Jump to ContentJump to Main Navigation
Parliaments in TimeThe Evolution of Legislative Democracy in Western Europe, 1866-2015$

Michael Koß

Print publication date: 2018

Print ISBN-13: 9780198766919

Published to Oxford Scholarship Online: December 2018

DOI: 10.1093/oso/9780198766919.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2019. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 17 September 2019

Introduction: Legislators in the Steel-Hard Casing

Introduction: Legislators in the Steel-Hard Casing

(p.1) 1 Introduction: Legislators in the Steel-Hard Casing
Parliaments in Time

Michael Koß

Oxford University Press

Abstract and Keywords

This chapter identifies the problem of legislative democracy. As a response to growing pressures to increase procedural efficiency in the wake of the Industrial Revolution and the advent of more inclusive suffrage formulae, legislators face two procedural alternatives: to centralize control over the legislative agenda, to create powerful committees. Talking legislatures combine centralized agenda control and weak committees, working ones decentralized agenda control and powerful committees, and hybrid ones centralized agenda control and powerful committees. According to the dynamic partisan perspective adopted in this book, a centralization of agenda control only occurs as a response to anti-system obstruction. Given legislators’ demand for mega-seats, the creation of powerful committees is the default way to rationalize legislative procedures. If, however, legislators fail to procedurally respond to anti-system obstruction they risk a breakdown of legislative procedures. This is why this book ultimately focuses on legislative democracy rather than legislative organization.

Keywords:   Legislative democracy, agenda control, committee power, anti-system parties, legislative obstruction, legislative mega-seats, talking legislatures, working legislatures, hybrid legislatures

On 20 February 1882, the Liberal British Prime Minister William Ewart Gladstone proposed that a majority of the House of Commons should be able to terminate legislative debates upon the initiative of the Speaker. At the time, Gladstone’s proposal met fierce resistance by both the official Conservative opposition and the ‘third party’, the Irish Nationalists. In defending the proposal, the Marquess of Hartington, Secretary of State for India, insisted that the fate of British democracy depended upon the introduction of a closure procedure, the technical term for the envisaged reform:

The Government are responsible to the House and to the country for the conduct of Business. They come before you to tell you that under the existing Rules of Procedure in this House they cannot undertake responsibility. If they cannot conduct the Business of the country, they are not fit to remain in Office. They have laid before you proposals which, in their opinion, will enable them, and other Governments after them, to conduct with dignity and efficiency the Business of the country, and by these proposals they are prepared to stand. If there are others who think they can, without these changes, conduct the necessary affairs of the country, and if they can persuade the House of their capability to do so, we shall cheerfully resign our functions; but, Sir, so long as we are responsible for the conduct of the necessary Business of the country, we must appeal to the House to give us those powers by which alone, as we think, our work can be effectually performed.

(HC Deb 20 March 1882, cols. 1337–8)

Hartington’s appeal to the House of Commons and the public at large contrasts sharply with the far more parochial complaint that Albin Ström, a rank-and-file Social Democratic legislator, raised about a procedural reform concerning the committee system in the Second Chamber of the Swedish Riksdag in 1932. Ström complained about the ‘joblessness’ of many of his fellow legislators. He argued:

The real work here in the Riksdag is done in committees. In most cases, open questions are settled there, and arguments in the chamber, like the one I am advancing just now, are rarely or never taken into account. Furthermore, a corresponding custom has developed according to which, (p.2) in the event of a plenary debate on a committee proposal, one should be a committee member or at least a deputy member, or proposer of an amendment, or party leader to dare to raise one’s voice. Other legislators have nothing to do with such committee reports. This custom is strictly followed, with the consequence that the more senior legislators who uphold and maintain this committee tyranny show a very angry face to those who break with this custom.

(AK 25 May 1932, 42)1

Hartington and Ström called for different reforms, both ideal-typical, which encompassed the full variety of options available to render legislative procedures in elected assemblies more efficient. On the one hand, Hartington focused on the chamber where responsible governments should be procedurally enabled to implement ‘their’ legislative programme. On the other hand, Ström emphasized committees as the major arena where all legislators should be allowed to scrutinize proposals independent of government or opposition status. This full procedural variety is also reflected by the reform debates during which Hartington and Ström spoke. In the House of Commons, the closure proposal triggered nineteen debates in which 217 speeches were held by 112 legislators (17 per cent of all). In the Riksdag’s Second Chamber, there was only one debate on the envisaged reorganization of the committee system in which 18 legislators held 32 speeches (7 per cent of all). With respect to the preparatory work informing the plenary debates, both legislatures differed even more starkly. In the House of Commons, the nineteen debates were informed by a single 128-word proposal tabled by the Prime Minister (see HC Deb 20 February 1882, col. 1124). In the Riksdag’s Second Chamber, the single debate was prepared by a 201-page commission of inquiry report (SOU 1936 no. 21), a 70-page government proposal (Prop 1932 no. 105), and a 93-page committee report (KU 1932 no. 26).

The House of Commons and the Riksdag demonstrate fundamentally different legislative procedural behaviour, emphasizing the two institutional mechanisms on which a large share of scholarship on legislative organization focuses: control over the plenary agenda and the power of legislative committees (e.g. Cox 1987; Krehbiel 1991; Huber 1996a; Martin and Vanberg 2011). In this respect, the British House of Commons and the Swedish Riksdag represent the ideal types of ‘talking’ and ‘working’ legislatures (see Steffani 1979, 96–106). The focus in talking legislatures is on plenary proceedings and the legislative agenda is controlled by the government majority. This implies that the committee stage of legislation is of secondary importance. In contrast, committees are the major arena of working legislatures where plenary decisions on legislative proposals are anticipated. This implies that there is no need to privilege government majorities in the chamber. The following chapters aim (p.3) to discover the conditions under which legislative procedure evolves towards the ideal type of talking or working legislatures.

Yet, a third alternative to talking and working legislatures exists: the breakdown of legislative procedures. This option was implicitly referred to by Hartington when he argued that only the introduction of closure would ensure the continued passage of legislation. His argument was proven right in the long run. The experiences of many European legislatures during the interwar years of the twentieth century are examples of the dangers associated with a breakdown of (legislative) democracy (see Luebbert 1991) which continues to this very day. Throughout this book, the term ‘legislative democracy’ is therefore preferred to the more commonly used term ‘legislative organization’ (see e.g. Carey 2006). Legislative democracy entails that all legislation is subject to consent by elected assemblies. This notion better accounts for what is at stake: the maintenance of a fundamental feature of democracy. Legislative democracy is just as important as electoral (ensuring the popular selection of legislators and—directly or indirectly—cabinet members) and liberal democracy (guaranteeing fundamental civil liberties to all citizens) (Rueschemeyer, Huber Stephens, and Stephens 1992, 10; Berins Collier 1999, 24). Against this background, the aim of this book is to examine not only why legislatures develop towards the talking or working ideal type, but also how legislative democracy is maintained—and under which conditions it fails.

1.1 Anti-System Obstruction and Procedural Rationalization

The core argument of this book can be summarized as follows. From the late nineteenth century onwards, like other modern individuals, legislators found themselves in a Weberian steel-hard casing2 which forced them to rationalize their activities due to the unprecedented expansion of their legislative workload. All else being equal, this triggered a development towards the working legislature ideal-type. Powerful committees offer legislators important offices which entail particular legislative powers. In this respect, working legislatures provide mega-seats which are desirable from both a career and a policy perspective (Carroll, Cox, and Pachón 2006; Martin 2014a). As we have seen, Ström was calling for the extension of such mega-seats in the Swedish Riksdag. In contrast, talking legislatures provide such legislative mega-seats primarily on one single, albeit very powerful, committee: the cabinet. For this (p.4) reason talking legislatures are less attractive, especially for rank-and-file legislators, and therefore remain an exception whose emergence is triggered by extraordinary circumstances: most notably, a substantial threat to legislative democracy by means of anti-system obstruction.

Only when anti-system parties opposing democracy as such or the political system they operate in (Capoccia 2002) aim to prevent the passage of legislation through the pursuit of obstruction (Bücker 1989) does it become feasible that a majority of legislators will forfeit their ancient procedural privileges embodied in powerful normative notions such as the ‘grand inquest of the nation’, ‘assembly rule’, ‘freedom of speech’, and ‘governmental consensualism’. In these cases, government majorities receive procedural powers like those Hartington called for in the House of Commons. As a consequence, committees remain weak or even absent. However, the presence of anti-system obstruction does not automatically cause talking legislatures to emerge. But if anti-system obstruction is not defused by granting the majority privileges in plenary proceedings (i.e. by a development towards a talking legislature), this triggers the breakdown of legislative democracy. If no such anti-system obstruction occurs, legislators prefer to empower committees for two reasons: first, because they are unwilling to unilaterally forfeit their ancient procedural privileges; second, because the alternative procedural path towards talking legislatures provides them with fewer mega-seats. This is how working legislatures emerge.

The quotations at the outset of this introduction also illustrate that the major actors in the procedural development of legislatures are not political parties, but two groups of legislators differing in their relation to legislative mega-seats on committees and/or the cabinet. On the one hand, there are those already holding mega-seats, like Hartington (who will henceforth be called parliamentary ‘leaders’) and, on the other hand, there are those desiring to obtain mega-seats, like Ström (who will henceforth be called parliamentary ‘followers’). The notions of ‘leaders’ and ‘followers’ are not necessarily linked to the existence of political parties which, as the dynamic partisan perspective adopted here suggests, co-evolve with legislative procedures (see also Rohde and Shepsle 1987). However, even at a pre-party stage, there are charismatic, ambitious, and skilful leaders who use their personal influence to assemble legislative majorities, and then there are the potential members (i.e. followers) of these majorities who must be persuaded to join the voting bloc. As we shall see, once parties and party groups emerge, these serve as gatekeepers of persistence which regulate access to mega-seats rather than being agents of procedural change. Consequently, the presence of party groups helps to explain the self-reinforcing path-dependent development of talking and working legislatures (see Mahoney 2000; Rixen and Viola 2014).

Taking seriously the historical commonplace that legislative procedures and political parties co-evolve implies that party system features such as (p.5) the effective number of parties (Carroll, Cox, and Pachón 2006), their polarization (Diermeier and Vlaicu 2011), or seat shares (Binder 1997, Dion 1997) can hardly unilaterally shape the evolution of legislative democracy. Likewise, the creation of powerful committees is not necessarily causally related to patterns of government formation (but see André, Depauw, and Martin 2016). Empirically, powerful committees emerge both before and after the advent of parliamentary democracy and, by implication, coalition governments. Finally, the dynamic partisan perspective adopted here suggests that institutional veto points available to parties such as bicameral arrangements, proportional or majoritarian electoral systems, or constitutional courts (McCubbins 2005; McGann 2006; Taylor 2006) hardly account for the development towards talking or working legislatures because they shape parties as much as they are shaped by them (see Colomer 2005).

1.2 The Case for a Longitudinal Analysis

Empirically, the ‘legislative state of nature’ (Cox 2006, 143), giving equal floor access for all legislators and with weak committees, proved surprisingly resilient. This is also illustrated by the two procedural reforms mentioned at the outset of this chapter. For all their differences, they were similar in one respect: both were only successful in the long run. The closure was eventually passed by the House of Commons in October 1882, but only applied twice afterwards. This rendered it necessary to pass it again in 1887. Only since 1902 have British governments fully dominated plenary proceedings as Hartington had originally envisaged. Similarly, a committee system providing all legislators with a seat on a permanent committee was only adopted by the Swedish Riksdag in 1970, thirty-eight years after Ström’s complaint in the Second Chamber. These examples already suggest that an analysis of the evolution of legislative democracy requires a long-term perspective. But when exactly does such a longitudinal analysis have to start? This question is of particular relevance since all legislatures originated as ‘assembl[ies] of notables of equal status’ (Loewenberg 2011, 14). Just as representative institutions were originally not designed to be elected (Manin 1997), they were not to be dominated by members with procedural privileges, such as the holders of mega-seats on the cabinet or committees. Rather, all over Europe parliamentarism emerged out of assemblies representing estates rather than individuals (Hintze 1930; 1931). In this sense, pre-modern representative assemblies were conceived of as court-like bodies whose consent was required for taxation in the context of European feudalism. Against this background, not only the specific development towards talking or working legislatures warrants an explanation, but also the departure from the legislative state of nature in general.

(p.6) The transition from assemblies to legislatures became feasible once these bodies were exposed to two pressures, one stemming from the Industrial Revolution which led to a proliferation of new policy areas—most notably economic and social policies—warranting regulation (i.e. legislation), the other deriving from increasingly inclusive suffrage formulas served as an incentive for legislators to demonstrate their commitment towards growing electorates through increased legislative activity. By the final third of the nineteenth century, growing economic progress and electoral responsibility simultaneously increased the number of legislative proposals and reduced the time available for their discussion on the floor. This transformed inquest-like assemblies, which previously were largely occupied with the discussion of grievances, into proper legislatures (Hintze 1930; 1931). In Western Europe, it was after 1866 that legislators found themselves in a steel-hard casing. Max Weber had originally used this notion to describe the impact of rationalization—which he understood as the interplay of systematization, specialization, individualization, disciplining, and bureaucratization—on modern individuals (Weber [1905] 1988, 224). Weber argued that ‘the waste of time is the primary and gravest of all sins’ in the steel-hard casing (Weber [1905] 1988, 167). It is exactly the ensuing pressure towards a more efficient use of time which allowed for the ‘paradox of hierarchy’ (Loewenberg 2007, 59) to emerge among formerly equal members of legislative assemblies. Another aim of this book, therefore, is to simultaneously study (emerging) parliaments in time through an examination of different distributions of legislative time.

Procedural rationalization of legislatures meant to relax the pressure of the steel-hard casing is by no means merely a historical phenomenon. Two examples will suffice here. First, the basic allegation of adherents of ‘parliamentary decline’ arguments remains unchanged since its first appearance (see Bryce 1921, 335–44; Schmitt [1926] 1996): legislative democracy runs the risk of becoming overstretched because, due to society’s ever-growing complexity, the sheer number of issues warranting regulation prevents meaningful legislative deliberation (e.g. Steffani 1979, 162; Benz 1998, 203; Döring 1996, 42). Contemporary macro trends such as globalization and Europeanization allegedly even aggravate this allegation because ‘post-parliamentary governance’ can be used to bypass national legislatures altogether (Andersen and Burns 1996). Second, recent developments in Kosovo, Macedonia, and most notably Turkey illustrate that the breakdown of legislative democracy continues to this day.3 What is more, legislative democracy becomes endangered (p.7) in established democracies, too. In its 2017 annual report on freedom in the world, the non-governmental organization Freedom House warned of the dual threat posed by populists and autocrats which leads to increasingly ‘fragmented parliaments with no governing majority’, especially in democracies such as Spain, Germany, Britain, France, and the US (Puddington and Roylance 2017, 8).

1.3 Why Legislative Democracy? and Why Western Europe?

Given the past and present threats to legislative democracy, this book aims to trace its evolution in Western Europe between 1866 and 2015—that is, from the emergence of the steel-hard casing of procedural rationalization up to the present. A focus on legislation allows for systematic comparisons over time and space. Furthermore, this perspective is of primordial importance since public resources of political systems are usually distributed by means of legislation (see Heller, Kyriacou, and Roca-Sagalés 2016, 684) which is better protected against judicial review than executive decrees (Saiegh 2011, 5). The universe of cases encompasses all Western European national legislatures (i.e. elected assemblies whose consent is required for the passage of legislation) existing throughout (the better part of) the 1866–2015 period. Full electoral (or even liberal) democracy is not necessarily required. The focus of this book will be on emerging parliamentary democracies in Western Europe. Such a focus is warranted for several reasons. First, it was in Europe that democratization paralleled the emergence of Weber’s steel-hard casing of rationalization. Second, the first wave of democratization in Europe is unprecedented in scope and scale (Ertman 1997). Third, parliamentarism has its roots in Western Europe and derived from the decision-making institutions originally developed under feudalism (Hintze 1930; 1931). Finally, too little theory-building in legislative studies has so far focused on legislatures beyond those in presidential systems in general and the United States Congress in particular (Dion 2011, 751–2; Saiegh 2014, 500). It is such theory-building (and testing) the present book focuses on.

In order to allow for a variety of controlled comparisons of (emerging) parliaments in time, this book will primarily focus on the lower chambers of four countries which are most different with respect to the fragmentation of (p.8) parliamentary party systems and the distribution of procedural privileges: Great Britain, France, Sweden, and Germany. The ultimate aim of the comparative qualitative case studies is to come up with falsifiable ‘master variables’ (Luebbert 1991, 5) explaining the evolution of legislative democracy. More specifically, the analysis focuses on ninety cases encompassing all procedural reforms of agenda control and committee power discussed on the parliamentary floor independent of their success between 1866 and 2015. Negative cases of rejected or only partially adopted reforms were included to minimize selection bias (see Capoccia and Ziblatt 2010, 943–4).

Even though ‘[t]he study of institutional origins has become one of the most vibrant research areas in comparative politics’ (Rodden 2009, 335), the recent historical turn in democratization studies (Capoccia and Ziblatt 2010) has not yet addressed legislatures. Surprisingly, the evolution of legislative democracy has only found limited scholarly attention, at least from a comparative perspective. The only systematic long-term comparison of European legislatures is provided by Sieberer et al. (2016), whose analysis, however, only begins in 1945. Congleton (2011) provides a truly longitudinal analysis of several European legislatures going back to medieval times, but focuses on government-executive relations and not legislative procedure. Single-case, long-term analyses exist primarily for the Congress of the United States and the British House of Commons (Martin, Saalfeld, and Strøm, 2014, 10–11).4 In this respect, the existing research on legislatures contrasts starkly with the large comparative literatures on electoral democracy, most notably the origins of electoral systems (Ahmed 2010; 2013; Boix 1999; Colomer 2005; Leeman and Mares 2014) and the secret ballot (Kam 2017; Teorell, Ziblatt, and Lehoucq 2017).

The relative neglect of legislative democracy also comes as a surprise given the well-established consequences of different forms of legislative organization. Apart from political parties, procedural rules—which can be codified as mere standing orders, formal laws, and even constitutional rules—are regarded as the most important structuring variables shaping political decisions in Europe (Sieberer, Müller, and Heller 2011, 951). Accordingly, decisions about such procedural rules directly affect public policies. The impact of government privileges—such as the ability to close debates, referred to at the outset of this introduction—in the legislative process illustrates the relation between procedural rules and policies: more legislative privileges lead to fewer, but longer, bills being passed (Döring 2001), fewer amendments being made prior to passage (Döring 1995b), and less time being spent on debating these bills in plenary. Furthermore, such government privileges provide incentives for (p.9) government (Huber 1996b; Heller 2001) and opposition cohesion (Dewan and Spirling 2011).

Against this background, it comes as no surprise that legislatures in which governments possess procedural privileges are often regarded as particular examples of declining legislatures: in the most sophisticated ranking of parliamentary powers, the French and British legislatures have particularly low scores as the only Western European legislatures without the ability to impact policy-making or select cabinet members (Sieberer 2011, 746). Given the emphasis of the scholarly literature on the effects of causes (such as the presence or absence of procedural government privileges), this book aims to uncover the causes of these effects in order to reassess their validity.

1.4 A Dynamic Partisan Perspective on Talking and Working Legislatures

Having defined the problem and relevance of legislative democracy, it is now time to further specify the distinction between talking and working legislatures. The following analysis of the evolution of legislative democracy between 1866 and 2015 focuses on two procedural features: agenda control and committee power. Once the pressure towards a rationalization of the legislative process begins to influence legislators, they face two procedural options in order to cope with the growing demand for legislation (see also Döring 1996, 43): to centralize control over the plenary agenda or to create powerful legislative committees. Agenda control refers to the whole process of passing legislation. Three stages are of particular importance (Döring 1995a; Rasch 2014): the ability to decide which legislative proposals are discussed in a legislature (timetable control), the power to amend these proposals (positive agenda control) and to terminate debates, moving on to decision-making (negative agenda control). Governments able to unilaterally select the legislative proposals to be discussed, to offer the last amendments to them, or to terminate legislative debates, possess such procedural privileges. In this case, control over the legislative agenda is centralized. This decreases the pressure of the steel-hard casing since it vertically differentiates the legislatures by privileging governments or majorities. Weak or absent legislative committees are the procedural corollary of centralized agenda control; the government prerogatives over the plenary stage most likely suffice to override unwanted committee revisions to legislation. This reduces incentives for committee empowerment since committees are hardly regarded as providing desirable mega-seats. Rather, the cabinet can be conceived of as a single powerful executive committee (Bagehot [1867] 2001, 11; Weber [1919/20] 2014, 211). (p.10) The focus of legislators’ activity is on talking during the plenary stage (including the option to tactically obstruct the passage of legislation), which is why this institutional design resembles a talking legislature.

The alternative response to the pressure to rationalize legislative procedure is the creation of powerful committees dealing with legislation. This option increases efficiency through a horizontal differentiation of the plenary arena. Committees are powerful if they are able to alter proposals sent to them and to gather information (Martin and Vanberg 2011, 35; André, Depauw, and Martin 2016, 111). For this reason, powerful committees are permanent and able to rewrite bills. Additionally, the structure of powerful committees is congruent with the executive departments. Powerful committees are the procedural corollary of decentralized agenda control: only if government majorities lack privileged access to the plenary agenda can committees effectively impact legislation and exert their influence. If powerful committees exist, legislative decisions are effectively taken in the committee stage and the plenary becomes the arena for their ratification. This institutional design, in which legislators emphasize committee work, therefore resembles a working legislature.

To explain the emergence of talking legislatures (combining centralized agenda control and weak legislative committees) and working legislatures (combining decentralized agenda control and powerful legislative committees), this book relies on historical and rational choice institutionalism. Its aim is to ‘read history forward’ (Capoccia and Ziblatt 2010, 939) in the context of case studies which allow for controlled cross-case comparisons (see Møller 2013). These case studies are informed with hypotheses derived from a model based on the assumptions of distributive bargaining (Knight 1992). The key dowry from historical institutionalism is the emphasis on qualitative process-tracing analysis for developing a causal mechanism (Beach and Pedersen 2013) capable of explaining legislative democracy’s evolution.

According to the dynamic partisan perspective adopted here, procedural rule change can be conceived of as a nested game (Tsebelis 1990) between parliamentary leaders and followers. The former already possess mega-seats, be they on the assembly’s steering body, legislative committees, or chairs of (rudimentary) parliamentary groups. In contrast, followers also sit in the legislature, but not on mega-seats. However, followers also desire to obtain such mega-seats, be it by substituting for current leaders, or through the creation of additional positions. This perspective suggests that followers are generally more likely to empower committees than to centralize agenda control. Conversely, leaders might well prefer to create just one powerful committee in which legislative power is concentrated: the cabinet. However, there are by definition more followers than leaders. Therefore, followers prefer the creation of a multitude of legislative committees, which provide them with more mega-seats, to only one executive committee, where all (p.11) mega-seats are consolidated (independent of the fact that mega-seats on the cabinet are more powerful than on legislative committees).

The longitudinal analysis carried out here suggests an explanation for the fact that the outcome of procedural reforms more often than not reflected followers’ preferences: political parties and legislative procedures co-evolve. Despite the fact that powerful parties do not yet exist and that majorities have not gained institutional privileges, legislators find themselves in the steel-hard casing in the years following 1866. This suggests that followers are in a better bargaining position than leaders. Parliamentary leaders cannot rely on partisan resources to force fellow legislators to forfeit the procedural privileges they possess, since all legislators are equals in the legislative state of nature. In contrast, followers can threaten to withdraw support from leaders and to align with other leaders promising to create more mega-seats on powerful committees which effectively differentiate the plenary arena horizontally.

The longitudinal comparison makes it possible to view the emergence of dynamic partisanship as a key explanatory factor for understanding the development of different types of legislatures. Legislators do align with parties or party labels, but these are not yet stable over time. Parties emerge as the dominant actors in the process of procedural rationalization, but whether this occurs in talking or working legislatures does not depend on the features of party systems co-evolving with legislative procedures but on the occurrence of anti-system obstruction. Talking legislatures only become a feasible option in the contingent event of anti-system obstruction, which has the potential to alter the balance of power between leaders and followers.

Anti-system obstruction can be conceived of as a contingent shock to legislative procedure. What renders anti-system obstruction so relevant from a causal perspective is its fundamental challenge to legislative democracy. Anti-system parties differ from other parties because they reject the political system they operate in, be it because they reject democracy as such or because of irredentism, the goal of creating their own polity (Capoccia 2002). This negative attitude provides anti-system parties with a sort of cohesion other parties lack, especially in historical perspective. Even more importantly, anti-system parties have an incentive to systematically obstruct legislative business, ultimately aiming to bring all legislative business to a standstill (Sartori [1976] 2005, 118). Anti-system obstruction is necessary (but not individually sufficient) to cause followers to give up their inherited procedural privileges, most notably equal access to the plenary agenda. Under the condition of anti-system obstruction, parliamentary leaders frame procedural reform as a choice between the persistence of legislative democracy under decentralized agenda control or its collapse. The Irish Nationalists in the House of Commons (1902) and the Communists in the French National Assembly (1958) are two empirical examples of anti-system parties whose obstruction gave rise to critical junctures which led to the centralization of agenda control. If, however, (p.12) legislators are not willing to forfeit equal access to the agenda despite an anti-system threat, legislative democracy breaks down. This was the case in the German Reichstag in 1930 and, to a lesser extent, in the French Chamber of Deputies in 1937.

In the absence of any anti-system threat, committees are empowered and the agenda remains decentralized. This occurred in the French Third Republic after 1876, in the Swedish Riksdag after 1866, and in the German Bundestag after 1949. Committees may develop into necessary monitoring devices under coalition governments (Martin and Vanberg 2011; Carroll and Cox 2012). However, similarly to the centralization of agenda control, the major reason for the empowerment of committees—followers’ demand for greater participation in legislative procedures—is not related to party system features. Once working and talking legislatures emerge, specific behavioural-institutional equilibria (Colomer 2005) appear which are maintained by political parties. The different constellations underlying substantial procedural decisions (taken in intra-party mode) and more incremental ones (taken in inter-party mode) suggests that in the case of legislative democracy, institutional choice is decoupled from institutional development (see Pierson 2004).

Both in talking and working legislatures, legislators develop a preference for talk (in the plenary) and work (on committees) respectively. Especially for members of the opposition, the former preference implies also the option of tactical obstruction directed against particular proposals (and not legislation as such). Such tactical obstruction explains the emergence of hybrid legislatures which combine features of talking and working legislatures. Tactical obstruction plays virtually no role in working legislatures because here the focus is on legislative scrutiny, most notably in committees. In talking legislatures, however, even loyal opposition legislators adopt tactical obstruction. If tactical obstruction becomes too intense under centralized agenda control, reforms are taken in the opposition mode in which majority and minority confront each other (King 1976, 17–18). In this case, leaders of the majority call for an empowerment of committees. Their rationale is to provide opposition legislators with an alternative to tactical obstruction: participation in powerful committees. However, as long as the agenda remains centralized this is not an attractive option for the opposition, since a government controlling the agenda is able to reverse any committee changes on the floor. The opposition therefore rejects such a procedural reform whose outcome depends on the cohesion of the majority. When this cohesion is provided, a hybrid legislature consisting of centralized agenda control and powerful committees emerges. This was the case in the French National Assembly in 2009.

The most important normative implication of the analysis carried out in this book is that that there is no evidence for a decline of legislatures. Scholars of legislative studies have come to similar conclusions, arguing that even in (p.13) talking legislatures, individual legislators are able to affect policies (e.g. Huber 1996a; Dewan and Spirling 2011). Against this background, the ongoing debate about the institutional powers of legislatures (e.g. Fish and Kroenig 2009; Sieberer 2011) misses the point. This debate starts from the ahistorical assumption that legislators face an unconstrained choice of procedural rules. They might do so in laboratory experiments, but not in the real world. If anti-system obstruction occurs, legislators do not face the choice between working or talking legislatures or any hybrid design, but between talking legislatures (i.e. a centralization of agenda control and disempowerment of committees) and the breakdown of legislative democracy. All else being equal, legislators do indeed prefer to create powerful committees rather than to centralize agenda control. However, they do not choose the latter alternative to suppress minorities on behalf of egoistic leaders or for other questionable reasons, but to contain anti-system threats. This suggests that talking legislatures like the British House of Commons and the French National Assembly do not indicate a decline of parliaments, but the resilience of legislative democracy.

1.5 The Evolution of Legislative Democracy in Western Europe: An Outlook

The book proceeds as follows. Chapter 2 discusses agenda control and committee power as the most important features of talking and working legislatures. Agenda control is regarded as centralized if governments possess privileges in two or more of its defining dimensions (timetable, positive, and negative control). Correspondingly, legislative committees are regarded as powerful if permanence and rewrite authority are present. Chapter 2 also discusses existing explanations for the distribution of agenda control and committee power. From this discussion follows the methodological decision to develop a model using qualitative process-tracing to employ controlled cross-case comparisons of four types of procedural reforms (path changes plus path-dependent, incremental, and failed reforms).

Chapter 3 presents the dynamic partisan perspective on procedural change adopted here. Leaders are expected to prefer to create mega-seats in the cabinet (and centralize agenda control), whereas followers prefer to create (powerful) legislative committees. Leaders’ and followers’ procedural preferences are operationalized by means of the framing of their respective reform proposals. Expectedly, leaders emphasize a majoritarian vision of legislative democracy, while followers espouse a proportional vision. All else being equal, followers enjoy a better bargaining position because the proportional vision is more in line with the legislative state of nature. Therefore, the (p.14) procedural path chosen hypothetically depends on the occurrence of anti-system obstruction which is able to alter followers’ preferences. Chapter 3 closes by discussing the temporal, substantial, and spatial boundaries of the cases selected here.

Chapter 4 identifies ninety procedural reforms from the British, French, Swedish, and German legislatures during the years 1866–2015. For all these reforms original data have been collected, mostly consisting of parliamentary records (committee reports and plenary proceedings). The details and sources of all reforms can be found in the Appendix. Remarkably, all legislatures started their procedural development in the legislative state of nature. Only nineteen of the ninety reforms are substantial (nine path changes and ten path-dependent reforms) while seventy-one are expressions of institutional conservatism (fifty-one incremental and twenty failed reforms). Chapter 4 also contextualizes these reforms by discussing the evolution of the legislative workload, extra-legislative institutions, and party systems in the four countries under investigation.

Chapters 5–7 constitute the main analysis and trace the ninety reform processes discussed in Chapter 4. The chapters are organized according to the nature of the anti-system parties present in the respective legislatures. Chapter 5 covers the period up to 1917, before the emergence of ideological anti-system parties. During this period only relational anti-system parties, those who aimed primarily to break away from the polity within which they operated (Capoccia 2002), existed. Chapter 6 analyses the 1918–90 period, during which time ideological anti-system communist and fascist parties which directly opposed democracy emerged. Chapter 7 focuses on the 1991–2015 period, which saw the diminished importance of the traditional anti-system parties of the left (due to the end of Communist rule in Europe) as well as the rise of new populist radical right-wing parties (see Mudde 2007). Each of the major chapters focuses on two paired comparisons.

Chapter 5 argues that a centralization of agenda control was successful in the British House of Commons, because systematic attempts of Irish nationalists to delay the passage of legislation amounted to anti-system obstruction, but failed in the German Reichstag. The German Social Democrats, despite being deemed as anti-system by conservatives, only obstructed legislative business tactically. Without anti-system obstruction, committees were empowered in both the French Chamber of Deputies and the Swedish Riksdag due to follower demand, even though democratic coalition and minority governments existed only in the former legislature.

Chapter 6 argues that the occurrence of anti-system obstruction also explains why a centralization of agenda control succeeded in the French National Assembly in 1958, but not during previous reform attempts or in the German Reichstag during the Weimar Republic. Irrespective of the presence of ideological anti-system parties in the French and German (p.15) legislatures since the 1920s, the absence of obstruction (in the French case) or legislators’ procedural responses to it (in the German case) prevented any substantial reform. Since legislators adapted to procedural path changes by developing a preference for talk (in the plenary) or work (in committees), respectively, neither the disappearance of the Irish anti-system threat in the House of Commons nor the emergence of a Communist party in the Swedish Riksdag’s Second Chamber affected the respective procedural paths chosen in the pre-1917 period.

Chapter 7 argues that an empowerment of committees under centralized agenda control occurred after 1991 in the French National Assembly as a response to the sharp rise of tactical obstruction by loyal opposition parties. Powerful committees were created to provide opposition legislators with a meaningful alternative to obstruction. With no similar increase of tactical obstruction in the British House of Commons, no substantial procedural reform occurred. In all talking and working legislatures, legislators stuck to their respective preferences for talk and work. This also explains why the appearance of a potential anti-system party in Sweden (the Sweden Democrats) had no procedural consequences.

Chapter 8 systematizes the findings of Chapters 5–7 and provides the causal mechanism explaining the emergence of talking, working, and hybrid legislatures. Apart from anti-system obstruction, in order to be established, talking legislatures depend on two additional necessary conditions which are jointly sufficient for a centralization of agenda control and the disempowerment of committees: a critical juncture and followers’ surrender of their inherited procedural privileges. Anti-system obstruction being absent, followers’ demand for mega-seats on legislative committees triggers a development towards working legislatures. Chapter 8 also emphasizes that legislative obstruction is subject to equifinality and discusses the scope conditions of this causal mechanism. The chapter notes that the procedural development of congresses in presidential systems is most likely susceptible to more multiple causation than that of parliaments. In conclusion, it appears that procedural reform in Western European legislatures over the last 150 years was primarily aimed at maintaining legislative democracy. Against this background, Chapter 8 concludes with a discussion of the alleged decline of legislatures and addresses options for countering the dual threats to legislative democracy posed by autocrats and populists.


(1) Unless otherwise indicated, all translations are those of the author.

(2) Previously, Max Weber’s original term ‘stahlhartes Gehäuse’ had been translated as ‘iron cage’ (see, e.g., DiMaggio and Powell 1983). Recently, Jan-Werner Müller (2011, 28) has convincingly pointed out that ‘steel-hard casing’ is a better translation.

(3) Both in the Kosovo and in Macedonia, nationalist oppositions blocked all legislative work for extended periods since 2015 (Kosovo) and 2017 (Macedonia). The Turkish constitution adopted in 2017 abandoned parliamentary democracy and allowed the president to regulate basically all policy areas (with some exceptions pertaining to the budget) by executive decrees. Even though the precise impact of the new constitution cannot be ascertained at the time this introduction is written, it is highly likely that legislative democracy will, if not break down, be strictly curtailed under the new constitution.

(4) On Congress, see Krehbiel 1991; Cox and McCubbins 1993; 2005; Binder 1997; Dion 1997; Schickler 2001. Spirling (2014) provides an overview of recent scholarship on the House of Commons.