«Notes and Comments The Contradictory Effects of Consensus Democracy on the Size of Government: Evidence from the Swiss Cantons ADRIAN VATTER MARKUS ...»
The second indicator is Rae’s index of fragmentation of the party system,21 computed on the basis of the different parties’ shares of seats in the cantonal parliaments. Our calculations rely on an annual Swiss publication.22 In these two measures we have included the most pivotal variables of the ﬁrst dimension of Lijphart’s concept of consensus democracy in our analysis.23 Consensus democracy of the second dimension (competitive veto points; institutional veto player) has been translated into three measurable and observable variables. The degree of decentralization is measured by an indicator of ﬁscal centralization developed by analogy with Lijphart’s measure and also used by Castles, Lane/Ersson, Keman and Schmidt.24 It denotes the tax revenue of the canton as a percentage of the total tax revenue of the canton and the municipalities.
The tax-share measure is based on the reasonable assumption that the scope of the activities of the central (cantons) and non-central government (municipalities) can be measured in terms of their revenues. The corresponding data was provided by the Federal Finance Administration. The inﬂuence of direct democracy on the size of public sector is speciﬁed by two variables. First, the formal institutions of direct democracy are measured via the index of ﬁnancial referendums deﬁned by Stutzer and Frey,25 which includes the number of signatures required to be collected, the deadline (F’note continued) of observations, only one was calculated for the whole estimate. In addition, the values of the Variance Inﬂation Factor (VIF) obtained in the tests are tolerable, so that multicollinearity is not a cause for concern.
As the division of power between the cantonal and municipal level varies from canton to canton, the data has been aggregated for the sake of comparability (Christoph A. Schaltegger, ‘Ist der Schweizer Foderalismus ¨ zu kleinraumig?’, Swiss Political Science Review, 7 (2001), 1–18). The ﬁgures denote per capita spending after ¨ deducting the contributions from the federation, as quoted by the functional division of the Swiss Federal Statistical Ofﬁce (Swiss Federal Statistical Ofﬁce, Statistisches Jahrbuch der Schweiz (Bern/Zurich: NZZ Verlag, various ¨ volumes).
Klaus Armingeon, ‘Konkordanzzwange und Nebenregierungen als Handlungshindernisse’, Swiss Political ¨
Science Review, 2 (1996), 277–303; Lijphart, Patterns of Democracy; Manfred G. Schmidt, ‘When Parties Matter:
A Review of the Possibilities and the Limits of Partisan Inﬂuence on Public Policy’, European Journal of Political Research, 30 (1996), 155–83; Schmidt, Demokratietheorien.
Markus M. L. Crepaz, ‘Consensus Versus Majoritarian Democracy: Political Institutions and their Impact on Macroeconomic Performance and Industrial Disputes’, Comparative Political Studies, 29 (1996), 4–26, p. 9;
Lijphart, Patterns of Democracy, pp. 65ff.
Douglas W. Rae, ‘A Note on the Fractionalization of Some European Party Systems’, Comparative Political Studies, 1 (1968), 413–18.
Annee politique Suisse, Schweizerische Politik (Bern: Institute of Political Science, 1990ff).
´ Lijphart, Patterns of Democracy, pp. 62f.
Lijphart, Democracies; Lijphart, Patterns of Democracy, p. 193; Francis G. Castles, ‘Decentralization and the Post-War Political Economy’, European Journal of Political Research, 36 (1999), 27–53; Jan Erik Lane and Svante O. Ersson, ‘Is Federalism Superior?’ in Bernard Steunenberg and Frans A. van Vught, eds, Political Institutions and Public Policy: Perspectives on European Decision Making (Dordrecht: Kluwer Academic Publishers, 1997), pp. 85–113; Hans Keman, ‘Federalism and Policy Performance’, in Ute WachendorferSchmidt, ed., Federalism and Political Performance (London: Routledge, 2000), pp. 196–227; Schmidt, Demokratietheorien.
Alois Stutzer and Bruno Frey, ‘Starkere Volksrechte – Zufriedenere Burger: eine mikrookonometrische
¨ ¨ ¨Untersuchung fur die Schweiz’, Swiss Political Science Review, 6 (2000), 1–30, p. 25.
¨ Notes and Comments 363 for submitting them and the level of expenditure, which can be challenged in a referendum. The authors claim that the index reﬂects the barriers to direct citizen involvement in ﬁscal matters that exist in the different cantons. However, there is evidence that the existence of formal rights of citizen participation does not necessarily mean that these rights will be exercised.26 Therefore, we examine the effect of the use of the instruments of direct democracy in practice alongside formal institutional design. The variable of the actual use of direct democracy denotes the annual number of ﬁnancial referendums in a canton. The data for both variables is based on Moser and the Schweizerische Politik.27 In addition to the above, we use prominent theories in cross-national public policy research to identify the relevant control variables. The class and partisan approach stresses the importance of the party composition of the government. Therefore, the ﬁrst control variable is the percentage of left-wing parties (social democrats and greens) in the cantonal governments, calculated on the basis of the data in Schweizerische Politik.28 The hypothesis of socio-economic determination puts the following variables at the centre of the analysis: the degree of urbanization (proportion of inhabitants living in urban areas), the unemployment rate and the population aged over 64 as a proportion of those between 20 and 64 years. The source of all these ﬁgures is the Swiss Federal Statistical Ofﬁce.29 Moreover, we include the annual percentage growth of the economy in the calculations. The socio-cultural dimension, which is relevant in the context of the Swiss cantons, is represented by a language variable.30 Finally, a comparison of sub-national units needs to take
Institutions of direct The easier the access to Index of ﬁnancial Negative democracy the right of ﬁnancial referendums by Stutzer referendum, the smaller and Frey (low values will be the public sector. correspond to high institutional barriers)
Cf. Markus Freitag and Adrian Vatter, ‘Direkte Demokratie, Konkordanz und wirtschaftliche Leistungskraft:
Ein Vergleich der Schweizer Kantone’, Swiss Journal of Economics and Statistics, 136 (2000), 579–606, p. 603.
Christian Moser, Abstimmungen, Initiativen und fakultative Referenden in den Kantonen (Bern: FSP, 1990ff); Annee politique suisse, Schweizerische Politik.
´ Annee politique suisse, Schweizerische Politik.
´ Swiss Federal Statistical Ofﬁce, Statistisches Jahrbuch der Schweiz.
The results of federal votes have recurrently shown that the citizens in the German-speaking part of Switzerland tend to prefer a liberal and subsidiary state, while French and Italian speakers approve a more statist model. We expect a negative correlation between the proportion of German speakers and the extent of the public sector in a canton.
364 Notes and Comments
into account the transfers carried out in the context of the redistribution of income within the federation. Based on data from the Federal Finance Administration, we therefore include the real per capita transfers from the federation to the cantons and municipalities.
All the variables have been compiled on an annual basis. For reasons of causality and for theoretical considerations, the explanatory variables are assumed to take effect after a lag of one period, provided that they vary over time. Table 1 gives an overview of the different variables, their measurements and the correlations that are to be expected.
We have estimated two models in order to test the impact of the potential determinants on differences in public sector size.
Table 2 reports the following results for the period from 1990 to 2000: in principle, we have corroboration of the postulated hypothesis that different aspects of consensus democracy have opposite effects on the size of government. The inﬂuence of the central factors, i.e. executive power-sharing, party-system fragmentation, decentralization and direct democracy, goes in the expected direction and is signiﬁcant in terms both of revenue and of expenditure in almost all estimations. Extensive rights of direct citizen participation and a decentralized state structure have a curbing effect on the public sector, whereas collective veto points (e.g. a multiparty system and a grand coalition) lead to state expansion.
Notes: Non-standardized regression coefﬁcients with panel-corrected standard errors in parentheses; *Signiﬁcant at the 10% level (two-tailed test); **Signiﬁcant at the 5% level (two-tailed test); ***Signiﬁcant at the 1% level (two-tailed test).
366 Notes and Comments In three of four cases, the instruments of direct democracy clearly conﬁne the public sector, measured in terms of tax revenue and public expenditure, in a statistically signiﬁcant manner. The more difﬁcult it is to launch a ﬁnancial referendum and the less frequently this is actually done, the higher public expenditure will be. Policies for the beneﬁt of one particular social group and that entail high expenditure tend to fail in popular votes, given the ﬁscally conservative preferences of the majority of the electorate. However, we do not see the Robin Hood effect arising from strong (re)distributive preferences in the electorate in our results.31 Contrary to the postulated positive association between direct citizen involvement and tax rates, state revenue is negatively correlated with the right to hold a ﬁscal referendum.
With respect to decentralization, its restraining inﬂuence on policy makers is also conﬁrmed. The more decentralized the cantons are in ﬁscal terms, the smaller the public sector will be. State revenue and expenditure are evenly affected by ﬁscal federalism: they both decrease with an increase in municipal autonomy.
Unlike the vertical division of power, the horizontal division of power inherent in consensus arrangements promotes state expansion. Both indicators of the collective veto points have a signiﬁcant impact on state revenue and on state expenditure in this analysis.32 In this vein, both executive power sharing and the mechanism of extensive logrolling in multiparty legislatures boost the size of the public sector in the Swiss cantons. With regard to the partisan composition of government, an increase in the share of left-wing parties in government does not lead to more interventionism, either on the expenditure or on the revenue side. As to the socio-economic variables, they seem to play a major role. Thus, the age structure of the population and urbanization are important determinants of government revenue and spending. Government expenditure, in particular, depends on further socio-economic factors, namely on economic growth and on the unemployment rate. Finally, monetary transfers from the federation turn out to be signiﬁcant on both sides of the cantonal account.
This research note’s starting point was the hypothesis that the two dimensions of consensus democracy have opposite effects on the size of government. This hypothesis has been substantiated with respect to the Swiss cantons in the period between 1990 and 2000. Essentially, we have found that institutional barriers to majority rule and features of multipartism are two distinct dimensions of consensus democracies that have contradictory effects on the size of the public sector.
Elements that are also termed ‘competitive veto points’ (or ‘institutional veto players’), such as decentralization and direct democracy, fetter the state not only in terms of revenue but also of expenditure. By contrast with studies of veto player theory, collective veto points (or partisan veto players), such as oversized government coalitions or multiparty legislatures, tend to facilitate public sector expansion. In general, our ﬁndings demonstrate that consensus democracies comprise two kinds of veto points, both of which signify power diffusion and which have opposing effects on government action.
Cf. Uwe Wagschal, ‘Direct Democracy and Public Policymaking’, Journal of Public Policy, 17 (1997), 223–45, p. 224.
Cf. Vicki Birchﬁeld and Markus M.L. Crepaz, ‘The Impact of Constitutional Structures and Collective and Competitive Veto Points on Income Inequality in Industrialized Democracies’, European Journal of Political Research, 34 (1998), 170–200; Lijphart, Patterns of Democracy; Manfred G. Schmidt, Sozialpolitik in Deutschland: Historische Entwicklung und internationaler Vergleich (Opladen: Leske and Budrich, 1998).
Notes and Comments 367
Note: The higher the index of ﬁnancial referendums, the lower the barriers for citizens entering the political process. The higher the index of government coalition, the more a given canton ﬁts consensus democracy criteria.