Ideological Signals of German Parties in a Multi-Dimensional Space: An Estimation of Party Preferences Using the CMP Data

Abstract

In this paper, we present a method based on the CMP data to identify parties' ideological signals. Although our approach uses the same data as others, our method is different in that we distinguish the methods for the directional and proximity models which have long been relevant for the theoretical literature. We apply our method for all German Bundestag parties between 1949 and 2005. One problem with policy preferences is that various – and sometimes only slightly deviating – approaches lead to rather different results. Thus, we test our method in terms of its robustness using the Jackknife method. We also show confidence intervals around policy positions.

Notes

A. Downs, An Economic Theory of Democracy (New York: Harper, 1957).

M. Laver and K.A. Shepsle, Making and Breaking Governments. Cabinets and Legislatures in Parliamentary Democracies (Cambridge: Cambridge University Press, 1996).

N.J. Schofield, ‘Existence of a “Structurally Stable” Equilibrium for a Non-collegial Voting Rule’, Public Choice 51/3 (1986), pp.267–84.

I. Sened, ‘A Model of Coalition Formation: Theory and Evidence’, Journal of Politics 58/2 (1996), pp.350–72.

Most notably the work of M. Laver and W. Ben Hunt, Policy and Party Competition (New York: Routledge, 1992); and K. Benoit and M. Laver, Party Policy in Modern Democracies (London: Routledge, 2006).

See M. Laver and J. Garry, ‘Estimating Policy Positions from Political Texts’, American Journal of Political Science 44/3 (2000), pp.619–34; J.B. Slapin and S.-O. Proksch, ‘A Scaling Model for Estimating Time Series Policy Positions from Texts’, American Journal of Political Science 52/3 (2008), pp.705–22.

See I. Budge, H.-D. Klingemann, A. Volkens and J. Bara, Mapping Policy Preferences: Estimates for Parties, Electors, and Governments 1945–1998 (Oxford: Oxford University Press, 2001).

Laver and Gary, ‘Estimating Policy Positions from Political Texts’.

F.U. Pappi and S. Shikano, ‘Regierungsabwahl ohne Regierungsneuwahl’, Politische Vierteljahresschrift 46/4 (2005), pp.513–26; M. Debus, ‘Office and Policy Payoffs in Coalition Governments’, Party Politics 14/5 (2008), pp.515–38.

Laver and Hunt, Policy and Party Competition.

Benoit and Laver, Party Policy in Modern Democracies.

Slapin and Proksch, ‘A Scaling Model for Estimating Time Series Policy Positions from Texts’.

See for this topic also S. Hans and C. Hönnige, ‘Noughts and Crosses. Challenges in Generating Political Positions from CMP-Data’, Kaiserslautern Occasional Papers in Political Science No. 2/2008.

I. Budge, D. Robertson and D. Hearl, Ideology, Strategy and Party Change: Spatial Analysis of Post-War Election Programmes in 19 Democracies (Cambridge: Cambridge University Press, 1987).

F.U. Pappi and S. Shikano, ‘Ideologische Signale in den Wahlprogrammen der deutschen Bundestagsparteien 1980 bis 2002’, Working Paper No. 76 at the Mannheim Centre for European Social Research, 2004.

Because the term ‘social policy’ can be interpreted in different ways, we make clear that we use this notion in the sense of ‘policy concerning the social order’ (including questions of justice, law and order, liberty as regards minorities, and so on), not with respect to social systems, which are a subfield of economic policy.

For a historical overview on the Manifesto Research Group and the CMP, see A. Volkens, ‘The Manifesto Research Group since 1979 with Particular Reference to Quality Control’, in M. Laver (ed.), Estimating the Policy Position of Political Actors (London: Routledge, 2001), pp.33–49.

The other manifesto categories could not clearly be identified as ‘left’ or ‘right’. Some of them are rather neutral or pure valence issues; others are too diffuse to classify them as ‘left’ or ‘right’.

H.-D. Klingemann and A. Volkens, ‘Struktur und Entwicklung von Wahlprogrammen in der Bundesrepublik Deutschland 1949–1998’, in O.W. Gabriel, O. Niedermayer and R. Stöss (eds.), Parteiendemokratie in Deutschland (Opladen: VS Verlag für Sozialwissenschaften, 1997), pp.507–27; S. Franzmann and A. Kaiser, ‘Locating Political Parties in Policy Space. A Reanalysis of Party Manifesto Data’, Party Politics 12/2 (2006), pp.163–88.

See for example Laver and Hunt, Policy and Party Competition.

We use ‘left’ and ‘right’ in the usual sense: With regard to economic policy, a right-wing position stands for laissez-faire and a state as weak as possible, whereas as a left-wing position favours regulation and the expansion of public welfare systems. In social policy, a right-wing position stands for law and order, cultural conservatism, and traditional values; a left-wing position highlights liberalism, post-materialist values, and open-mindedness for individual lifestyles. Thus, the terms ‘left’ and ‘right’ have different meanings in the various policy dimensions. We only use these words to avoid overlong descriptions of the axes' poles, knowing that the meaning of ‘left’ and ‘right’ might be disputed. For a full list of party accronyms please see Appendix 2.

Cf. the salience theory of Budge et al., Ideology, Strategy and Party Change.

Downs, An Economic Theory of Democracy.

In some cases, there were single quasi-sentences which could not be classified to one of the 56 manifesto categories for substantive reasons. For Germany, this is a minor problem, because the categories are always able to cover over 95 per cent of every manifesto (with two expectations: 1949, SPD: 93.9 per cent; 1949, DP: 85.7 per cent). After 1949, they capture more than 99 per cent of the quasi-sentences (again with two expectations: 1953, DP: 97.6 per cent and 1987, Green Party: 97.8 per cent). However, for our approach uncoded quasi-sentences are not a problem, because we consider a macro-category ‘others’ (Table 1) in which these quasi-sentences fall.

For the interpretation of the terms ‘left’ and ‘right’ see note 21.

For other approaches, e.g. with regard to legislative decisions, it is more helpful to use meso categories by grouping the categories according the single cabinet members' competencies.

Pappi and Shikano, ‘Ideologische Signale in den Wahlprogrammen der deutschen Bundestagsparteien 1980 bis 2002’.

Laver and Hunt, Policy and Party Competition.

This distinction is similar to the classification in valence issues and positional issues of Pappi and Shikano, ‘Ideologische Signale in den Wahlprogrammen der deutschen Bundestagsparteien 1980 bis 2002’, and Franzmann and Kaiser, ‘Locating Political Parties in Policy Space’. Our definition, however, deviates from placing manifesto categories within the macro-category ‘others’, because this category cannot be interpreted as policy dimension. Thus, it would not make sense to code and sum up this macro-category's manifesto issues.

A. Volkens, ‘Manifesto Coding Instructions’, paper presented at the Wissenschaftszentrum Berlin für Sozialforschung (2nd revised edition), 2002, p.127.

For some categories, their classification might be inappropriate with regard to their naming on the first view. The second view which considers the detailed descriptions of the manifesto categories includes various experts we asked, who came to the same results (see for example the discussion on category 403). However, we later tested our method for robustness; we included error terms with respect to single debatable classifications (cf. the application of the Jackknife method as described below).

This categorisation can easily be adapted for other purposes (with a finer classification on the meso-level) or political systems (with further relevant policy dimensions if existing).

In the narrower sense of Downs, An Economic Theory of Democracy, and applied, for example, by G. Tsebelis, ‘Decision Making in Political Systems: Veto Players in Presidentialism, Parliamentarism, Multicameralism and Multipartyism’, British Journal of Political Science 25/3 (1995), pp.289–325.

E.g. S.A. Matthews, ‘A Simple Direction Model of Electoral Competition’, Public Choice 34/2 (1979), pp.141–56; I. Budge, ‘A New Spatial Theory of Party Competition: Uncertainty, Ideology and Policy Equilibria Viewed Comparatively and Temporally’, British Journal of Political Science 24/3 (1994), pp.443–67.

The product of two sign functions is not necessarily 1 since both functions refer to different variables.

The addition of 1 is necessary to map the zero onto itself, since ln(1) = 0.

For the relationship between positional and directional preferences see E. Linhart and S. Shikano, ‘A Basic Tool Set for a Generalized Directional Model’, Public Choice 140/1 (2009), pp.85–104.

Contrary to the estimation of ideal directions and positions, there is no reason to logarithmise the manifesto categories' frequencies for the estimation of interest weightings. Thus, here we use the rough frequencies. If, as in our application, a remaining category exists, the salience values should be renormalised for application in which only the main policy dimensions are considered.

As assumed, for example, by G. Rabinowitz and S.E. Macdonald, ‘A Directional Theory of Issue Voting’, American Political Science Review 83/1 (1989), pp.93–121.

Cf. Matthews, ‘A Simple Direction Model of Electoral Competition’.

Or only as angles, if one follows Matthews (ibid.).

This relationship also holds if we exclude the splinter parties of the early phase.

See also E. Linhart and S. Shikano, ‘Die Generierung von Parteipositionen aus vorverschlüsselten Wahlprogrammen für die Bundesrepublik Deutschland (1949–2002)’, Working Paper No. 98 at the Mannheim Centre for European Social Research, 2007; K. Benoit, M. Laver and S. Mikhaylov, ‘Treating Words as Data with Error: Estimating Uncertainty in the Comparative Manifesto Project Measures’, American Journal of Political Science 53/2 (2009), pp.495–513. For the bootstrap technique, see B. Efron and R. Tibshirani, An Introduction to the Bootstrap (New York: Chapman and Hall, 1993); S. Shikano, ‘Bootstrap und Jackknife’, in J. Behnke, T. Gschwend, D. Schindler and K.-U. Schnapp (eds.), Methoden der Politikwissenschaft (Baden-Baden: Nomos, 2006), pp.69–80.

The detailed results are available in Linhart and Shikano, ‘Die Generierung von Parteipositionen aus vorverschlüsselten Wahlprogrammen für die Bundesrepublik Deutschland (1949–2002)’.

See Linhart and Shikano, ‘A Basic Tool Set for a Generalized Directional Model’.

For an empirical test of a coalition theory based on this policy space, see S. Shikano and E. Linhart, ‘Coalition Formation as a Result of Policy and Office Motivations in the German Federal States: An Empirical Estimation of the Weighting Parameters of Both Motivations’, Party Politics (forthcoming).

For example, Laver and Garry, ‘Estimating Policy Positions from Political Texts’; Slapin and Proksch, ‘A Scaling Model for Estimating Time-Series Party Positions from Texts’.