Heterotemporal Parliamentarism: Does Staggered Membership Renewal Matter?

Klaus H. Goetz1, Christian Stecker2, and David Willumsen3

Department of Political Science, University of Munich*

A rich literature on political business cycle has produced ample evidence for how the electoral calendar influences legislative behaviour and the functioning of parliaments. Against this background surprisingly little is known about the effects of staggered membership renewal (SMR). Staggered mandates desynchronize the electoral calendars of legislators within a single chamber. This creates different “classes” of legislators that operate with different time horizons and thereby face different behavioural incentives at any given point in time. Extending the literature on political business cycles to SMR-legislatures our paper asks: In what ways do staggered elections influence legislative behaviour and – in turn – the working of the entire parliamentary chamber? At least two basic effects of staggering are plausible. It can be expected to have stabilising effect on parliamentary activity, by reducing the chances of wholesale change in the legislature’s composition. Second, it can lead to more cycles in parliamentary activity, by putting the legislature in permanent election mode as the next election is always close(r) for at least one group of legislators than the other(s). Much of the exact effect of staggering on parliamentary activity will depend on inter-class interaction, i.e. whether legislators with different time horizons will engage in cooperation or mutual restraint. Intertemporal logrolls between legislators, for example (Shepsle et al., 2009), would allow each legislator to optimally pursue his re-election thereby making overall legislative activity more short-termist. This paper constitutes a first attempt towards addressing these two possibilities. With a focus on the procedural aspects of parliamentary staggering, the paper analyses the case of the German Bundesrat, which is composed of representatives of governments of the federal states and wields significant legislative power. As the Bundesrat’s composition is determined by staggered Land elections, the time budgets and likely time horizons of Bundesrat members vary greatly. Analysing different dimensions of parliamentary activity, including resolutions and bill initiatives of Bundesrat members, we show that staggering indeed has a significant impact on the parliamentary behaviour of (groups of) members and the Bundesrat overall.

Paper prepared for the ECPR Joint Sessions, Salamanca, 2014. First draft – Comments are very welcome.

1 [email protected] 2 [email protected] 3 [email protected]; corresponding author * This study is part of the ongoing comparative research project “Staggered Membership ”Effects and Patterns Comparing ׃Chambers Second in Horizons Time Differential and Renewal funded by the DFG; see http://www.politicaltime.eu/ for more details. We thank Sandra Gadinger for helpful research assistance. 1

1. Introduction

More than 20 democracies have a bicameral legislature where the members of the upper house are elected based on a staggered formula4, yet the phenomenon is almost entirely unstudied (see Shepsle et al., 2009 for an exception). The, usually implicit, justification for employing staggered terms in legislatures is that this will lead to more stable policy outcomes and to less short-termism in decision-making, as the effect of temporary swings in the public mood is reduced. However, whether this actually occurs is almost entirely unexplored. Most importantly, as will be discussed below, the effect of staggered terms of office is unexplored in the literature when it comes to the effects of individuals being able to influence their chances of obtaining additional terms of office. This paper sets out to explore the effects of desynchronised political time in second chambers. It asks the question: what happens when there are different time horizons within a single institution? More specifically, what are the effects of staggering on legislative behaviour and institutional outcomes?

The rest of this paper consists of five sections. First, the literature on the effects of staggered terms in various institutional settings will be briefly reviewed, along with a discussion of the various justifications advanced for using staggered terms in upper chambers. This will then be combined with the literature on political business cycles. The argument will be made that staggered terms in upper chambers can be expected to lead to increases in short-termist behaviour by legislators. Second, the institutional structure and legislative process of the German Bundesrat will be briefly summarised, and we will discuss how the Bundesrat is a most-likely case for observing an increase in the number of political business cycles due to staggered terms. Third, the data analysed and the modelling approach will be introduced. Fourth, the models will be presented and the results of these discussed. A final section concludes.

4 Calculated by the authors based on the IPU’s PARLINE database: http://www.ipu.org/parline/

2. The Effects of Staggered Terms

As noted above, there is little extant literature on the phenomenon of staggering in legislatures. This is also the case when it comes to other institutions; while the phenomenon of staggering has been studied, the literature has focussed on it as a binary concept in terms of institutional design – does a legislature have staggered membership or not? However, as will be shown below, this has ignored the ability of individual members of (staggered) institutions to influence their chances of obtaining additional terms of office. This paper will next briefly review the literature on staggered terms, as it applies to central banks, regulatory and other public agencies, and company boards of directors. Further, it will outline the justifications given in the literature for employing staggered terms in legislatures.

2.1. Staggering in Central Banks

The main focus with regards to staggered terms within the literature on central bank governing boards is the extent to which they limit the ability of an appointing actor to place a majority of appointees onto the board; that is, whether staggered terms help ensure a politically independent central bank (see e.g. Lindner, 2000). Generally assumed, however, is that appointees are faithful agents of their principals. Thus Waller (1989, 1992) explicitly models overlapping central bank board member terms, but the board members cannot vary their behaviour in his models, and De Paso (2000, p. 474), explicitly states that “governors appointed by each [political] party are assumed to adhere to the monetary policy desired by the party who appointed them”. More recent work on the individual behaviour in central banks also leaves out the potential effects of the desire for re-appointment to another term on central bankers’ voting behaviour (Hix et al., 2010).

2.2. Regulatory and other public agencies

Using the length of terms of office and/or whether these terms are renewable is a standard approach when seeking to measure the independence of regulatory agencies (see e.g. Gilardi and Maggetti, 2010; Hanretty and Koop, 2012), having been employed, amongst others, for studying the independence of EU agencies (Wonka and Rittberger,

2010), public broadcasters (Hanretty, 2010), administrative authorities in France (Elgie and McMenamin, 2005), and competition authorities (Guidi, 2013). The relative timing of agency members’ appointments is not explored, however. Indeed, some work simply treats re-appointments as “essentially the same as new appointments” (Snyder and Weingast, 2000, p. 287 footnote 21), indicating the limited interest in the potential influence of appointees’ desire for additional terms of office.

2.3. Company Boards of Directors

A substantial literature on the effect of staggered boards5 on firm and shareholder value exists in the research on corporate governance (see e.g. Bebchuk and Cohen, 2005; Jiraporn et al., 2012; Duru et al., 2013). However, as with central banks, those serving on boards are generally assumed to be perfect agents of those that appoint them. The primary use of staggered boards as an independent variable is as part of an index of shareholder power (with staggered boards seen as giving less power to shareholders) or an index of protection from takeover (Gompers et al., 2003; Cremers and Nair, 2005; Jiraporn and Gleason, 2007). The effects of a company being more protected from takeovers range from different levels of leverage (Jiraporn and Gleason, 2007; Jiraporn and Liu, 2008), more transparency (Jiraporn et al., 2012), higher dividend payments to shareholders (Jiraporn and Chintrakarn, 2009), and lower levels of spending on research (Faleye, 2009). Staggered terms are also argued to result in management being protected from the discipline of the market (Daines and Klausner, 2001), making them put in less effort for the shareholders, leading to, amongst other things, reduced shareholder value (Bebchuk et al., 2002a, 2002b).

2.4. Why staggered Terms in Legislatures?

As far back as the Federalist Papers, it was argued that no government can “be truly respectable, without possessing a certain portion of order and stability” (Hamilton et al., 2001, p. 324), and staggered legislative terms are then argued to bring this about. This in turn is seen as advancing the quality of policy outcomes, by limiting the uncertainty and

5 This literature also uses the term “classified board” for this set-up, since board members are organised into different classes (much like the 3 classes of US Senator).

inefficiency associated with rapid policy reversal (see also Majone, 1996). Staggered terms in legislatures are empirically associated with bicameralism, which is itself often justified as being a stability-inducing institution. However, the literature on bicameralism places very little emphasis on the effect of staggered terms. Rather, this literature is dominated by its effects on reducing policy cycling (Riker, 1992a, 1992b; Patterson and Mughan, 1999), representation (Campion, 1953; Money and Tsebelis, 1992; Patterson and Mughan, 1999), parliamentary-executive relations (Russell, 2001) and the quality of legislation (Weare, 1965; Tsebelis and Rasch, 1995; Rogers, 2001; Heller, 2007). Even the classic book on bicameralism, studying a country which has a senate with staggered terms (France) does not mention either staggered terms or overlapping terms (Tsebelis and Money, 1997). In a notable exception to the lack of research on the effects of staggered terms on legislative behaviour, Shepsle et al. (2009) find evidence that the theorised stabilising effects of staggered terms may not occur, by showing that the way the terms of US Senators overlap lead to more, rather than less, election-oriented spending.

The underpinning of the argument that staggered legislative terms leads to policy stability is two-fold. Firstly, it is argued that by restricting the share of the membership of an institution that has to be renewed at any given time, the effects of temporary swings in the public mood on public policy are reduced, because only part of the institutional make-up is affected. Second, it is argued that the attempts by the share of the institution next up for re-election (by definition, it should be noted, less than a majority) to behave in a short-termist manner will be limited, as this will be opposed by the members with longer time remaining of their term. In the case of legislators, this is a particularly pressing concern, due to their direct influence on public policy. By staggering terms, the share of legislators with an incentive to provide particularistic goods to their constituents due to an upcoming re-election campaign is minimised, and their legislative colleagues, not up for re-election at the same point in time, are then expected to deny them these re-election-oriented goods. When these ‘deniers’ then subsequently are up for re-election, they are then expected to be denied particularistic goods by those they themselves denied these goods to. However, whether this will occur is an unexplored matter, and as will be shown below, there are strong reasons to expect that legislators will do their utmost to obtain additional terms of office through their 5

control of the public purse. Further, it should also be noted that while passing legislation is a collective endeavour, a number of other crucial functions of legislatures, in particular their role in oversight, can be conducted by individual parliamentarians, and aimed towards re-election rather than good public policy.

2.5. Political Business Cycles

A substantial literature exists on the potential use by politicians of the resources they control through the state to improve their chances of re-election. Starting with the work of Nordhaus (1975), whether politicians are able to manipulate the economy to perform well immediately prior to elections, and in this process create a political business cycle, has been subject to substantial debate. Since such manipulations are predictable, it is argued that such political business cycles lead to companies and unions anticipating the actions of politicians in their contract negotiations, which in turn simply leads to increased inflation, rather than any increased economic output (Rogoff and Sibert, 1988; Alesina and Roubini, 1992).

A substantial empirical literature suggests that politicians actively attempt to manipulate the economy prior to elections. Thus, Blais and Nadeau (1992), studying Canadian states, found that there is a 1% increase in government spending in election years, and Yoo (1998) finds that, in Japan, there exists a tax cycle, where taxes are lowered before elections to appeal to voters. Galli and Rossi (2002) find evidence of pre- election budget increases in German Länder. Similarly, Foucault, Madies and Paty (2008) find that French municipalities increase their spending in the run-up to local elections.

In addition to general increases in spending, a strategy exists that targets favoured or crucial electoral groups prior to elections, at the cost of other groups, and so avoiding general budget increases (which may be a liability in elections). Looking at Colombian municipal elections, Drazen and Eslava (2010) find a pre-election increase in targeted spending to specific groups (and a decrease in spending on other items), as well as evidence that voters respond favourably to this; Schneider (2010) finds that incumbent governments in the German Länder try to increase their vote share by moving money between different parts of the budget to appeal to specific groups, but avoiding general budget increases or deficits. Further work suggests that manipulation

of the economy prior to elections is conditional on the risk of losing office; those governments that are confident of re-election have little reason to manipulate the economy in the run-up to an election, in particular if the spending cannot be clearly linked to them (see e.g. Schultz, 2009; Aidt et al., 2011; Potrafke, 2012)

Generally speaking, the use of public spending to influence the economy immediately prior to elections, and so improve the chances of a specific party (or set of parties) has been identified across the democratic world. Thus, public money has been found to be used to influence elections in Brazil (Sakurai and Menezes-Filho, 2008, 2011; Brollo and Nannicini, 2012; Monteiro and Ferraz, 2012; Caselli and Michaels, 2013), Portugal (Veiga and Pinho, 2007; Veiga and Veiga, 2007), Spain (Solé-Ollé and Sorribas-Navarro, 2008), Australia (Worthington and Dollery, 1998), India (Arulampalam et al., 2009), Canada (Lecours and Beland, 2009), Norway (Sørensen, 2003), Sweden (Dahlberg and Johansson, 2002; Johansson, 2003), as well as the United States (Grossman, 1994; Levitt and Snyder, 1995; Alvarez and Saving, 1997; Levitt and Snyder, 1997; Ansolabehere and Snyder, 2006), indicating that this is a very widespread phenomenon.

In addition to these findings at the level of countries, regions and cities, substantial work also indicates that such re-election-oriented use of public money also exists at the level of individual legislators. In particular, such an effect has been identified in the case of the US Congress (a non-exhaustive list: Mayhew, 1974; Shepsle and Weingast, 1981; Weingast et al., 1981; Rich, 1989; Adler and Lapinski, 1997; Knight, 2005), but also in the parliaments of Germany (Stratmann and Baur, 2002) and Italy

(Golden and Picci, 2008).

In addition to the use of public spending to appeal to voters, politicians can also attempt to improve their chances of electoral victory through non-financial legislation, signalling voters that their interests beyond economic matters are being looked after, giving politicians incentives to time popular legislation so that it is passed in the run-up to elections. Thus, as we observe a political business cycle, we may also observe a legislative business cycle (Lagona and Padovano, 2008). While such legislative business cycles have not been studied extensively, there is some evidence for their existence. Brechler and Gersl (2011), studying the Czech parliament, find two peaks in legislative

productivity, one immediately after elections (fulfilling promises) and one as elections grow nearer (appealing to voters for another term). Lagona, Maruotti and Padovano (2012) find evidence of both a decree cycle and a legislative cycle in Italy, with the issuing of decrees peaking early in a legislative term, while law production peaks towards the end of the term. Some evidence of a legislation cycle has also been found in the European Parliament, driven not by electoral concerns but by the regular reallocation of agenda powers in the EP (Kovats, 2009). However, it should also be noted that when studying coalition governments, Martin (2004) finds that the issues that are important to governing parties are legislated on early, while more divisive issues are introduced later (if at all). Thus, a peak in legislation towards the end of a term may simply be the result of a combination of procrastination by legislators and the fact that legislation takes time to prepare, leading to a natural skew in the distribution of legislation over the course of a term.

2.6. Political Cycles in the German Bundesrat

In this section, based on a general theoretical discussion we formulate specific hypotheses for the German case. When combining the insights of the literature on political and legislative business cycles discussed above with the previously discussed literature on staggered terms, it becomes clear that the core assumption regarding the effect of staggered terms, namely that they lead to stability, may be untenable due to the desires of legislators (but also any other individuals working within any institution employing staggered terms) to obtain additional terms of office, and their ability to influence this through the powers of their office. Such instability may manifest itself in a number of areas, such as parliamentary questions, the introduction of amendments, public spending, and cultivating personal votes through floor voting.

The key aspect here is the extent to which voters have long-term memories of legislators’ actions. If voters have perfect memories, then the timing of legislative behaviour relative to elections is irrelevant; much more realistically, if voters have short-term memories6, then the timing of behaviour relative to elections matter. That is,

6 The “What Have You Done For Me Lately” (WHYDFML) principle (Shepsle et al., 2009).

voters’ imperfect memories may lead to political business cycles. Recall that staggering has been argued to lead to policy stability, longer time horizons and larger time budgets, due to the limited share of members of an institution facing re-election/re-appointment at any given time. However, this logic can be reversed, suggesting that staggering leads to less stability, not more: Those legislators nearest re-election campaigns may be allowed to obtain particularistic/campaign goods by their colleagues with the understanding that, when these same colleagues run for re-election themselves, they will be extended the same courtesy. Alternatively, the legislators not up for re-election may simply not be able to prevent the re-election-oriented legislative behaviour of their colleagues. In both cases, we would expect that, the nearer a legislator is to her re- election date, the more she will engage in behaviour to improve her chances of winning re-election.

Cyclical legislative behaviour may come about in three ways. First of all, each legislator may engage in certain legislative behaviour based on the extent to which she faces an upcoming election, thus leading to the classic case of a political business cycle, where electorally popular policies and public spending are focussed towards the end of the legislative term.

With regard to the German Bundesrat this can be translated into more specific expectations. Cyclical behaviour may be first expected within state parliaments. In line with theory, state governments may be more inclined to pass popular legislation near the end of term. Yet, as has been illustrated at length (see e.g. Jeffery and Hough, 2001; Burkhart, 2005), state elections are heavily influenced by issues that are to be decided at the federal level. State governments can only indirectly influence these issues via bill initiatives in the Bundesrat. Hence, for the German states political cycles may partly structure the timing of bill initiatives and resolutions within the Bundesrat7. In the case of the German Bundesrat, we would thus have (up to) 16 different political cycles taking place, one for each Land. This gives us

7 The null hypothesis regarding the effect of an approaching election is of course that legislative behaviour is independent of the amount of time left in a legislator’s term of office.

Hypothesis 1: The closer the election date of a Land government is, the more resolutions and legislative initiatives it will propose in the Bundesrat.

Secondly, the legislative behaviour in the Bundesrat may be influenced by the electoral cycle at the national level (i.e. of the Bundestag). State parties forming regional governments are also part of their national party organization. Although there is some programmatic heterogeneity (Müller, 2013) the different organizations at different government levels are vertically well-integrated (Renzsch, 1999; Detterbeck and Jeffery, 2009). Various coordination mechanisms ensure that the collective interest of the entire party is pursued. Submitting bill initiatives to the Bundesrat may therefore be seen as a collective strategy by which state parties try to increase the vote share of their party during national elections. This gives us

Hypothesis 2: The nearer national elections are, the more resolutions and legislative initiatives members of the Bundesrat will propose.

Third, there is the potential effect of shared time horizons of cohorts within a legislative chamber. If voters apply the WHYDFML principle, there is an incentive for those legislators nearest election to cooperate in order to maximise their chances of re- election. Those legislators further away from re-election then defer to all those within their chamber that faces an imminent election. This would then lead to an intertemporal log-roll, based on the expectation that a similar courtesy will be extended to them when it is their time to face the voters (Shepsle et al., 2009). This gives us

Hypothesis 3a: Legislators who do not face an imminent election reduce their legislative activity whenever a fellow legislator, regardless of party affiliation, approaches an election.

Finally, as a specific case of the above, legislative behaviour in a legislative chamber may also be influenced by a combination of the electoral horizons of legislators and the partisan links that they share. If, following the WHYDFML principle, legislators heavily discount legislative activity or success early in their term and at the same time recognise that their long-term interests are best served by the chamber being controlled by members of their own party, then we would expect that party links would matter greatly for legislative behaviour. This leads us to the “Partisan Moment of Glory” hypothesis: To maximise the collective power of their parties, and since, as noted above, legislative time is limited (and time used by one actor cannot be used by another), legislators who share partisan affiliation defer to those amongst them nearest election, in order to maximise their exposure to voters applying the WHYDFML principle. This gives us

Hypothesis 3b: Legislators who do not face an imminent election reduce their legislative activity whenever a fellow party member approaching an election.

Having reviewed the literature on staggered terms, and discussed how, when viewed from the perspective of the literature on the political business cycle, we can expect staggered terms to lead to a multiplication of such cycles, with more legislative behaviour oriented towards the re-election goal, this paper now turns to the institutional structure of the case at hand: The German Bundesrat.

3. The Legislative Process in Germany

Article 50 of the constitution stipulates that the German states participate directly in the decision-making at federal level through the Bundesrat. This participation is institutionalised in various forms. Most importantly, the Bundesrat may veto so called consent bills (“Zustimmungsgesetze”). Consent bills deal principally with issues that influence the financial situation of the federal states. With regard to objection bills (“Einspruchsgesetze”) it enjoys a suspensive veto. Objection bills deal with issue that are of the competence of the federal state alone (Basic Law of Germany, 2010, Articles 72-

74). The right to initiate legislation is among the less prominent participatory powers. Every state (or group of states) may introduce legislative proposals into the Bundesrat (Article 76 Sub-section 1 Basic Law). The Bundesrat decides with the majority of its votes about the introduction of bills into the Bundestag (Patzelt, 1999; Ismayr, 2008: 393). Bills that secure this majority enter the Bundestag where the usual legislative process starts.

All legislation at the federal level in Germany falls into one of two categories, based on the federal structure of the German state, that is, two types of bills exist: “Consent bills” (Zustimmungsgesetze) and “Objection bills” (Einspruchsgesetze). Consent bills deal principally with issues that influence the financial situation of the federal states (Bundesländer); objection bills deal with issue that are of the competence of the federal state alone (Basic Law of Germany, 2010, Articles 72-74). In terms of the legislative process, the key difference between the two types of law concerns the status of the Bundesrat’s veto. For objection bills, the Bundesrat’s may be overturned the Bundestag; in other words, the Bundesrat has only a suspensory veto in this case. Consent bills, on the other hand, require that the Bundesrat actively approves a bill, giving it a full, formal veto over such legislation (Patzelt, 1999; Alter, 2002; see also articles 70-78 of the German Basic Law).

Regardless of whether a bill is a consent or objection bill, the legislative process is identical until the final (veto) stage. A bill can originate in either the Bundesrat, the Bundestag or from the federal government. In the first round of deliberations, the Bundesrat comments on a proposed piece of legislation, which is then introduced into the Bundestag. After three readings and passage in the Bundestag, the bill is then sent back to the Bundesrat for a second round of deliberations. It is at this stage that the distinction between consent and objection bills becomes crucial. In the case of objection bills, the Bundesrat may reject a bill passed by the Bundestag; however, this “objection” to the bill (which needs to be passed with an absolute majority of votes in the Bundesrat8) can be overridden by an absolute majority of the Bundestag, which is almost always commanded by the government of the day. However, if the Bundesrat

8 Currently 35 votes (69/2 + 1)

passes its objection with a 2/3rds majority, a similar majority is required to over-ride in the Bundestag. While the Bundesrat thus only has a suspensory veto in the case of the Objection bills, there is evidence that when the partisan composition of the Bundesrat differs from that of the Bundestag, objections are used as a legislative tool by the Bundesrat, to delay the passage of laws, and so exploit the impatience of the federal government to extract policy concessions (Manow and Burkhart, 2008).

For consent bills, however, the veto of the Bundesrat is absolute, and not merely suspensory as in the case of objection bills. In other words, for consent bills, if the Bundesrat does not support a bill passed by the Bundestag, the bill fails. However, in that case, a conference committee can be convened at the request of either the government, the Bundestag or the Bundesrat. The conference committee, which has 16 members each from the Bundestag and the Bundesrat (each Land appointing one member9, may recommend a compromise, which nonetheless has to be accepted by both houses (Patzelt, 1999, pp. 75–80).

Voting weights in the Bundesrat are based on the population at the Land level, with each land having between 3 and 6 votes10. Worth noting is that votes in the Bundesrat have to be cast en bloc (Basic Law of Germany, 2010, Article 51:3); thus, for example, a coalition government at the Land level cannot split its votes in the Bundesrat. Further, the Bundesrat needs an absolute majority of votes in favour in order to take decisions (Basic Law of Germany, 2010, Article 52:3), making abstaining equal to an “against” vote. As the membership of the Bundesrat is composed of members of the governments of the sixteen German Länder, it sits permanently, rather than having legislative terms. As such, changes in its composition are dictated by the results of Land

9 It is worth noting that the composition of the conference committee is thus different in terms of voting power than the Bundesrat as a whole, as each Land has the same voting weight in the conference committee, which is not the case in the Bundesrat plenum (see below). 10 Each Land has a minimum of three votes in the Bundestag; those Länder with more than two million inhabitants have four votes in the Bundesrat, Länder with more than six million inhabitants five votes, and finally the Länder with more than seven million inhabitants have six votes (Basic Law of Germany, 2010, Article 51:2). As in most federal systems, these voting weights give a disproportionate influence to smaller states. For example, Nordrhein-Westfalen, the biggest land by population, has 8.7% of the votes in the Bundesrat (6 of 69), but around 22% of the population of Germany.

elections (Basic Law of Germany, 2010, Article 51:1); when the government of a Land changes, as does the composition of the Bundesrat.

Thus, the Bundesrat is an extreme example of a legislature with staggered terms. While, for example, the US Senate has 3 sets of senators with different time left until their next election, and the French Senate has two, the members of the Bundesrat has up to 16 different expected election dates (see Table 4 in the appendix). This in turn means that, for the present purposes, these staggered terms provide substantial variation in electoral incentives with regards to the behaviour to be analysed. The next section introduces the data to be analysed, details the theoretical expectations we would have based on this data, and discusses the estimation techniques to be used.

4. Data and Methods

Two types of data will be analysed in this paper: The Resolutions (Entschließungen) of the Bundesrat, and the Initiatives of the Länder in the Bundesrat, that is, the introduction of legislation by a Land (or a coalition of Länder)11. Both sources of data were collected by the authors from the DIP12 and GESTA13 databases14, and collected on a daily basis. For Bundesrat Initiatives, the data covers the period 1972-2013, and for Resolutions, the period 1976-2005 is covered. In both cases, the data collected consists of the population of all such behaviour in the period covered.

In total German state governments submitted 1561 bills to the Bundesrat between 1972 and 2013 and 592 resolutions between 1976 and 2005. As can be seen in Figure 1 and Figure 2 activity varies considerably across states and time. Bavaria and Baden-Württemberg (alone accounting for altogether 758 initiatives and 304 resolutions) are among the top suppliers of these initiatives. It is also evident that states

11 Future work will also include data on floor speeches and the budget; here, it is worth noting that the budget is an Einspruchsgesetz (objection bill), meaning that the power of the Bundesrat is limited to a suspensory veto, as discussed earlier. 12 http://dip.bundestag.de/ 13 http://www.bundestag.de/service/glossar/G/gesta.html 14 Thereby we build on the work of Burkhart and Manow (2007) as well as Miller and Stecker (2008); see also Bräuninger and König (1999).

often engage in collective initiatives. As a general pattern it has been identified that states that are governed by parties that are in the opposition at federal level are more active in submitting initiatives.

Figure 1: Number of single and collective bill initiatives by state

Figure 3 and Figure 4 visualize as to what extent the legislative activity of the German states in the Bundesrat is driven by the electoral cycle at the federal level. The bars represent the number of initiatives that are submitted during a quarter year. The dashed vertical lines mark the date of a general election at federal level. Particularly with regard to bills it is nicely shown that the electoral cycle at national level influences the legislative activity of the members of the Bundesrat. Around all 12 federal elections in our period of investigation legislative activity is increased, resulting in particularly clear peaks of bill numbers in 2013, 2009, 2002 and 1994.

Figure 2: Number of single and collective resolutions by state

Surprisingly in light of the theory of business cycles, however, activity is not always or exclusively raised before an election takes place. In 1998, 2002 and 2009 activity rather increases shortly after the election. This post-election peak can be understood as resulting from the ‘second-orderness’ of elections in the German states. To a considerable extent state-level electoral campaigns focus on federal issues. Specifically in their electoral manifestos and coalition agreements regional party branches promise to address disadvantageous federal legislation via Bundesrat initiatives. Hence, submitting Bundesrat initiatives right after elections can be seen as delivering on promises made during electoral campaigns – a cyclical behaviour that is in line with our theory.

Figure 3: Number of bills per quarter and electoral cycle at the federal level

Figure 4: Number of resolutions per quarter and electoral cycle at the federal level

Figure 5 visualizes how the electoral cycle at state level influences the legislative activity of German state governments. Four selected states including the top suppliers Baden-Württemberg and Bavaria are presented. Again the bars represent the number of initiatives that the state submitted to the Bundesrat in a quarter year. The dashed vertical lines now represent the event of an election in that state. The pattern of an electoral cycle in legislative activity most clearly emerges in the state of Hesse and to a lesser degree also in Bavaria and Baden-Württemberg. Again we observe the surprising pattern that legislative activity may be raised before and/or after the state election. In North Rhine-Westphalia such a cycle is less obvious. Similarly in most other states an electoral cycle is not readily identifiable upon visual inspection.

Figure 5: Number of resolutions per quarter and electoral cycle at the land level

In terms of modelling the decision of a Land government to introduce a bill or a resolution in the Bundesrat, a number of issues regarding the nature of the data have to be considered. Most obviously, the data used for the dependent variables, namely the number of bills and resolutions introduced on a given day by a Land government in the Bundesrat, should be understood as being count data; on any given day, a Land government decides to introduce some number of bills or resolutions, starting at zero 18

(i.e. no introduction) and then increasing in discrete steps of 1. Due to the non-equal and non-fixed amount of time between Land elections, it is not possible to aggregate the data to provide a count of actions during a specific period (e.g. a month); rather, the unit of time has to be a (working) day15. As the standard deviation (for both dependent variables) was significantly larger than the mean, a negative binomial model was used.

Since the data consists of repeated observations of the same cross-section of actors (i.e. the Länder in the Bundesrat), it has to be analysed as panel data. This introduces the possibility of serial correlation of error terms both across time (due to unobserved unit-specific characteristics) and across panels (due to common exogenous shocks). There is also a substantial risk of non-constant variance of error terms within a panel, leading to panel heteroskedasticity. To deal with these issues, the models were estimated using fixed effects, and with robust standard errors16.

As can be seen from Figure 3 and Figure 4, there has been a general increase over time of the number of resolutions and bills initiated in the Bundesrat. This also holds if examining the data for the Western and Eastern Länder separately; the increase in initiatives is not driven by the increase in the number of Länder in 1990, although this of course had an effect17. This over-time increase suggests that the data may not be stationary. Conducting Hadri-Lagrange tests for the presence of a unit root indicates that there is evidence for non-stationarity in the data used for both dependent variables. The null hypothesis of no unit root can be rejected, regardless of whether it is conducted only on the Eastern Länder, the Western Länder or all Länder since German reunification, as well as for most of the individual Länder18.

15 No resolutions or bills were introduced on a Saturday or Sunday throughout the whole period under study.

16 The models were run using the xtnbreg command in Stata 12.

17 See appendices 8.2 and 8.3 for a breakdown of activities by Land.

18 Since the Hadri Lagrange test requires a strongly balanced panel, it was not possible to test all observations in a single test, due to the different dates for the first observations for the Western and Eastern Länder. The Eastern Länder did not join the Bundesrat on a the same day; rather,

However, introducing fixed time effects to control for this is problematic, for two reasons. First of all, since the time unit of the data is a single day, this would require several thousand dummy variables, which would lead to a very substantial increase in the computational power needed for a maximum likelihood estimation, to the point of being entirely unfeasible. Second, even introducing year dummies to correct for non- stationarity is problematic, since, if there has been a general increase over time in the tendency of Land governments to use their power of initiative within the Bundesrat to appeal to voters, then including time dummies would capture such an effect as well, underestimating the effect of electoral cycles on behaviour in the Bundesrat. Thus, to correct for stationarity, we follow Carter & Signorino (2010) and include a variable in the model which measures the number of days since a Land last initiated a bill or resolution, respectively, along with the squared and cubed version of this variable. This will capture a general increase in the tendency of Länder to initiate bills and resolutions, since this will necessarily lead to a decrease in the time between two instances of the same activity (Beck et al., 1998).

To test hypothesis 1 (days to election), we include a variable measuring, for each day in the data, the number of days until the next election in that Land. For elections that have not yet taken place at the time of writing, we assumed that the next election will take place either 4 or 5 years (depending on the constitution of the Land) after the last election held in that Land.

To test hypothesis 2 (days to federal election), we included a variable indicating the number of days remaining until the next federal election. To test hypothesis 3a (days until any Land election), we included a variable indicating the number of days until the next election in any of the 16 (or 11, pre-unification) Länder. We tested hypothesis 3b (days until co-partisan election/’moment of glory’) in two ways, reflecting two different

they joined as each Land from the GDR established a government. The difference between the first Eastern Land to join the Bundesrat (Meckelburg-Vorpommern) and the last (Sachsen and Thüringen) was however only 9 working days. When performing the Hadri-Lagrange tests, the first observations (all zeroes) were therefore (temporarily) dropped in order to ensure a balanced panel. All observations were used in the subsequent analysis.

ways in which to understand what governs Land activity in the Bundesrat. First, we included a variable indicating the number of days until the next election in any Land where the government included representatives of the same party as the Land in question19. For coalition governments, the lowest of the several values were used, reflecting the assumption that each coalition party wields a veto over what is proposed by the Land in the Bundesrat (and so allowing co-partisans to help the Land amongst them nearest an election). An alternatively specification assumes that what is proposed by a Land in the Bundesrat is controlled by that Land’s prime minister (Ministerpräsident), and measures the smallest number of days to election of any Land where the prime minister is a co-partisan. As all Land governments in the period studied includes either the CDU or SPD (or both), no observations are lost.

Table 1 provides descriptive statistics of the variables used in the analysis of the determinants of initiation. It should be noted that there are no gaps in the data; all resolutions and bills in the period covered are recorded in the dataset.

Table 1: Descriptive statistics

Variable Obs Mean Std. Dev. Min Max Number of bills initiated 146978 0,0199 0,1705 0 9 Number of resolutions initiated 136039 0,0080 0,0946 0 6 Days until next election 146978 807,82 489,87 1 1905 Days until next federal election 146978 691,01 406,88 1 1476 Days until any Land election 146978 126,39 124,18 1 791 Days until co-partisan Land election 146978 241,95 240,63 1 1829 Days until fellow PM Land election 146978 221,80 211,83 1 1780

It should be noted that, in addition to the incentives to propose legislation as elections approach in order to appeal to voters, there are also incentives for increased legislative activity immediately after elections. To a considerable extent campaigns for state elections focus on federal issues. In their manifestos regional parties address policy

19 We considered the CDU and the CSU as the same party for these purposes.

areas such as labour and social policy that are to be decided at the federal level (Pappi and Seher, 2009). Thereby they explicitly announce to initiate bills via the Bundesrat in order to change unfavourable policies via the national legislative process. Accordingly, submitting bills to the Bundesrat after one has gained control of a state government can be seen as an act of pledge fulfilment – although these bills rarely have a chance of being enacted (Harle and Stecker, 2011). This gives us. This, in combination with hypothesis 1 suggests that the legislative activity of Land governments in the Bundesrat will form a U- shape around state election dates. In order to control for this, we include a dummy variable equal to 1 whenever less than a year has passed since the last election in a Land.

5. Models and Analysis

As discussed above, we modelled the data as count panel data; since we have two ways of operationalising hypothesis 3b, we ran the models using each operationalisation in turn. Further, we also ran the models with and without the control variable for recent land elections, giving four models per dependent variable. The results are shown in Table 2 and Table 3. Results are presented in incidence rate ratios for ease of interpretation, meaning that effects over 1 indicate that a variable makes initiation more likely as it increases, while effects less than 1 indicate that the variable makes initiation less likely to occur as it increases.

Starting with the initiation of resolutions, as shown in Table 2, we see that there is no support for hypothesis 1 (days to next election), nor is there any significant effects of the federal electoral cycle (hypothesis 2). Together, this indicates a lack of a direct electoral cycle. However, when looking at hypothesis 3a (days to any Land election), we find a statistically significant effect (at the 1% level), indicating that the nearer any Land is to an election, the more active the other Länder become in terms of initiating resolutions in the Bundesrat (note that since the variable indicates the number of days until the election, an effect size less than 1 would indicate that a Land becomes less active the farther it is from an election, which is the equivalent of becoming more active the fewer days there are to an election). The effect size is relatively small, however. For every 10 days the Land nearest election is to Election Day, there is a 1% increase in the number of resolutions initiated by other Länder. However, it should be noted that this

effect is sensitive to how H3b (co-partisan nearer election) is operationalised. When H3b is operationalised in the sense of coalition parties having a mutual veto over Bundesrat initiatives, the effect of H3a is present; when H3b is operationalised in the sense of the partisan affiliation of the Land Ministerpräsident, there is no statistically significant effect. Regardless of the operationalisation of H3b, there is no evidence of intra-party cooperation in terms of the initiation of resolutions in the Bundesrat.

Table 2: Determinants of Resolution Initiation

(1) (2) (3) (4) Days to election 1.000043 1.000065 1.000117 1.000141 (0.000) (0.000) (0.000) (0.000) Days to Federal Election 0.999862 0.999861 0.999877 0.999876 (0.000) (0.000) (0.000) (0.000) Days to any Land election 0.999125** 0.999378 0.999062** 0.999310 (0.000) (0.000) (0.000) (0.000) Days to co-partisan election 1.000289 1.000283 (0.000) (0.000) Days to PM election 1.000045 1.000040 (0.000) (0.000) Year after election 0.861202 0.857908 (0.083) (0.083) Days since initiation 0.996495*** 0.996467*** 0.996539*** 0.996514*** (0.000) (0.000) (0.000) (0.000) Days since initiation (squared) 1.000001*** 1.000001*** 1.000001** 1.000001*** (0.000) (0.000) (0.000) (0.000) Days since initiation (cubed) 1.000000 1.000000 1.000000 1.000000 (0.000) (0.000) (0.000) (0.000) Constant 0.204148*** 0.207414*** 0.197086*** 0.200084*** (0.038) (0.039) (0.037) (0.038) N 136039 136039 136039 136039 aic 12067.01 12070.03 12066.60 12069.49 bic 12135.76 12138.78 12145.17 12148.06 chi2 . . . . * p<0.05, ** p<0.01, *** p<0.001; table shows IRRs; S.E.s in brackets

Turning to the determinants of bill initiation, a different pattern appears. While, as with resolutions, there is no statistically significant effect on Bundesrat behaviour of the number of days until a Land government faces re-election (hypothesis 1), there is a strongly significant effect of the number of days until the next federal election (hypothesis 2). The nearer the next federal election is, the less active the Länder

become; alternatively, the farther away the next federal election is, the more active Länder are. One possible explanation for this is that the electoral focus of the federal election is the Bundestag, which would be interfered with by activity by the Bundesrat. However, it should also be noted that this effect is relatively slight in magnitude; there is a roughly 2% decrease in activity for every 100 days nearer a federal election.

Interestingly, there is no evidence of cross-land collusion (H3a), but rather the opposite – every 10 days the Land nearest election is nearer to its election date leads to a 2% increase in the number of resolutions initiated, which is a greater effect size than what was found for resolutions above. However, unlike for resolutions, this holds regardless of how hypothesis 3b is operationalised, and also holds when controlling for any effect promise-fulfilling behaviour in the Bundesrat. The increase in activity as another Land nears election is of course the reverse of what we expected; as opposed to collusion, we find some evidence that the government of other Länder try to obstruct the Land nearest election, by attempting to draw attention to themselves (and so away from the Land in question. This argument is strengthened by the presence of intra-party cooperation (H3b), where co-partisans become less active the nearer one of their number is to an election (regardless of how this is operationalised. Thus, it appears that Land governments, while not directly becoming significantly more active as their election nears, nonetheless see their activity in the Bundesrat as relevant to their electoral prospects, as evidenced by the presence of both obstructing effects and attempts to help out by co-partisans20.

20 We also ran the models including a variable capturing the effect of promise-fulfilment from federal elections, including a dummy variable equal to 1 if less than a year had passed since the last federal election, and zero otherwise. The substantive results were the same; for the determinants of bill initiation, this added variable had a strongly significant and substantively relatively large effect, which the „days to federal election“ variable stopped being significant (and also went from a positive to a (mildly) negative effect). This indicates that the Länder take an active part in helping the federal government fulfil its election promises.

Table 3: Determinants of Bill Initiation

(1) (2) (3) (4) Days to election 0.999993 0.999975 1.000026 1.000002 (0.000) (0.000) (0.000) (0.000) Days to Federal Election 1.000187*** 1.000192*** 1.000191*** 1.000194*** (0.000) (0.000) (0.000) (0.000) Days to any Land election 0.998732*** 0.998548*** 0.998709*** 0.998533*** (0.000) (0.000) (0.000) (0.000) Days to co-partisan election 1.000243* 1.000240* (0.000) (0.000) Days to PM election 1.000446*** 1.000440*** (0.000) (0.000) Year after election 0.935217 0.946366 (0.056) (0.057) Days since initiation 0.994697*** 0.994738*** 0.994725*** 0.994762*** (0.001) (0.001) (0.001) (0.001) Days since initiation (squared) 1.000007** 1.000007** 1.000007** 1.000007** (0.000) (0.000) (0.000) (0.000) Days since initiation (cubed) 1.000.000 1.000.000 1.000.000 1.000.000 (0.000) (0.000) (0.000) (0.000) Constant 0.088277*** 0.087091*** 0.087187*** 0.086219*** (0.007) (0.007) (0.007) (0.007) N 146978 146978 146978 146978 aic 270973,10 270859,40 270980,60 270871,00 bic 271764,90 271651,20 271871,50 271761,80 chi2 204.52 216.64 205.71 217.45 * p<0.05, ** p<0.01, *** p<0.001; table shows IRRs; S.E.s in brackets

6. Conclusion

This paper provides a first look at the effects of the electoral incentives created by staggered electoral terms on behaviour in the German Bundesrat in terms of the initiating of resolutions and bills. While only slight effects, which were sensitive to differing operationalisations, were found in terms of resolutions, an interesting pattern emerged when we analysed the drivers of bill initiation. Here, we found that, in addition to a decrease in activity as a federal election nears, there is on the one hand a pattern of Länder increasing their activity when other Länder are nearer an election, in effect seeking to prevent these Länder from obtaining a finite amount of media and voter attention, and on the other hand, an attempt by co-partisans to counter this effect by becoming less active in the same situation. The combination of these two effects points

to a potentially very interesting pattern of behaviour in the Bundesrat, where, unlike what Shepsle et al (2009) found in the case of the US Senate, there is not only no cross- party collusion to benefit those legislators nearest election, but rather the opposite, namely within-party cross-time cooperation. That is, we found that co-partisans, recognising the importance of helping their fellow party members get re-elected (and so maintain or improve the voting power of their party in the Bundesrat) both attempt to undermine other Länder’s re-election attempts and support such attempts by their co- partisans by varying their behaviour in the Bundesrat according to the electoral cycles of the other members of the institution, providing evidence that staggered electoral terms lead to increases in re-election-seeking behaviour by legislators.

7. Bibliography

Adler, E.S., Lapinski, J.S., 1997. Demand-Side Theory and Congressional Committee Composition: A Constituency Characteristics Approach. American Journal of Political Science 41, 895–918. Aidt, T.S., Veiga, F.J., Veiga, L.G., 2011. Election results and opportunistic policies: A new test of the rational political business cycle model. Public Choice 148, 21–44. doi:10.1007/s11127-010-9644-3 Alesina, A., Roubini, N., 1992. Political cycles in OECD economies. The Review of Economic Studies 59, 663–688. Alter, A.B., 2002. Minimizing the risks of delegation: multiple referral in the German Bundesrat. American Journal of Political Science 299–315. Alvarez, R.M., Saving, J.L., 1997. Deficits, Democrats, and Distributive Benefits: Congressional Elections and the Pork Barrel in the 1980s. Political Research Quarterly 50, 809–831. doi:10.1177/106591299705000405 Ansolabehere, S., Snyder, J.M., 2006. Party Control of State Government and the Distribution of Public Expenditures. Scandinavian Journal of Economics 108, 547–569. doi:10.1111/j.1467-9442.2006.00470.x Arulampalam, W., Dasgupta, S., Dhillon, A., Dutta, B., 2009. Electoral goals and center- state transfers: A theoretical model and empirical evidence from India. Journal of Development Economics 88, 103–119. doi:10.1016/j.jdeveco.2008.01.001 Basic Law of Germany, 2010. Bebchuk, L.A., Coates IV, J.C., Subramanian, G., 2002a. The Powerful Antitakeover Force of Staggered Boards: Theory, Evidence, and Policy. Stanford Law Review 54, 887– 951. Bebchuk, L.A., Coates IV, J.C., Subramanian, G., 2002b. The Powerful Antitakeover Force of Staggered Boards: Further Findings and a Reply to Symposium Participants. Stanford Law Review 55, 885–917. Bebchuk, L.A., Cohen, A., 2005. The costs of entrenched boards. Journal of Financial Economics 78, 409–433. doi:10.1016/j.jfineco.2004.12.006 Beck, N., Katz, J.N., Tucker, R., 1998. Taking Time Seriously: Time-Series-Cross-Section Analysis with a Binary Dependent Variable. American Journal of Political Science 42, 1260–1288. Blais, A., Nadeau, R., 1992. The electoral budget cycle. Public Choice 74, 389–403. Bräuninger, T., König, T., 1999. The checks and balances of party federalism: German federal government in a divided legislature. European Journal of Political Research 36, 207–234.

Brechler, J., Gersl, A., 2011. Political Legislation Cycle in the Czech Republic (IES Working Paper No. 21/2011). Institute of Economic Studies, Faculty of Social Sciences, Charles University in Prague. Brollo, F., Nannicini, T., 2012. Tying Your Enemy’s Hands in Close Races: The Politics of Federal Transfers in Brazil. American Political Science Review 106, 742–761. doi:10.1017/S0003055412000433 Burkhart, S., 2005. Parteipolitikverflechtung. Über den Einfluss der Bundespolitik auf Landtagswahlentscheidungen von 1976-2000. Politische Vierteljahresschrift 46, 14–38. Campion, Lord, 1953. Second chambers in theory and practice. Parliamentary Affairs 7, 17–32. Carter, D.B., Signorino, C.S., 2010. Back to the Future: Modeling Time Dependence in Binary Data. Political Analysis 18, 271–292. doi:10.1093/pan/mpq013 Caselli, F., Michaels, G., 2013. Do Oil Windfalls Improve Living Standards? Evidence from Brazil. American Economic Journal: Applied Economics 5, 208–238. doi:10.1257/app.5.1.208 Cremers, K.J.M., Nair, V.B., 2005. Governance mechanisms and equity prices. The Journal of Finance 60, 2859–2894. Dahlberg, M., Johansson, E., 2002. On the vote-purchasing behavior of incumbent governments. American Political Science Review 96, 27–40. Daines, R., Klausner, M., 2001. Do IPO Charters Maximize Firm Value? Antitakeover Protection in IPOs. Journal of Law, Economics & Organization 17, 83–120. Detterbeck, K., Jeffery, C., 2009. Rediscovering the Region: Territorial Politics and Party Organizations in Germany, in: Swenden, W., Maddens, B. (Eds.), Territorial Party Politics in Western Europe. Palgrave Macmillan, New York, pp. 63–85. Drazen, A., Eslava, M., 2010. Electoral manipulation via voter-friendly spending: Theory and evidence. Journal of Development Economics 92, 39–52. doi:10.1016/j.jdeveco.2009.01.001 Duru, A., Wang, D., Zhao, Y., 2013. Staggered boards, corporate opacity and firm value. Journal of Banking & Finance 37, 341–360. doi:10.1016/j.jbankfin.2012.09.002 Elgie, R., McMenamin, I., 2005. Credible Commitment, Political Uncertainty or Policy Complexity? Explaining Variations in the Independence of Non-majoritarian Institutions in France. British Journal of Political Science 35, 531–548. doi:10.1017/S0007123405000281 Faleye, O., 2009. Classified Boards, Stability, and Strategic Risk Taking. Financial Analysts Journal 65, 54–65. Foucault, M., Madies, T., Paty, S., 2008. Public spending interactions and local politics. Empirical evidence from French municipalities. Public Choice 137, 57–80. doi:10.1007/s11127-008-9312-z Galli, E., Rossi, S.P., 2002. Political budget cycles: the case of the Western German Länder. Public Choice 110, 283–303.

Garcia De Paso, J.I., 2000. Partisan appointments to the central bank: Policy uncertainty and the democratic deficit. Journal of Macroeconomics 22, 471–489. Gilardi, F., Maggetti, M., 2010. The independence of regulatory authorities, in: Levi-Faur, D. (Ed.), Handbook on Regulation. Edward Elgar Publishing Limited, Cheltenham. Golden, M.A., Picci, L., 2008. Pork-Barrel Politics in Postwar Italy, 1953–94. American Journal of Political Science 52, 268–289. Gompers, P., Ishii, J., Metrick, A., 2003. Corporate governance and equity prices. The Quarterly Journal of Economics 118, 107–156. Grossman, P.J., 1994. A political theory of intergovernmental grants. Public Choice 78, 295–303. Guidi, M., 2013. Delegation and varieties of capitalism: Explaining the independence of national competition agencies in the European Union. Comparative European Politics. doi:10.1057/cep.2013.6 Hamilton, A., Jay, J., Madison, J., 2001. The Federalist, The Gideon edition. ed. Liberty Fund, Indianapolis. Hanretty, C., 2010. Explaining the De Facto Independence of Public Broadcasters. British Journal of Political Science 40, 75–89. doi:10.1017/S000712340999024X Hanretty, C., Koop, C., 2012. Measuring the formal independence of regulatory agencies. Journal of European Public Policy 19, 198–216. doi:10.1080/13501763.2011.607357 Harle, I., Stecker, C., 2011. Die Initiativtätigkeit des Bundesrates im Lichte der Parteipolitisierungshypothese. Zeitschrift für Parlamentsfragen 42, 325–334. Heller, W.B., 2007. Divided Politics: Bicameralism, Parties, and Policy in Democratic Legislatures. Annual Review of Political Science 10, 245–269. doi:10.1146/annurev.polisci.10.071105.112758 Hix, S., Høyland, B., Vivyan, N., 2010. From doves to hawks: A spatial analysis of voting in the Monetary Policy Committee of the Bank of England. European Journal of Political Research 49, 731–758. doi:10.1111/j.1475-6765.2010.01916.x Jeffery, C., Hough, D., 2001. The Electoral Cycle and Multi-Level Voting in Germany. German Politics 10, 73–98. doi:10.1080/772713264 Jiraporn, P., Chintrakarn, P., 2009. Staggered Boards, Managerial Entrenchment, and Dividend Policy. Journal of Financial Services Research 36, 1–19. doi:10.1007/s10693-009-0059-6 Jiraporn, P., Chintrakarn, P., Kim, Y.S., 2012. Analyst following, staggered boards, and managerial entrenchment. Journal of Banking & Finance 36, 3091–3100. doi:10.1016/j.jbankfin.2012.07.013 Jiraporn, P., Gleason, K.C., 2007. Capital structure, shareholder rights, and corporate governance. Journal of Financial Research 30, 21–33. Jiraporn, P., Liu, Y., 2008. Capital Structure, Staggered Boards, and Firm Value. Financial Analysts Journal 64, 49–60.

Johansson, E., 2003. Intergovernmental grants as a tactical instrument: empirical evidence from Swedish municipalities. Journal of Public Economics 87, 883–915. Knight, B., 2005. Estimating the value of proposal power. The American Economic Review 95, 1639–1652. Kovats, L., 2009. Do elections set the pace? A quantitative assessment of the timing of European legislation. Journal of European Public Policy 16, 239–255. doi:10.1080/13501760802589255 Lagona, F., Maruotti, A., Padovano, F., 2012. The opposite cycles of laws and decrees. Condorcet Centre for Political Economy Working Paper 2012-01-ccr. Lagona, F., Padovano, F., 2008. The political legislation cycle. Public Choice 134, 201– 229. doi:10.1007/s11127-007-9219-0 Lecours, A., Beland, D., 2009. Federalism and Fiscal Policy: The Politics of Equalization in Canada. Publius: The Journal of Federalism 40, 569–596. doi:10.1093/publius/pjp030 Levitt, S.D., Snyder, J.M., 1995. Political Parties and the Distribution of Federal Outlays. American Journal of Political Science 39, 958–980. Levitt, S.D., Snyder, J.M., 1997. The Impact of Federal Spending on House Election Outcomes. The Journal of Political Economy 105, 30–53. Lindner, A., 2000. Long-term appointment of central bankers: costs and benefits. European Journal of Political Economy 16, 639–654. Majone, G., 1996. Temporal Consistency and Policy Credibility: Why Democracies Need Non-Majoritarian Institutions. EUI Working Paper 96. Manow, P., Burkhart, S., 2007. Legislative Self-Restraint under Divided Government in Germany, 1976-2002. Legislative Studies Quarterly 32, 167–191. Martin, L.W., 2004. The government agenda in parliamentary democracies. American Journal of Political Science 48, 445–461. Mayhew, D.R., 1974. Congress: The Electoral Connection. Yale University Press, New Haven, CT. Miller, B., Stecker, C., 2008. Consensus by Default? Interaction of Government and Opposition Parties in the Committees of the German Bundestag. German Politics 17, 305–322. doi:10.1080/09644000802300551 Money, J., Tsebelis, G., 1992. Cicero’s Puzzle: Upper House Power in Comparative Perspective. International Political Science Review 13, 25–43. doi:10.1177/019251219201300103 Monteiro, J., Ferraz, C., 2012. Does oil make leaders unaccountable? Evidence from Brazil’s offshore oil boom. Working Paper. Müller, J., 2013. On a Short Leash? Sub-National Party Positions between Regional Context and National Party Unity. Journal of Elections, Public Opinion and Parties 23, 177–199.

Nordhaus, W.D., 1975. The political business cycle. The Review of Economic Studies 42, 169–190. Pappi, F.U., Seher, N.M., 2009. Party Election Programmes, Signalling Policies and Salience of Specific Policy Domains: The German Parties from 1990 to 2005. German Politics 18, 403–425. doi:10.1080/09644000903055831 Patterson, S.C., Mughan, A., 1999. Senates and the Theory of Bicameralism, in: Patterson, S.C., Mughan, A. (Eds.), Senates. Bicameralism in the Contemporary World. Ohio State University Press, Columbus. Patzelt, W.J., 1999. The Very Federal House: The German Bundesrat, in: Patterson, S.C., Mughan, A. (Eds.), Senates. Bicameralism in the Contemporary World. Ohio State University Press, Columbus. Potrafke, N., 2012. Political cycles and economic performance in OECD countries: empirical evidence from 1951–2006. Public Choice 150, 155–179. doi:10.1007/s11127-010-9695-5 Renzsch, W., 1999. Party competition in the German federal state: New variations on an old theme. Regional & Federal Studies 9, 180–192. Rich, M.J., 1989. Distributive Politics and the Allocation of Federal Grants. American Political Science Review 83, 193–213. Riker, W.H., 1992a. The Justification of Bicameralism. International Political Science Review 13, 101–116. doi:10.1177/019251219201300107 Riker, W.H., 1992b. The merits of bicameralism. International Review of Law and Economics 12, 166–168. Rogers, J.R., 2001. An Informational Rationale for Congruent Bicameralism. Journal of Theoretical Politics 13, 123–151. doi:10.1177/0951692801013002001 Rogoff, K., Sibert, A., 1988. Elections and macroeconomic policy cycles. The Review of Economic Studies 55, 1–16. Russell, M., 2001. What are Second Chambers for? Parliamentary Affairs 54, 442–458. Sakurai, S.N., Menezes-Filho, N., 2011. Opportunistic and partisan election cycles in Brazil: new evidence at the municipal level. Public Choice 148, 233–247. doi:10.1007/s11127-010-9654-1 Sakurai, S.N., Menezes-Filho, N.A., 2008. Fiscal policy and reelection in Brazilian municipalities. Public Choice 137, 301–314. doi:10.1007/s11127-008-9329-3 Schneider, C.J., 2010. Fighting with one hand tied behind the back: political budget cycles in the West German states. Public Choice 142, 125–150. doi:10.1007/s11127- 009-9480-5 Schultz, K.A., 2009. The Politics of the Political Business Cycle. British Journal of Political Science 25, 79. doi:10.1017/S0007123400007079 Shepsle, K.A., Van Houweling, R.P., Abrams, S.J., Hanson, P.C., 2009. The senate electoral cycle and bicameral appropriations politics. American Journal of Political Science 53, 343–359.

Shepsle, K.A., Weingast, B.R., 1981. Political Preferences for the Pork Barrel: A Generalization. American Journal of Political Science 25, 96–111. Snyder, S.K., Weingast, B.R., 2000. The American System of Shared Powers: The President, Congress and the NLRB. Journal of Law, Economics & Organization 16, 269–305. Solé-Ollé, A., Sorribas-Navarro, P., 2008. The effects of partisan alignment on the allocation of intergovernmental transfers. Differences-in-differences estimates for Spain. Journal of Public Economics 92, 2302–2319. doi:10.1016/j.jpubeco.2007.06.014 Sørensen, R.J., 2003. The political economy of intergovernmental grants: The Norwegian case. European Journal of Political Research 42, 163–195. Stratmann, T., Baur, M., 2002. Plurality Rule, Proportional Representation, and the German Bundestag: How Incentives to Pork-Barrel Differ across Electoral Systems. American Journal of Political Science 46, 506–514. Tsebelis, G., Money, J., 1997. Bicameralism. Cambridge University Press, Cambridge. Tsebelis, G., Rasch, B.E., 1995. Patterns of Bicameralism, in: Döring, H. (Ed.), Parliaments and Majority Rule in Western Europe. Mannheim Centre for European Social Research (MZES), Mannheim. Veiga, L.G., Pinho, M.M., 2007. The political economy of intergovernmental grants: Evidence from a maturing democracy. Public Choice 133, 457–477. doi:10.1007/s11127-007-9208-3 Veiga, L.G., Veiga, F.J., 2007. Political business cycles at the municipal level. Public Choice 131, 45–64. doi:10.1007/s11127-006-9104-2 Waller, C.J., 1989. Monetary policy games and central bank politics. Journal of Money, Credit and Banking 21, 422–431. Waller, C.J., 1992. A bargaining model of partisan appointments to the central bank. Journal of Monetary Economics 29, 411–428. Weare, V., 1965. The House of Lords - Prophecy and Fulfilment. Parliamentary Affairs 18, 422–433. Weingast, B.R., Shepsle, K.A., Johnsen, C., 1981. The political economy of benefits and costs: A neoclassical approach to distributive politics. The Journal of Political Economy 642–664. Wonka, A., Rittberger, B., 2010. Credibility, Complexity and Uncertainty: Explaining the Institutional Independence of 29 EU Agencies. West European Politics 33, 730– 752. doi:10.1080/01402381003794597 Worthington, A.C., Dollery, B.E., 1998. The political determination of intergovernmental grants in Australia. Public Choice 94, 299–315. Yoo, K.-R., 1998. Intervention analysis of electoral tax cycle: the case of Japan. Public Choice 96, 241–258.

8. Appendices

8.1. Table of Elections

Table 4: Voting weights and recent elections in the German Länder21

Next scheduled Land Votes in BR N-1 election Last election election Baden- 6 26/03/2006 27/03/2011 2016 Württemberg Bayern 6 28/09/2008 15/09/2013 2018 Berlin 4 17/09/2006 18/09/2011 2016 Brandenburg 4 19/09/2004 27/09/2009 14/09/2014 Bremen 3 13/05/2007 22/05/2011 2015 Hamburg 3 24/04/2008 20/02/2011 2015 Hessen 5 18/01/2009 22/09/2013 2018 Niedersachsen 6 27/01/2008 20/01/2013 2018 Mecklenburg- 3 17/09/2006 04/09/2011 2016 Vorpommern Nordrhein- 6 09/05/2010 13/05/2012 2017 Westfalen Rheinland- 4 26/03/2006 27/03/2011 2016 Pfalz Saarland 3 20/09/2009 25/03/2012 2017 Sachsen 4 19/09/2004 20/09/2009 31/08/2014 Sachsen-Anhalt 4 26/03/2006 20/03/2011 2016 Schleswig- 4 27/09/2009 06/05/2012 2017 Holstein Thüringen 4 13/06/2004 30/09/2009 14/09/2014

21 Source: http://www.bundeswahlleiter.de/de/landtagswahlen/wahltermine/index.html

8.2. Bill Initiation Graphs.

8.3. Resolution Initiation Graphs