Resilience and Change in Federal Institutions: The Case of the German Federal Council

One of the oldest federalist systems, Germany offers itself as a case study for long-term developments in federalism. Drawing on a burgeoning literature on institutional continuity and change we investigate the development of a key institution of German federalism, the Federal Council (Bundesrat) from the foundation of the German Reich until today. Counter to claims that institutional change occurs mainly during “critical junctures,” the Federal Council has shown remarkable resilience: It persisted through World War I and the 1919 revolution and the writing of the centralist Weimar constitution. Dismantled in 1934, it returned in 1949 after years of dictatorship, war, and military occupation. Counter to an emphasis on institutional stability in “settled” times in earlier literature, the role of the Bundesrat has changed significantly over the last 60 years. Initially representing the interests of federal states, it has gradually developed into a powerful second chamber dominated by national-level politics. 1 We thank the participants in the Darmstadt workshop, and especially Fritz Scharpf, Jörg Broschek, and Arthur Benz, for extremely helpful comments on this chapter. An earlier version of this paper was presented at Syracuse University, where comments by Pablo Beramendi in particular proved invaluable to improving the overall argument.


Introduction
The study of federalism has undergone something of a renaissance in the last two decades (Erk and Swenden 2010;Hueglin and Fenna 2006). Transitions from authoritarianism to democracy in Latin America and Eastern Europe have been especially important in rekindling interest, but a more general trend toward political decentralization involving devolution of political responsibilities and sometimes the "activation" of previously dormant territorially based representation has also contributed to the current wave of interest in the effects of federalism on policy and politics generally (Gibson 2004).
One important stream of scholarship has focused on the stability of democratic federalism, asking the question of what makes federalism robust (or "self-enforcing") over time (e.g., Bednar, Eskridge and Ferejohn 2001;Bednar 2009). In line with the themes at the heart of this volume, this chapter seeks to go beyond the question of stability or breakdown to provide insight into how federal institutions evolve over time. As one of the oldest federal systems, Germany is a case that allows us to analyze the development of federal institutions over a much longer stretch of time than is possible with more newly minted varieties.
We focus special attention on the German Federal Council (the Bundesrat), a cornerstone of German federalism from the beginning, and an institution that has exhibited remarkable resilience over the past 140 years. When a unified German state was formed in 1871 by several independent princedoms, the Bundesrat was a central component in the constitution. This upper parliamentary chamber consisted of emissaries from all state governments, weighted by, but not proportional to states' size. The imperial government broke down when Germany was defeated in World War I, and after a revolutionary period, the Weimar constitution entered into force. The upper chamber devised for Germany's first democracy (renamed Reichsrat) again consisted of emissaries from all state governments, weighted by, but not proportional to states' size. Finally, following Germany's capitulation in World War II, after 12 years of dictatorship and three years of occupation, a new constitution was crafted in 1948 and became the Basic Law that governs Germany until today. It establishes an upper chamber (called, again, the Bundesrat) that consists of emissaries from all state governments, weighted by, but not proportional to states' size.
The resilience of the Bundesrat through otherwise order-shattering events in Germany's recent history presents an intriguing puzzle when viewed against the backdrop of "punctuated equilibrium" models that tend to dominate the literature on institutional change in comparative politics. Such models encourage us to think about institutional development as characterized by long periods of stable institutional reproduction punctuated periodically by moments of intense institutional reconfiguration. As such, this literature has us mostly looking for "big" institutional changes in historic breakpoints, and for stable institutional reproduction the rest of the time.
Whereas institutions are seen to be mostly constraining under conditions of stability, "critical juncture" moments are thought to ease the usual constraints on action and open the door to significant institutional transformation (e.g., Katznelson 2003).
One puzzle, then, around which the following analysis is organized is to explain the surprising durability of the Bundesrat through massive historical break points following Germany's defeat and reconstitution after the first and second world wars. These episodes would surely qualify as "critical junctures" by almost any definition, and they were moments in which the actors who were politically empowered by the transitions in fact preferred very different and more centralizing constellations. In both episodes, the Bundesrat figured centrally in conflicts between centralizers and advocates of strong states rights, and although the balance sometimes shifted, the Bundesrat survived.
However, there is a second puzzle as well, and in some ways it is the mirror image of the first.
Here the challenge is to explain significant changes in the role and functions of the Bundesrat since 1949, a period that most observers would almost certainly code as "settled" rather than "unsettled" times, to use Ann Swidler's terms (Swidler 1986). However, the operation of German federalism has in fact changed significantly in this period. The Federal Republic of Germany was re-founded after World War II as a "unitary federal state" (Hesse 1962), with all the ambiguities the phrase itself suggests. Since then German federalism has increasingly evolved into what Scharpf characterizes as a "joint decision system" prone to political paralysis and stalemate (Scharpf 2005). Changes in the role and functions of the Bundesrat figure prominently in these developments. Repeatedly defended against more unitarist impulses as a mechanism to safeguard state rights, the Bundesrat survived, ironically, only to become an instrument for the unitarization of power at the expense of state autonomy. The "functional conversion" of the Bundesrat -from a "states rights" chamber to a powerful player in national decision-making --was not anticipated by the institution's designers and defenders (Neunreither 1959: 713). The shift was also not accomplished in a single well-defined moment of institutional (re) design; this outcome, rather, emerged incrementally through a series of discrete political shifts and small but consequential judicial decisions in the post-WWII period that cumulated to produce a very significant reconfiguration of institutional arrangements in the absence of any single "break point." This chapter examines the evolution of German federalism and the Bundesrat in both aspects -formal institutional survival through historic break points and gradual functional change beneath the surface of apparent institutional continuity-applying insights from recent writing in historical institutionalism to the study of federalism.

The Evolution of German Federalism: An Overview
To frame the discussion to follow, this section provides a schematic overview of the evolution of German federalism over the past century and a half. We can capture the relevant changes with reference to a theoretical framework proposed by Gibson and Falleti (Gibson and Falleti 2004).
Gibson and Falleti distinguish two dimensions along which federal systems vary. The first dimension concerns the balance of power between national and state (provincial) governments, and it is captured by the distinction between "centralized" versus "peripheralized" variants of federalism. In strongly centralized systems power is concentrated at the national level (i.e., the center dominates the states), while in peripheralized systems state governments exercise considerable powers vis-à-vis the center. A second dimension concerns relations among the states themselves, and involves the distinction between "hegemonic" and "pluralistic" varieties of federalism (Gibson and Faletti 2004: 227). Hegemonic federalism depicts a situation in which a single state or province is able to exercise disproportionate influence over the others; in pluralistic varieties of federalism, power is more evenly dispersed among the various states.
Combining these two dimensions produces the analytic space shown in figure 1, with which we can track changes in German federalism over time. The original design of the Imperial Constitution of 1871 made Germany an almost textbook example of "hegemonic peripheralized" federalism, combining as it did significant decentralized powers for individual states but with one state (Prussia) clearly dominant. The Weimar period marked a shift to a more centralized but still hegemonic version of federalism; power became more centralized as the national government assumed control over fiscal matters, but plans to reconfigure the states (and dismantle Prussia) failed. In the post World War II period, the redrawing of the national and state maps (eliminating Prussia) produced a more plural variety of federalism, which by design at least also provided for very significant state autonomy (peripheralized and plural, therefore, in Gibson and Falleti's scheme). Over time, however, subtle but cumulatively profound changes moved Germany strongly in the direction of greater unitarization, as the autonomy of state legislatures shrank in tandem with an increase in the power of the Federal chamber at the national level.
[ Figure 1 here] The following sections examine these developments in turn, paying particular attention to changes in the role and functions of the Bundesrat as a central player -and field of conflict --in the evolution of German federalism. German Reich was born federal --a "union of twenty five German states of various sizes and forms of government" (Craig 1980: 39). The core institutions of the new German nation-state were the national executive (the emperor and his chancellor), a national parliament (Reichstag) elected by universal male suffrage and secret ballot, and a federal council (Bundesrat) composed of appointed delegates representing the governments of the constituent states.

Founding Institutions and the Origins of the Bundesrat
Prussia was clearly the dominant force, 2 but the federal design made Prussian hegemony in the unified Reich more palatable to the other states by evoking the structure of the delegate council that provided coordination during the years of the German Confederation (Deutscher Bund, 1815-1866), when representatives of autonomous states had convened as equals (Lehmbruch 2000: 78-79;2002: 41). Along with arrangements guaranteeing the states very significant levels of fiscal autonomy, 3 the establishment of a council in which state governments would be directly represented in national decision-making soothed some of the concerns of the more independent states, as it underlined the federative character of the new government. 4 In fact, each of the state delegations to the Bundesrat had its own embassy and its members enjoyed diplomatic status. In the subsequent functioning of the Reich as well, large states were regularly consulted before the introduction of important legislation (Lehmbruch 2000: 80;2002: 44).
Beyond its role in facilitating unification, the Bundesrat played a key part in stabilizing the authoritarian imperial government in a context marked by considerable pressures for democratic opening emanating both from neighboring countries and from within Germany itself (Nipperdey 1992). The democratically elected lower house (Reichstag) provided a kind of fig leaf legitimacy but real power rested with the Bundesrat, which was dominated by Prussia, and whose state legislature was notoriously undemocratic. 5 More precisely, and as Otto Pflanze has put it, the Bundesrat was a key component in an elaborate balancing act that allowed Bismarck to leverage multiple potentially disorganizing tendencies. "Centralistic and particularistic institutions were to be played off against each other. Pressure would be met by counter pressure: the nation against the dynasties, the confederation against Prussia, Reichstag against Bundesrat, parliament against parliament, centralism against particularism, the centripetal against the centrifugal…" (Pflanze 1990: 346-48).
Given its history and role in the German imperial state, the Bundesrat was strongly associated with authoritarianism. 6 Deeply anti-democratic in its composition, the institution represented a "cartel of princes and bureaucrats" (Lehmbruch 2002: 37) who actively inhibited the development and exercise of democracy. The opposition Social Democrats understandably committed themselves in this period to dismantling the Bundesrat and establishing a unitary system of government. As Engels put it in 1891, it would be impossible "to revolutionize society while there are [state] reservation rights," and noted that "the proletariat can only use the form of the one and indivisible republic" (Engels 1990(Engels [1891: 227-28). Specifically, Preuß favored a "decentralized unitary state" with power concentrated at the national level and where states would survive but in a completely reconfigured form. Preuß regarded the existing states as relicts from an authoritarian past (Apelt 1946: 60). He envisioned redrawing the map to create eight wholly new states (above all: more equal in size) and subordinated to the powers of the central government. Given the dominance of Social Democrats at both the national level and in state governments 7 and the strong preference for a unitary system of government on the part of those charged with crafting a new constitution, one might have expected a complete break with traditional institutional arrangements.

Politics in Critical Junctures: The limits to institutional redesign
Yet the refounding exhibits a more complex mix of continuities and changes (Schultz 1987: 198).
When the Weimar constitution was signed into law on August 11, 1919, it not only left traditional 7 As Schulz notes, one year after the toppling of the monarchy, Social Democrats were the majority party in thirteen of Germany's states, including in the three largest states (Schultz 1987: 198). state boundaries intact, but also established the Reichsrat, an institution strikingly similar in its composition and design to the imperial Bundesrat. As before, state representatives to the Council would not be elected but would instead be direct emissaries of state governments. 8 Even though the center of power shifted sharply toward the national assembly (Reichstag) (Schultz 1987: 104), the Reichsrat retained some significant rights, specified in Articles 60-67, 69 and 74 of the Weimar Constitution (see also Craig 1980: 419). The constitution stipulated that every legislative initiative by the government had to be presented to the Reichsrat. If the Reichsrat did not consent, its dissenting opinion needed to be presented together with the proposal to the Reichstag.
Moreover, the Reichsrat could object to any law passed by the Reichstag, and overruling such objections required either a 2/3 majority in the Reichstag or a majority in a plebiscite called by the president.
Prussia was a main target of Preuß's reform ambitions, and the failure of his plan to break up the state by redrawing state boundaries meant that Prussia survived the transition to democracy.
Decision-making within the new government was more centralized, a consequence above all of fiscal reforms (the Erzberger reforms) that freed the central government from dependence on the states for financing. And although some limits were placed on Prussian influence in the Reichsrat, 9 Prussia would continue to loom relatively large in German politics. Invoking Gibson and Falleti (2004) again, one could speak of a transition to a more centralized version of a hegemonic federal regime -the hegemon, Prussia, survived the transition although the center of power (especially in financial matters) shifted to the national government.

Federalism in the SPD
9 In a provision designed to limit Prussian influence, no state could have more than 2/5 of the votes in the Reichsrat (Prussia had about 60% of the population). For the Prussian territory, half of the delegates would represent the provinces and half the state of Prussia itself.
Politics of the SPD, one of the bigger parties supporting centralization, is particularly instructive.
In the nineteenth century Germany's Social Democratic Party had developed a reputation as the country's most strictly organized, centralized and disciplined party, a legacy of the original organizational design but also the result of decades of semi-clandestine activity under the anti socialist laws (Nipperdey 1961: 386;Michels 1911Michels , 1966Michels [1925). But the reputation is at odds with the reality, since the party in fact was more decentralized and diverse (organizationally as well as ideologically) than commonly thought. Political conditions in Germany's various states diverged considerably in the imperial period. The Left was vigorously repressed in Prussia but in some of the southern and southwestern states the Social Democratic Party enjoyed more freedom to organize and to contest elections.
Local and state level party leaders adapted their strategies to local conditions. In Prussia, repression and constraints on organization -extending well beyond the expiration of the antisocialist laws in 1890 ---fuelled a radical ideology. There, anti-combination laws prevented statewide assemblies (the first state party conference in Prussia did not take place until 1905), and electoral laws were skewed to systematically under-represent the Left. However in other German states, including Baden, Hessen, Bavaria, Württemberg, and Saxony, a more liberal climate prevailed. In these states, social democrats organized and contested elections, achieving representation in state governments by 1900 (Miller and Potthoff 1986;Schultz 1987;von Freyberg 1975). In these states, party leaders regularly faced important choices about whether to align themselves with bourgeois parties in electoral coalitions and frequently traded their support on various government bills for more democratic electoral laws.
The diversity of conditions under which the SPD operated in different states set the scene for consequential battles within the party of how much autonomy regional leaders ought to enjoy. 10 Party elites in the more liberal states chafed under what they considered the overly doctrinaire line of (Prussian) central party leaders. Already in 1891, Georg von Vollmar, the powerful reformist leader of the Bavarian Social Democratic Party, had rankled national leaders by suggesting that the party should cooperate with the dominant bourgeois powers on military issues and policies toward independent farmers. Such ideas were considered treasonous in some party circles and the issue of centralization versus state autonomy in the party was a repeated subject of debate at party congresses in the early 1900s.
At the 1905 congress the southern German states sponsored a proposal for the federalization of the party structure itself, precipitating a heated debate over party organization. Although the southerners' proposal for decentralization was defeated at that congress, the issue did not go away. It reached a head in 1908 when southern delegations were charged with breach of discipline for having voted for the national budget, and some of the hardliners at party headquarters (e.g., Clara Zetkin) sought to have them expelled. 11 But unity trumped discipline and no sanctions were imposed. While the party congress reaffirmed its commitment to centralization, there nonetheless remained significant space for autonomy even if the southerners were never successful in their bid to get the party reorganized officially around federal principles.
These divisions did not vanish in the transition to democracy after WWI; indeed, in the rocky period following Germany's capitulation, the emerging leadership in southern states -led by Bavaria --put all their efforts into securing state autonomy. Within the Bavarian state party, reformist and revolutionary elements locked horns on all conceivable programmatic issues, but they spoke with one voice on the question of state autonomy and federalism. Erhard Auer, Vollmar's protégé and, like his mentor, a dedicated reformist, was firmly committed to state autonomy. But so was the Bavarian revolutionary Kurt Eisner, who wanted the same outcome but for very different reasons, namely to avoid subordination to the more moderate party elite that 11 The national party explicitly rejected supporting the national budget on principle, since the budget gave the government resources to pursue its authoritarian policies. had in the meantime assumed power in Berlin (Nipperdey 1961: 387-90;Schultz 1987; see also Mehringer 1992). Bavaria was not alone, either. Already in December 1918 the southern German states had held a "conference of states" (Staatenkonferenz) to establish a common position to bring into national discussions.
State governments were up and running much earlier than the national government. Separate revolutions in different regions gave power to self-confident and independent minded (also now, democratic) state governments (Peukert 1987: 46). Most of these were quickly able to solidify their power (Apelt 1946: 40). Partly this was a function of high continuity in local and state bureaucracies, as new democratic governments "slipped into the empty hull" of the previous state structures (Schultz 1987: 104), leaving the administration intact and largely unchanged.
Moreover, in the context of extreme national upheaval, the country's president Friedrich Ebert needed regional elites to restore order and re-establish authority. Just as Bismarck had needed state elites to solidify authoritarianism, so too did national political elites in the early years of the democratic Weimar Republic need them to stabilize democracy, and for this compromises were necessary.

The role of Länder bureaucrats
When state representatives received Preuß's constitutional draft on January 20 th 1919, they vehemently opposed it as too unitarist. The states were invited to a joint conference with the federal government on January 25 th 1919. Participating in this conference on behalf of the states were not only their revolutionary governments (mostly SPD or USPD), but also large numbers of bureaucrats, many of whom had already been part of the imperial Bundesrat (Jasper 1992: 125f).
The tone of the meeting has been characterized as "more ancien regime than revolution," and many scholars regard the convening of the conference as a definite decision for a more federal solution (Rose 1964: 22;Apelt 1946: 177).
Delegates to the conference put up ferocious resistance to the idea of restructuring the states.
Even southern elites argued against ideas that had been floated to divide Prussia into a number of smaller states, not out of loyalty to Prussia but because they viewed this as a strategy to weaken the power of the states generally. State representatives successfully argued for the establishment of a permanent council of states (Staatenausschuss) to accompany the constitutional process. The federal government would put all of its recommendations before the council (Schultz 1987: 187 and passim). In all areas where agreement with the states could not be reached, both positions were to be presented to the constitutional convention, though only in questions of the integrity of state territory were the states to have an outright veto (Apelt 1946: 68).
The states' representatives in the council would not be elected, but rather, direct emissaries of state governments --thus mirroring the past Bundesrat and foreshadowing the future Reichsrat.
Once convened, the council was also successful in convincing the government to agree to pass a preliminary constitution that would serve as a legal basis during the negotiations of the constitutional assembly. On Feb 6, 1919, a "preliminary law governing central-state power" [Vorläufiges Gesetz über die Reichsgewalt], was passed by the national assembly (Nationalversammlung) without significant debate. This short document predetermined many of the later structures of government, including the existence of a second chamber based on direct representation of state governments at the national level (Jasper 1992: 127).
In sum, the result of this first episode in institutional redesign was heavily laced with elements from the previous imperial model (Craig 1980), an instance of what Slater and Simmons might call a "critical antecedent" (Slater and Simmons 2010). The new constitution involved some diminution of state power, but much survived the rupture, and the Weimar constitution would likely have been more federalist still had Bavaria (the most vociferous proponent of state autonomy) not sunk into chaos. The crucial Erzberger reforms regulating state finance centralized taxation authority. However, efforts to redraw state boundaries and reconfigure German federalism in a more profound way were defeated.
The Bundesrat thus survived a transition that had transferred power to a national political elite previously committed to its elimination. The reason has in large part to do with the fact that federalist thinking had long infiltrated the party itself, and even radical socialists sought refuge in federal institutions (in this case, paradoxically, to buffer themselves from the moderates in power at the national level). For its part, the beleaguered national government saw the newly founded democratic state governments as allies in its efforts to institutionalize democratic rule. And, as during the imperial period, radically diverging political constellations in different states during the rest of the Weimar years kept the discussion of deeper reforms to the Bundesrat and to Germany's federal structure off the agenda.

Reinstituting Federalism: the Bundesrat in the Federal Republic
Federal institutions were completely dismantled in the years of National Socialism, but reemerged after the Second World War. When the Federal Republic of Germany was founded out of the three zones occupied by France, the United Kingdom, and the United States, the allies placed certain limits on the nature of the constitution, outlined in the so called "Frankfurt Documents." These were to be the basis of the constitution (named the "Basic Law" or Grundgesetz for its supposed provisional nature) crafted by the Parliamentary Council convening in Bonn. The allies were unambiguous in demanding that Germany be reconstituted as a federal state but as Scharpf and Lehmbruch both emphasize, they confronted powerful actors in Germany who preferred a more centralized constellation (Lehmbruch 2002: 59;Scharpf 2009: 18f). 12 Since a completely unitary system was ruled out by the occupying powers, contests focused on more versus less centralized models.
Some of the most important conflicts were played out over the composition and role of the Bundesrat. Regionally based parties and parts of Christian Democratic Union (CDU), as well as politicians from the southern states were especially interested in preserving the traditional Bundesrat model (Lehmbruch 2002: 7, 59;Lambrecht 1975: 28). By contrast, the Social 12 Although a full treatment is beyond the scope of this chapter, it is worth mentioning that the allies also insisted on a rather decentralized system of dual taxation in which indirect taxes would go to the national government and direct taxes to the states. In order to avoid conflicts, the constitution of 1949 included only preliminary rules on fiscal federalism. Elsewhere, (Broschek 2010(Broschek , 2011Renzsch 1991) have analyzed the ways in which incremental expansion of vertical revenue sharing and "horizontal" (inter-state) equalization schemes evolved step by step between 1955 and 1969, at which point a major overhaul of the tax system explicitly reestablished a system which, by requiring joint decisions in legislation, increased the power of the Bundesrat.
Democratic Party and the liberal Free Democratic Party preferred a popularly elected senate like in the United States. They were joined in this position by a significant fraction in the CDU, including its leader Konrad Adenauer, as well as the country's future first president Theodor Heuss (Morsey 1974: 67;also Morsey 2000;. In the end the Bundesrat system (of state-government delegates) prevailed, a decision widely seen as a victory of federalist forces (Neunreither 1959;Lambrecht 1975;Merkl 1965;Lehmbruch 2002: 63). The conservative state premier of Bavaria, Hans Ehard, who was deeply involved in the deliberations, declared the reinstitution of the Bundesrat "a decisive victory for the federalist spirit" (cited in Morsey 1974: 77). 13 After the extreme centralization of power under Hitler the proponents of the Bundesrat model saw the institution as central to the project of "stabiliz[ing] the rights of the states against federal powers" (Neunreither 1959: 729).
The revival of the Bundesrat model was anything but foreordained. With the exception of Bavaria, Hamburg and Bremen, 14 the German states did not reemerge after the war in their historic form. Most of the states that would send delegates to the upper chamber were wholly new constructions, their boundaries above all a function of the occupation zones. However, and as in the early Weimar years, state governments had been able to reconstitute themselves before the national government was up and running. As in the previous transition, Prime Ministers of the Länder convened a Parliamentary Council through which they could collectively exert influence in discussions over constitutional redesign --another critical antecedent à la Slater and Simmons (Slater and Simmons 2010). 15 As Karl Arnold, Prime Minister of North-Rhine Westphalia and Bundesrat with the votes of a majority of its members, except where the Bundesrat objects with a 2/3 majority, in which case the Bundestag needs a 2/3 majority to overrule the objection (Art. 77, 1-4). the first president of the Bundesrat argued in his inaugural address: "The Länder, represented by their governments, are powers within the Federal Republic of Germany that cannot be ignored.
They were the first to re-emerge after the total defeat" (cited in Rapp 1974: 23).
In principle, and as before, the national leadership of the SPD was less enamored of strong state rights. However, for them the Bundesrat model was part of a compromise that would clearly subordinate this institution to the popularly elected Bundestag in the legislative process, something that would have been hard to attain with the Senate model (Morsey 1974: 74f). The reinstitution of the Bundesrat was thus in an important sense a "common carrier" of diverse and even conflicting objects (Schickler 2001). States' rights advocates believed the institution would serve as a bulwark to preserve state autonomy, and unitarists saw in it a less formidable competitor for national legislative power than a popularly elected Senate. The following section explores why both sides were wrong.

Limits to Stability: The Changing Nature of Post-War German Federalism
The role played by the Bundesrat in the Federal Republic would have disappointed its federalist and unitarist supporters in about equal measure, for over time it developed into an institution that, contrary to the federalists, has not preserved the autonomy of the individual state legislatures, and, contrary to the unitarists, has at the same time developed into a (sometimes disruptive) veto player in national-decision making. Oeter (1998: 264-65) characterizes the transformation of the Bundesrat as follows: "originally intended by its proponents as a 'federal brake' [on central power] along the lines of the Bundesrat in Bismarck's constitution…the Bundesrat in fact developed into a promoter of unitarization" (Oeter 1998: 264-65).
The division of powers between central government and states that was worked out in the Bonn constitution was from the start somewhat ambiguous, as indeed the term "unitary federal state" (Hesse 1962) itself implies. As noted above, the state governments were in place before the founding of the Federal Republic and the allies "insisted on a federal solution" and "also made sure that the Länder would dominate the design and adoption of the constitution" (Scharpf 2005).
At the same time, however, the idea of a unitary state had taken hold in the population in general and in the national political parties in particular, so that as Fritz Scharpf notes, the Basic Law that was ultimately ratified in 1949 was also infused with a strong strain of unitarism.
In some ways tensions were built into the constitution itself. Article 72 set out the basic terms of the relationship between central and state legislative jurisdictions-concurrent rights, reserved competencies, and conditions under which federal legislation trumped state rights -though the general trend in the early years was toward interpreting these in ways that enhanced federal dominance. However, the constitution also clearly gave states primary responsibility for administering federal law, building on Germany's strong tradition of decentralized administration (Neunreither 1959: 713-14). In light of this, Article 84 (1) of the German Basic Law stipulated that federal laws that require execution and administration through the states must secure their consent via the Bundesrat. This provision, originally seen as innocuous, in fact opened the door for what became a "surprising expansion" of the Bundesrat's involvement in legislative matters (Neunreither 1959: 718). 16 The increase in the Bundesrat's role and influence in national decision making was unanticipated, and in its aggregate effect, unwanted by all sides as well. The change was not abrupt but proceeded gradually as more and more legislation came to be interpreted as requiring Bundesrat approval. While initially there was some ambiguity as to whether Bundesrat approval was required only for the administrative sections of proposed laws, political developments and expansive interpretations of the law by the Constitutional Court drove an increase in the number of laws requiring Bundesrat assent. Figure 2 documents the growth of laws subject to Bundesrat approval after 1949.

Figure 2 about here
As the influence of the Bundesrat in national politics grew, the autonomous legislative powers of the state legislatures declined, or as Oeter puts it, "the states' gradually diminishing independent powers were replaced with increasing participation [through the Bundesrat] in the exercise of power at the national level" (Oeter 1998: 265). The paradox is that the Bundesrat, established to safeguard state autonomy and widely conceived as a victory for federalists, was in this way effectively transformed from a "states rights" chamber into a vehicle for the unitarization of power in Bonn/Berlin. The result, as one observer put it, is that today's Bundesrat is a far cry from the one envisioned by the founders (Neunreither 1959: 713).

The 'Double Movement' of German Federalism
The evolution of German federalism and, with that, of the role and functions of the Bundesrat involved a double movement -an increase in the power of the federal government over state governments in legislative decision-making, and a concomitant increase in the role and power of the Bundesrat in national-level politics. This section deals with each of these aspects in turn.
Judicial actions and political dynamics both operated from early on to reinforce the dominance of the central government over the states in legislative decision making. An important court decision (in 1952), for example, underscored the impact of Article 125 of the Basic Law, which stipulated that all matters falling under concurrent legislation that had already been pre-empted by the federal level before 1949 would automatically be carried over into the Basic Law as federal competences. The court ruled that these matters would not be subject to state approval, in effect generating a barrier against decentralization in any of these areas (Oeter 1998: 202-206;Scharpf 2009: 9).
Legal rulings and political practice under Art 72,2 of the constitution also reinforced the general thrust toward federal dominance. This provision gave the central government the right to pass laws "if and to the extent that the establishment of equal living conditions throughout the federal territory or the maintenance of legal or economic unity renders federal regulation necessary in the national interest." The original intent of this clause (the so-called necessity clause or Bedürfnisklausel) under the occupation powers was to preserve significant residual powers for the states -i.e., the federal government could only take action required to equalize conditions. But in practice the clause was interpreted broadly to allow federal legislation in a wide range of areas.
In the 1950s and 1960s, state governments readily acceded to federal actions taken under the auspices of the "equalization" clause because economic conditions (above all, labor shortages) were driving costly competition between states on subsidies and public employee wages (Scharpf 1999). In this context, states were willing to yield regulatory power to the federal government as a way of solving their own collective action problems. A similar dynamic took hold when the German federal government began to offer grants-in-aid to states in issue areas falling under their (states') constitutional prerogative, such as education and health care. Frequently such grants came with some type of conditionality, e.g., that the states match federal funds. Nevertheless, states were in no position to turn down federal grants, even less so as their fiscal situation deteriorated in the first economic downturn of the late 1960s and early 1970s (Blair 1981: 224).
When the "Grand" coalition of CDU/CSU and SPD came to power in 1966 it undertook a major overhaul of federal arrangements. However, what transpired to a very significant extent simply followed and codified political realities as these had developed over the previous decades.
Longstanding areas of federal-state cooperation were formally written into the constitution as "joint tasks" (Art. 91), which are planned and financed jointly by state and federal governments.
The constitutional basis for grants-in-aid was established by Article 104a,4, and some of the fields where such financing had become prevalent (e.g., hospitals and education and training), were added to the list of concurrent powers (Art. 74. 13 and 19a) (Scharpf, Reissert, and Schnabel 1976;Lehmbruch 1976). In this sense, the strengthening of Bundesrat power in national decision making and of state administrative power were mutually reinforcing; Oeter even speaks of a dialectical relationship between the two (Hesse 1962;Oeter 1998: 265). The Bundesrat (and the state governments represented in it) thus retained significant power, even if state legislatures who did not have a voice in these proceedings, found themselves excluded from an increasing number of policy fields (Schmalenbach 1998: 44f). 17 As Behnke and Benz note, the trend since the 1980s has been in the other direction, with states trying to "decentralize powers and to push back federal interventions in affairs of Länder and local governments" (Behnke and Benz 2009: 224).
--- Figure 3 about here ---Thus, the other side of the double movement of German federalism was the expansion of the role of the Bundesrat. In this case too the trend was partly driven by political developments but further reinforced by key court decisions. Of these, the most significant ruling was a 1958 decision in response to several laws extending the federal price act. In this case the court ruled that Bundesrat approval was required in the law at issue, and it went on to render a very expansive interpretation of Article 84 (1) that held that it was not just the administrative provisions of a proposed law that required Bundesrat asset, but the entire act. The impact of this ruling, based on a "unity theory" [Einheitstheorie] of laws was not fully apparent until the 1970s when the Bundesrat and Bundestag were controlled by different political majorities. However, this decision translated into a significant increase of power of the Bundesrat, as "the administrative character of a single clause was sufficient to give the Bundesrat an absolute veto over the whole law" (Blair 1981: 96). 18

Party Politics and the Bundesrat
During the crafting of the constitution, proponents of both Senate and Bundesrat models suggested that the institution would be above party politics, a moderating influence on the more partisan Bundestag. However, as Lehmbruch has convincingly argued, these views rested on what turned out to be faulty assumptions both about partisan politics and the limited role the Bundesrat was expected to play (Lehmbruch 1976: 71). By 1961 the powers of the Bundesrat had grown beyond the intent of the designers and the German party system was increasingly characterized by a high degree of polarization between the CDU/CSU and the SPD. The specter of divided government -with Bundestag and Bundesrat controlled by different "poles" in the political spectrum became a distinct possibility.
The first period of "divided government" in 1969 under the SPD/FDP coalition had the effect of pulling the Bundesrat into national party politics, a problem that became recurrent and acutely visible thereafter (Lehmbruch 1976;Burkhart 2008). 19 Initially, the position of the small but pivotal FDP in several state coalitions with the CDU was able to moderate attempts by the CDU to use the Bundesrat as a tool for party politics. But as the CDU was able to win absolute majorities in an increasing number of states, it was able to make good on the announcement of CDU-leader Kiesinger that the Bundesrat would be "an important instrument for the opposition" (cited in Lehmbruch 1976: 133). The CDU/CSU's opposition did not take the form of explicit obstruction, but the opposition had a very strong position in negotiating favorable outcomes for laws requiring Bundesrat consent. 20 One consequence of these developments was that questions of states' rights were increasingly subordinated to or translated into partisan disputes and played out in conflicts between competing majorities in the Bundesrat and Bundestag. 21 Moreover, state prime ministers increasingly used campaigned with a slogan that emphasized the importance for the composition of the Bundesrat of the party's winning the state parliament. 20 In 1994, a constitutional amendment was passed requiring the Federal Constitutional Court to decide on matters of concurrent powers, and after 2000 a number of court rulings produced a somewhat more restrictive interpretation of the "necessity clause," thus restricting federal jurisdiction. But this did not change the basic dynamics described here. 21 Under Gerhard Schröder's Red/Green coalition, legislation that was opposed by a CDU-CSU dominated Bundesrat included health care reform, fiscal reform, welfare reform, and the reform of citizenship laws. the platform afforded them as members of the Bundesrat to position themselves as opposition leaders. This "tradition" was begun by Helmut Kohl, then Prime Minister in Rhineland Palatinate in 1975, when, addressing the Bundestag as a representative of the Bundesrat, he stated that he was "speaking for my friends in the German CDU/CSU" (cited in Lehmbruch 1976: 141). The role of the Bundesrat as a stage for national politics has been further emphasized by the importance of state party organizations for the recruitment of national politicians. Angela Merkel was not only the first Chancellor since Ludwig Erhard (1963)(1964)(1965)(1966), who had not been state prime minister, 22 but was one of only two major party candidates since Erhard without such experience.
The role of party politics, party organization, and divided government has thus led the Bundesrat further and further away from a "states rights" chamber. When Gerhard Schröder called for new elections in 2005, it was not on the basis of defeat in the lower house (Bundestag), but rather on the occasion of the CDU/CSU winning a 2/3 majority in the Bundesrat -thus guaranteeing the opposition the ability to block any law (not just those requiring assent) passed by the governing coalition. Efforts to ease the political deadlock caused by these dynamics through constitutional 22 Or the functional equivalent of ruling mayor of an independent city, such as Brandt, Schmidt, and Vogel, mayors

Conclusion: Stability and Change in Federal Institutions
The case of the Bundesrat is instructive for what it can teach us about processes of institutional change in federal institutions and beyond. The literature in comparative politics often falls back on a punctuated equilibrium model of change that expects significant transformations mostly in "critical juncture" moments and institutional stability in periods of "normal politics." What we observed for German federalism, however, is something like the opposite: surprising stability through unsettled times, and significant changes in periods of "normal politics" and even under the guise of considerable formal institutional stability.
As we saw, the Bundesrat proved surprisingly resilient through successive episodes of dramatic political change, being reinvented or revived in the aftermath of historic ruptures following the first and second world wars. Since that time, however, the Bundesrat's role in Germany's system of government has changed considerably and in ways that are deeply paradoxical -from a strong states' right chamber to a powerful and sometimes disruptive player in national decision making.
But what is crucial in the current context is that many of the most significant shifts in this transformation occurred outside of big break points through processes of incremental but cumulatively transformative change.
As noted in the introduction to this volume, most of the existing literature that takes federal institutions as the dependent variable focuses either on issues of constitutional design or of the stability of federal institutions. Regarding design, our study suggests that even the most dramatic instances of historical ruptures do not necessarily allow for textbook re-design; many features survive these ruptures. Moreover, the political role of the Bundesrat (Reichsrat) hardly corresponded to the expectations of its designers, which in turn leads us to be sceptical about inferring the effects of institutions from their formal design, or conversely, of reading the intent of an institution's designers off of the role the institution can later seen to be performing.
Analyses of the stability of federal institutions do implicitly address institutional change --as the absence of stability. But of course treating stability and change as polar opposites is clearly unsatisfactory. German federalism is indisputably stable on many important dimensions, and yet formal institutional stability has not prevented significant change in the functions of the Bundesrat over the last sixty years. Indeed, as Behnke and Benz suggest, a certain degree of flexibility is sometimes even a precondition for the stability of federal structures (Behnke and Benz 2009).
We have suggested studying the origins and development of federal institutions using insights from historical institutionalism. Following the work of Thelen (2004) and Pierson (2004), we emphasize the mechanisms of reproduction that keep institutions intact even through significant historical junctures as well as the importance of gradual changes in the form and functions of institutions over time. While until now the key works in historical institutionalism have concentrated on 'softer' institutions, mostly in the area of political economy (Hall 1986;Thelen 2004), our study shows that some of these insights can be fruitfully applied to formal political institutions as well. based on: Statistisches Bundesamt, Fachserie 14, R 3.1, 1997