Voting Patterns of the parliamentary opposition in Germany Christian Stecker University of Potsdam http://www.uni-potsdam.de/db/vergleich/?page_id=88 Paper prepared for the meeting of the Italian Political Science Association in Palermo, 2011 -- very first draft -- I am indebted to Johannes Kuhn and Tristan Klinglhöfer for excellent research assistance.
Introductory Remarks & Theoretical Considerations After the elections in 2005 the two major parties in Germany CDU/CSU and SPD formed a grand coalition controlling almost 75 per cent of the seats in the German Bundestag. The new government faced an opposition (Free Democrats (FDP), the Greens and the Left Party) that was exceptional in two important regards. First this opposition exhibited a huge ideological diversity with the FDP propagating strong free-market ideas and a Left Party that still carries the plea for a democratic socialism in its party manifesto. Second no parliamentary opposition since the late 1960s has been as powerless. Even when voting in unison the three opposition parties failed to reach the 1/3-majority hurdle to invoke important minority rights such as issuing an appeal to the constitutional court or scheduling an extraordinary session of the Bundestag. 1 Six years later in 2011 the situation on the benches of the opposition could barely be more different. An ideological homogenous opposition bloc of SPD and Greens (and a more remote Left Party) now confronts the conservative government of CDU/CSU and FDP. Furthermore, since CDU/CSU and FPD lost their majority in the second chamber, the Bundesrat, many major bills need the approval of the SPD-Greens opposition. Both opposition parties closely coordinate the activities of their state governments in the Bundesrat. This short journey through time indicates that the opposition is a quite heterogeneous phenomenon in Germany both in terms of ideological composition and power resources. For a proper understanding of voting patterns in the Bundestag, two imperatives follow: First, a meaningful theoretical account of voting behaviour cannot be built around the opposition alone. As in most multi-party democracies the German opposition is comprised of different actors with different set of goals that will most likely facilitate different behaviour (cf. Blondel 1997, p. 470). As any theory of action builds on actors, that is entities with a sufficiently coherent set of goals the term the opposition might be analytically too big to swallow. 2 Conflict and consensus in parliamentary voting result 1 Note, that there are no genuine opposition rights in Germany but only rights that can be used by a parliamentary minority (e.g 25% of all MPs to install a committee of enquiry). Accordingly, if all opposition parties fail to reach this minority requirement the German opposition effectively enjoys no rights at all. 2 Arguably, some of the alleged theory resistance of opposition research (Helms 2008, p. 7) might result from the fact that it (at least implicitly) attributes actor-qualities to the opposition
from the behaviour of the real actors of parliamentary politics: individual parties. Hence, although the paper will make use of the aggregated term opposition it takes where necessary individual (opposition) parties, their preferences and institutional opportunities as analytical starting point. Second an understanding of consensus in parliamentary voting requires a broader focus on the bargaining situation between government and opposition parties. To illustrate this point imagine two votes in each of the legislative terms described above. In the first vote, the FDP (as member of the powerless opposition) agrees with a bill of the grand coalition. In the second vote (six years later) the SPD opposition party agrees with a bill proposal of the CDU/CSU-FDP-government. How do we interpret this consensus? In the first instance the agreement of the FDP is likely to be completely voluntary maybe as a function of ideological proximity to the government s proposal. In the second instance voluntary agreement is also possible but another explanation seems more plausible: As the CDU/CSU-FDP-government anticipated the veto power of the opposition via the second chamber it sought policy compromise before the vote (Manow and Burkhart 2007). Hence, the SPD agrees with the government proposal because the proposal was significantly tailored to its policy preferences. Although the observable behaviour of the two parties in the two votes was identical, the reasons behind it differed considerably. In the first case agreement came about voluntarily whereas in the second it was facilitated by concessions of the government in light of an oppositional veto. For an interest in behavioural patterns and institutional effects these differences are of high importance. Is consensus only a function of veto power, may majority rule itself facilitate consensual behaviour (McGann 2006) or is there some sort of cultural drift towards consensus? Accordingly, the paper tries to acknowledge these differences in order to capture the substantive meaning of consensus. per se. This is safe ground in systems with two-party competition as here opposition and individual party coincide (Kaiser 2008, p. 21). Under multi-party opposition, however, making sense of the behaviour of the opposition becomes as demanding as making sense of the behaviour of a schizophren. In fact, most studies that speak of a general explanation of the opposition s behaviour in multi-party democracies implicitly focus their analysis on the behaviour of individual parties such as the largest opposition party (Mújica and Sánchez-Cuenca 2006, p. 93).
The rest of the paper is structured as follows. It will first lay out a very sketchy and stylized framework for understanding the voting behaviour of individual opposition parties in Germany. Thereby I discuss several hypotheses for the occurrence of consensus between opposition and government parties. The then following part describes the dataset and the measurement used for a preliminary test of the hypotheses. How do German opposition parties vote and why? In general, parliamentary voting is about policy. Votes force parties to take a position on a particular issue such as the tax system, labour market regulations or civil rights. If parties were motivated by policy alone their voting decision would follow a simple spatial logic. Any proposal that makes the party better off than the existing status quo should win its agreement. Accordingly, for consensus between government and opposition parties to occur there need to be a proposal on the table that makes both government and opposition parties better off. In a world with pure policy-seeking parties we should therefore perfectly understand the varying levels of consensus by knowing the position of the proposal, the individual actors and the position of the status quo (Tsebelis 2002). Yet, parties also pursue other goals than policy (Müller and Strøm 1999), most importantly gaining votes at elections. Party Competition for votes might result in a voting behaviour that deviates from a pure policy-seeking world. Let me elaborate on these deviations in considering the two prominent dimensions of party competition: positional competition and issue competition. The first is the positional dimension (Ezrow 2010, Grofman 2004) that revolves around the idea, that the political positions of parties relative to each other and to the electorate influence their vote share. At this dimension any agreement between two parties might incur electoral costs as the party labels get blurred and the two parties become indistinguishable in the eyes of the voters. This problem is well-known to government coalitions that are often forced to agree on an identical policy proposal (Boston and Bullock 2010). It might be a particular hindrance for government-opposition cooperation as becoming indistinguishable from the government seems to be the worst opposition strategy possible. Opposition party leaders may therefore be well-advised to reject government proposals even if it promises marginal policy gains. In doing so the opposition party consequently presents itself as a clear alternative to the government and might profit electorally from this
behaviour. 3 This strategy seems to have nothing but advantages in particular if oppositional votes are not needed for the passage of a bill. The second dimension, famously coined issue competition (e.g. Green-Pedersen 2007, p. 609), is less about positions than it is about the relative importance that political parties attach to different political issues. In relation to public attention parties try to put selective emphasis (Budge and Farlie 1983) on topics that are important to them. In relation to parliamentary voting it could be argued that saliency serves as a silencer or amplifier of political distances between parties. The silencer effect is most obvious in the case of parties that predominantly pursue a regional agenda and have no stakes in national policies. They may support governmental ideal points on national policies in return for concessions on their regional agenda. In Spain part of governmentopposition-consensus seems to be driven by this logic (Mújica and Sánchez-Cuenca 2006). On the other hand, parties will be less inclined to compromise on topics that are salient to them. In fierce battles over tax reform in Germany little differences often sufficed to block agreement between government and opposition parties (Zohlnhöfer 1999). Accordingly, when modelling inter-party conflict in empirical research saliency is often combined with positional differences in a multiplicative interaction term (Martin and Vanberg 2004, p. 20). In terms of spatial models the different logics of the policyseeking and vote-seeking world read as follows (Ganghof and Bräuninger 2006): In the vote-seeking world opposition parties do not consider policy gains over the status quo as relevant but may see any deviation from their own ideal point as sufficient for disagreement. Consensus in voting between government and opposition may (partly) be understood as a function of parties policy- and vote-seeking incentives. Of course these incentives do not only come into play at the stadium of the vote on a proposal but before the actual submission of the proposal. Political leaders will make different strategic considerations 3 Dewan and Spirling (Dewan and Spirling 2011, p. 338) recently suggested that there are strong incentives for the opposition to reject any government proposal in parliament even if policy compromise was possible. By committing to oppose the government, the opposition forces it into proposals that appeal to a majority formed by its own (i.e., government) backbenchers. If there is some degree of preference overlap between government and opposition MPs this commitment to disagree might lead to considerable policy gains of the opposition.
before submitting a proposal. They will try to anticipate the voting behaviour of other parties and within their own ranks (which is in sum the probability that their prosposal will be successful). They will also evaluate the expected reaction of the electorate. There will be many proposals particularly from opposition parties that are submitted without any chance of ever receiving a majority. As such they are not intended to change the status quo but to signal the position of a party (Stecker 2010). Government parties, however, are expected to get things done, that is to get their motions passed. If they fail to do so voters may feel that the government is incapable to act and withdraw their electoral support. Rational governments will therefore formulate proposals that are likely to pass through parliament. Bills that are contentious within the government will be kept off the agenda or changed into bills that are acceptable to the government ranks. This reasoning will also be partly at work when the consent of the opposition is required, that is when the opposition enjoys an effective veto opportunity, e. g. via a second chamber. 4 Under these circumstances it is unlikely that a rational government runs into the open knife of the opposition and submits its ideal point only to let it be rejected by parliament. It is rather to be expected that the government will adapt its proposals to this situation. The government may seek consensus with the opposition if it sees sufficient policy- and electoral benefits in this strategy. If not, the government may refrain from introducing any bill that will most certainly be blocked by the opposition. In the end oppositional consensus may be the result of legislative self-restraint (Manow and Burkhart 2007, Vanberg 2005) of the government in light of an oppositional veto. Hence, it is important to understand the institutional environment in which the voting game takes place if one wants to grasp the substantive meaning of consensus. In sum, consensus in parliamentary voting maybe understood as a function of the policyseeking and vote-seeking incentives of parties that submit and vote on proposals under a certain institutional environment. It could be argued that many hypotheses about the occurrence of consensus between government and opposition are mere variations of this basic logic in that they are proxies for policy preferences or institutional situations. I will now discuss several hypotheses for Germany and relate them to this basic logic. 4 Note, that analytically this situation is no different from minority government (Juul Christiansen and Damgaard 2008, p. 46)
It has been argued that the mover of the motion to be voted on is an important predictor for consensus. Government initiatives are expected to facilitate below-average consensus (Mújica and Sánchez-Cuenca 2006, p. 97). This expectation should be particularly true for Germany. In Germany, the status quo to be voted on will very often be close to the ideal point of the opposition as it was this opposition that established this status quo when being in government. With the exception of two episodes of grand coalitions, Germany experienced clear patterns of alternation between minimalwinning-connected coalitions of competing party blocs under the lead of one of the two large parties SPD or CDU. Hence, the structural divide between government and opposition coincides with a considerable divide in terms of policy preferences and electoral ambition. Higher rates of consensus should be seen on motions that were introduced by nongovernmental actors. One such set of actors are state governments that may introduce legislative drafts via the Bundesrat. These initiatives should (at least in constitutional theory) touch on territorial issues that stand outside the party-political conflict on the national level (Harle and Stecker 2011). In consequence, agreement should be easier to reach than in cases of government-sponsored legislation. Higher consensus is also to be expected from votes on proposals that originate from parties cutting across the government opposition-divide in parliament. Hypothesis: Government-sponsored bills see more conflictual votes than initiatives of state governments and mixed initiatives. Admittedly, this hypothesis is circular to some extent. In many instances the mover of the motion is simply a proxy for the policy position of the motion to be voted upon. This is most obvious in the case of mixed initiatives that are submitted both by government and opposition parties. The consensual vote on such motions simply restates the consensus that already existed in introducing the motion. The expectation that state government-initiatives induce higher consensus already embodies the idea that different policy areas territorial issues in this case exhibit different potentials for conflict and consensus. Issues that lie at the heart of party competition will see more conflictual votes than issues that do not structure party competition. It is arguable whether policy areas possess any intrinsic conflict potential or whether this is entirely dependent on the dominant lines of conflict in a particular
country (Budge, et al. 2006, Laver and Benoit 2006). Certainly all topics that are embedded in the classical left-right axis will facilitate fewer consensual votes. With regard to Germany, I will therefore hypothesize Hypothesis: Bills relating to financial, social, labour and economic policies will induce higher conflict in voting. As discussed above, whether a vote will be consensual or conflictual between government and opposition also depends on the institutional resources that they control and the anticipation of these resources by the government. Under majority coalitions and majority rule in parliament consensus should be rare as there is no need for the government to make concessions to the opposition. If the opposition commands effective veto power, this situation changes dramatically. The government now needs to compromise with the opposition if it wants its motions to pass. Oppositional vetoes come in different forms including minority vetoes within the first chamber such as in Finland before 1993 or via strong second chambers that are controlled by majorities that are in the minority in the first chamber. In Germany an effective veto opportunity of the opposition depends on two things. First, it depends on whether the opposition controls the median voter in the second chamber, the Bundesrat (Bräuninger, et al. 2010). The Bunderat is composed of the members of the governments of the states which are in most parts coalition governments. Although party discipline is far from perfect, any federal government may count on the votes of state governments that are identical in composition. Yet, for the longest time, the Bunderat was controlled by state governments that are not identical with the federal government or that were even identical with the government-alternative. In the last years of their government, for example, SPD and Greens faced a hostile majority of states that were governed by coalitions of CDU/CSU and FDP considered the natural government alternative of SPD and Greens. This constellations made compromise on major pieces of legislation indispensable (Saalfeld 2006) and contributed to the notion of Germany as being a permanent grand coalition state (Schmidt 2007). Second, the veto power of the opposition depends on the type of the bill. The German constitution discerns into so called consent (Zustimmungsgesetze) and objection (Einspruchsgesetze) bills. For the latter, the Bundesrat can only voice its objection which can then be overturned by a governmental majority (Kanzlermehrheit) in the Bundestag.
In the case of consent laws, however, the veto is not only dilatory but absolute. More precisely, in order for a consent bill to pass it needs to be approved by an absolute majority of state votes. If it fails this majority the bill is rejected. This means in effect, that any government without an own majority in the Bundesrat may be forced to compromise with the opposition. Hence the majority constellations in combination with the type of bill will be an important predictor for the degree of consensus. Hypothesis: Consensus will be higher if the opposition commands an effective veto via the Bundesrat (consent bills). Indeed, the work of Burkhart and Manow (Burkhart and Manow 2006) shows that the level of conflict in Bundestag s votes is significantly related to majority constellations in the Bundesrat. Miller and Stecker (Miller and Stecker 2008) prove the same logic for the work of the Bundestag s committees. Data and operationalisation To test these expectations I use a dataset that consists of information gathered by Burhkart (Manow and Burkhart 2007) and Miller/Stecker (Miller and Stecker 2008). It comprises information on the progress of 6500 individual bills in the Bundestag between 1974 and 2005. Among others these information include the initiator, the type of bill, the respective policy field and the bill s status (approved or rejected). At this stage it also (and most importantly) includes the results of the final vote after the third reading in the Bundestag for all 3640 bills that were approved by the Bundestag. Unfortunately (for empirical purposes) most final votes on bills are semi-open that is via raising hands or rising in seats (Saalfeld 1995). The results are recorded with varying levels of detail. In the worst case the chairing president of the session only notes that a bill is adopted or rejected. In most cases he or she gives a few more details such as that the bill was adopted with a broad majority or with the majority of the coalition. Inferring the voting behaviour of individual parties is therefore very time-demanding if not impossible. Therefore, the only reliable measurement of the dependent variable at this point is a dichotomy of broad consensus (including at least the largest opposition party) or conflictual (at least without the agreement of the largest opposition party). This measure is sufficient to test for the effects of the Bundesrat s veto in the case of consent bills as it may never be invoked without the largest opposition party. It is however
insufficient to trace the behaviour of smaller opposition parties such as the Greens, the FDP and the Left Party. This will be done with more elaborate data in the future. The independent variables were operationalized as follows. The identity of the mover of a motion is coded as a dummy variable for state government and mixed initiatives where government initiatives act as baseline. The policy field is coded as the policy field given by the official database on legislation in Germany. Admittedly, policy fields might only be a weak proxy for the conflict potential of a motion. Votes on justice might include highly contentious votes on abortion but also low-profile votes on the organization of prisons. Preliminary Analysis A first test of the hypotheses will be presented using descriptive statistics. Figure 1 plots the different share of consensual votes for different categories of votes. The first column overall consensus represents the global average. In 69% of all votes the government and the largest opposition party voted in unison. Figure 1: Consensus by different categories of bills (1974-2005) Remarks: N = 2251 bills.
This rate seems surprisingly high given that the largest opposition party always represents the main competitor of the largest government party. The averages of the other categories mainly follow the hypotheses. The effect of the identity of the mover of a motion is very pronounced with non-governmental initiatives seeing consensus in 85% of the votes. But as said before, this finding is rather tautological than surprising as the submission by both government and opposition parties (in the case of mixed initiatives) already forecloses consensus in voting. The last two columns stand for the difference in consensus between situations where the opposition enjoys an effective veto opportunity via its control of the Bundesrat and where not. There is a difference of 5 per cent but it is less clear-cut than expected. It might be inferred that an institutional veto of the opposition is not the main driving force behind government-opposition consensus. Figure 2: Consensus by policy areas (1974-2005) Figure 2 plots the degree of consensus in relation to the different policy areas of a motion. As expected, votes on social policy and finance induce below-average consensus. Yet, there are also a few surprises. Votes on economic policy are highly consensual. Exceptional conflict is exhibited by votes on environmental policy. With the advent of
the Greens in the early 1980s this area has gained in salience and seems also to be a main battleground for the two large parties. Conclusion The paper offered a first glance at the patterns of conflict and consensus between government and opposition in Bundestag votes. In 69 per cent of all votes on bills the largest opposition party group (either the SPD or the CDU/CSU) compromises with the government. Variance in consensus is driven by the identity of the mover of a motion, effective veto opportunities of the opposition and the policy area under consideration.
Bibliography Blondel, J., (1997) Political Opposition in the Contemporary World. Government and Opposition, 32 (4), pp. 462-486. Boston, J. and Bullock, D., (2010) Multi-party governance: Managing the unity-distinctiveness dilemma in executive coalitions. Party Politics, pp. Bräuninger, T., Gschwend, T. and Shikano, S., (2010) Sachpolitik oder Parteipolitik? Eine Bestimmung des Parteidrucks im Bundesrat mittels bayesianischer Methoden. Politische Vierteljahresschrift, 51, pp. 223-249. Budge, I. and Farlie, D. (1983) Party Competition Selective Emphasis or Direct Confrontation? An Alternative View with Data. in: H. Daalder and P. Mair (ed.) West European Party Systems: Continuity & Change, pp. 267 305 (London: Sage). Budge, I., Klingemann, H.-D., Volkens, A., Bara, J. and Tannenbaum, E. (2006) Mapping Policy Preferences II: Estimates for Parties, Electors, and Governments in Eastern Europe, European Union and OECD, 1990-2003 (Oxford: Oxford University Press). Burkhart, S. and Manow, P., (2006) Kompromiss und Konflikt im parteipolitisierten Föderalismus der Bundesrepublik Deutschland. Zeitschrift für Politik, 16 (3), pp. 807 824. Dewan, T. and Spirling, A., (2011) Strategic Opposition and Government Cohesion in Westminster Democracies. American Political Science Review, 105 (2), pp. 337-358. Ezrow, L. (2010) Linking Citizens and Parties. How electoral systems matter for political representation (New York, NY: Oxford University Press). Ganghof, S. and Bräuninger, T., (2006) Government Status and Legislative Behaviour. Partisan Veto Players in Australia, Denmark, Finland and Germany. Party Politics, 12 (4), pp. 521-539. Green-Pedersen, C., (2007) The Growing Importance of Issue Competition: The Changing Nature of Party Competition in Western Europe. Political Studies, 55, pp. 607-628. Grofman, B., (2004) Downs and Two Party Convergence. Annual Review of Political Science, 7 (1), pp. 25-46. Harle, I. and Stecker, C., (2011) Die Initativtätigkeit des Bundesrates im Lichte der Parteipolitisierungshypothese. Zeitschrift für Parlamentsfragen (zur Veröffentlichung angenommen), 42 (2), pp. 325-334. Helms, L., (2008) Studying Parliamentary Opposition in Old and New Democracies: Issues and Perspectives. The Journal of Legislative Studies, 14 (1-2), pp. 6-19. Juul Christiansen, F. and Damgaard, E., (2008) Parliamentary Opposition under Minority Parliamentarism: Scandinavia. The Journal of Legislative Studies, 14 (1-2), pp. 46-76. Kaiser, A., (2008) Parliamentary Opposition in Westminster Democracies: Britain, Canada, Australia and New Zealand. The Journal of Legislative Studies, 14 (1-2), pp. 20-45. Laver, M. and Benoit, K. (2006) Party Policy in Modern Democracies (London: Routledge). Manow, P. and Burkhart, S., (2007) Legislative Self-Restraint under Divided Government in Germany, 1976-2002. Legislative Studies Quarterly, XXXII (2), pp. 167-191. Martin, L.W. and Vanberg, G., (2004) Policing the Bargain: Coalition Government and Parliamentary Scrutiny. American Journal of Political Science, 48 (1), pp. 13-27. McGann, A.J., (2006) Social Choice and Comparing Legislatures: Constitutional versus Institutional Constraints. Journal of Legislative Studies, 12 (3-4), pp. 443-431. Miller, B. and Stecker, C., (2008) Consensus by Default? Interaction of Government and Opposition Parties in the Committees of the German Bundestag. German Politics, 17 (3), pp. 307-324. Mújica, A. and Sánchez-Cuenca, I., (2006) Consensus and Parliamentary Opposition: The Case of Spain. Government and Opposition, 41 (1), pp. 86-108.
Müller, W.C. and Strøm, K. (1999) Policy, office, or votes? How political parties in Western Europe make hard decisions (Cambridge: Cambridge University Press). Saalfeld, T. (1995) Parteisoldaten und Rebellen. Eine Untersuchung zur Geschlossenheit der Fraktionen im Deutschen Bundestag (1949-1990) (Opladen: Leske + Budrich). Saalfeld, T., (2006) Conflict and Consensus in Germany's Bi-cameral System: A Case Study of the Passage of the Agenda 2010. Debatte, 14 (3), pp. 247-269. Schmidt, M.G. (2007) Das politische System Deutschlands (München: C.H. Beck). Stecker, C., (2010) Causes of Roll Call Vote supply. Journal of Legislative Studies, 16 (4), pp. 438-459. Tsebelis, G. (2002) Veto players. How political institutions work (Princeton, N.J: Princeton University Press). Vanberg, G. (2005) Verfassungsgerichtsbarkeit und Gesetzgebung: Zum politischen Spielraum des Bundesverfassungsgerichts. in: P. Manow and S. Ganghof (ed.) Mechanismen der Politik. Strategische Interaktion im deutschen Regierungssystem, pp. 183-213 (Frankfurt: Campus). Zohlnhöfer, R., (1999) Die große Steuerreform 1998/99: Ein Lehrstück für Politikentwicklung bei Parteienwettbewerb im Bundesstaat. Zeitschrift für Parlamentsfragen, 30 (2), pp. 326-345.