbins and Samuel L. Popkin, eds., Cambridge: Cambridge University Press. Soroka, Stuart and Christopher Wlezien. 2009. Degrees of democracy: Politics, public opinion and policy. Cambridge: Cambridge University Press. Stimson, James A. 1999. Public opinion in America: moods, cycles, and swings, Transforming American politics. Boulder: Westview Press. Stimson, James A. 2004. Tides of consent: how public opinion shapes American politics. New York: Cambridge University Press. Stimson, James A., Cyrille Thiebaut and Vincent Tiberj. 2012. The evolution of policy attitudes in France. European Union Politics, 13: 293-316. Tsebelis, George. 2002. Veto players: How political institutions work. New York: Russell Sage Foundation / Princeton University Press. Weiss, Steffen. 2011. The German macro polity: Essays on party identification, policy preferences, and elections. PhD thesis University of Essex, Colchester.