Introduction

Modern societies depend increasingly on participation to achieve economic efficiency, both in the marketplace and in parliament. For example, in many universities, students assign marks to their teachers to help them be more effective; and many firms, public and private, ask customers to fill in forms that rate customer's satisfaction with the service provided. In politics the main form of participation is represented by the act of voting: It is, however, a very incomplete form of participation, and the economic consequences of this limitation need to be acknowledged and better understood. This essay points to three major issues. The first is related to the lack of representation for a significant share of the population, namely children and young people. The second issue regards the economic consequences of the relationship between younger and older voters, as revealed by the multipeak surface of the joint age–income distribution. A simple way to analyze this issue is to refer to the median voter model, exploiting the implied statistical relationship between median and average income. The third issue points to the problem of low political participation as a possible cause of higher violent crime rates, higher law enforcement costs, and a lower quality of life. Possible solutions to the issues raised are simply hinted, because they need to be specific to each country: in particular, the potential benefits of new information technology need to be explored.

Democracy and group census

Because it faces severe limits on its workability, democracy is not a panacea for politics. It works only on the margins of great issues. The few big issues it can handle are those on which there is broad consensus–such as the consensuses in the United Kingdom, Canada, and the United States on fighting World War II. Most forms of government could handle such issues roughly as well. Indeed, the United Kingdom handled World War II by ceasing to be democratically accountable for the duration of the war. For conflictual issues, democracy can work only against a background of rough coordination on order. Without that essentially prior coordination, democracy is trammeled or irrelevant. And even with the relevant coordination on order, if precise theoretical claims are at issue, democracy works only in the sense that it reaches a result–but not in the sense that it gets the right result. Often, indeed, a democrat would want decisions to not be made democratically.

Democracy is essentially a member of the mutual-benefit class of theories. If political divisions cut very much deeper than the marginal issues on which we can democratically compromise, democracy may no longer seem to produce mutual benefits. It then produces major–not marginal–winners and losers. Big disagreements bring us down. For example, democracy could not handle conflict over slavery in the United States, it was very shaky in handling socialism in the postwar United Kingdom, and it could not even get off the ground in independent Burundi with its first democratic election pilloried as merely an ethnic census.

Introduction

The quest for control of the public machinery has accompanied the entire history of political life, insofar as “the utility function of the agents is not identical with that of the ruler” (North, 1981, p. 25). Today, when we contrast the ramification of public action with the effectiveness of parliamentary control, it is easy to see that the quest is still going on, and if anything is more impelling. Should we conclude that the growth of the welfare state and the widening of democracy has brought about the dethronement of the voter? And who has gained effective power? The notion of bureaucratic capture is one answer, though it is difficult to assign the roles of captor and captive when the rules of the bureaucratic game are concocted inside the political game. Why should representative bodies accept to be dispossessed if they can enact new rules, deny money, and redress the balance of power at any time? Mathew McCubbins and Thomas Schwartz (1984) suggested that the control of the elected assemblies materializes through either “fire-alarm” or “police-patrolling” devices, the former being the case of the U.S. Congress called to act when needed, the latter of the more systematic control exerted by parliaments. Both systems, however, seem to face similar problems, and it is not easy to assess whether that metaphor effectively expresses real differences or generalizes scattered pieces of evidence. In either case, the intriguing question is to ask why the controls and the features of bureaucracies should be different under different democratic systems.

In this chapter we evaluate the assumptions underlying the model of intercameral bargaining developed in Chapter 4. We do so by analyzing six laws passed by the French legislature between 1971 and 1981. In Chapter 4 we argued that impatience to reach agreement drives bicameral bargaining and that impatience is generated by a number of factors: the desire of legislators to be viewed as efficient, the need to resolve fiscal or political crises, and the concern that legislators will defect from the political coalition supporting the bill. Moreover, we concluded that the relative power of each house in bicameral legislatures is a function of institutional constraints and the impatience of each legislature to reach a deal. In this chapter, we examine the assumption that impatience for legislative outcomes (1) drives the negotiating process and (2) determines the outcomes of bicameral negotiations in France (within the limits set by institutional constraints). As in Chapter 6, because we focus such attention on a single country, we also address the relevant French legislative analyses on relative house power. One explanation of senatorial power, expertise, is common to many countries, as noted in Chapter 1. The second explanation, presidential politics, is peculiar to France.

We proceed by defining the concept of impatience and our approach to operationalizing that concept. We then briefly present the two alternative models of French legislative behavior against which we pitch our own. In the third section, we evaluate the three competing hypotheses in light of six pieces of legislation. The detailed examination of the bills as they shuttle from one chamber to the other allows us to focus on the process of bicameral negotiations.

Bicameral legislatures are those whose deliberations involve two distinct assemblies. As such, they are relatively modern political institutions that only gained popularity in the eighteenth and nineteenth centuries, despite much earlier origins in the fourteenth-century English parliament. Nonetheless, earlier political institutions share some of the features of more modern bicameral legislatures, and the intellectual debate over the merits of multiple deliberative assemblies is centuries old. In this chapter, we trace the evolution of bicameral institutions and the intellectual justifications that underpin those structures. We do so by following two threads of analysis, what we call the political and efficient dimensions of bicameralism.

We begin with a definition of the political and efficient dimensions of bicameralism, as these two themes surface repeatedly in the text. We proceed via a chronological overview of the institutional evolution of dual deliberative structures in various eras, accompanied by an intellectual history. In Sections 1 through 3, we review the pre-bicameral institutions of ancient Greece and Rome, the development of the first, class-based bicameral legislature in Britain, and the creation of an alternative federal model in the United States. In Sections 4 and 5, we trace the divergent paths of institutional development in Europe that privilege the distinctions between federal and unitary political systems. Finally, we summarize the contemporary debate on bicameralism, which tends to emphasize either the political or the efficient dimension of bicameral legislatures.

THE TWO DIMENSIONS OF BICAMERALISM

Legislation changes the previously established way of doing things. Whether it regulates behavior in a new way or deregulates behavior and permits individual choice, legislation changes the status quo. In bicameral legislatures, the status quo may be modified in two ways.

Given the diverse national paths in the development of bicameral legislatures, it is not surprising to find substantial institutional variation. In this chapter, we explore these institutional differences. The characteristics of lower and upper legislative houses differ on a number of dimensions, and analysts have relied on these differences to explain cross-national variations in upper house power. Although there are variations in size, legislative term of office, turnover, membership, representativeness, and institutional power, two particular characteristics have been emphasized. The first is the membership of the two houses, based on selection methods and categories of citizens represented. The second is the relative power of the two houses as reflected in their mechanisms for resolving intercameral differences. Here we focus on these two critical dimensions of variation.

Political analysts who examine the variation in bicameral institutions argue that bicameralism produces disparate results across countries. Lijphart (1984) attributes variation to the degree of congruence between the two legislative houses and power asymmetries, whereas Mastias and Grangé (1987) focus on upper house legitimacy as the important independent variable.

Lijphart (1984: 99) defines “congruence” as similarity of political composition. Regardless of the variations in selection methods, if the two houses have similar political representation, they are deemed congruent. Disparities in power range from full symmetry, where agreement of the two houses is necessary to enact a law, to total asymmetry, where one house is granted decision-making power. Using these two categories, Lijphart constructs three types of bicameralism. He argues that “strong” bicameral legislatures are characterized by significant differences in composition and by relatively symmetric power. “Weak” bicameral legislatures are characterized either by asymmetric power or by congruent chambers.

As we discussed in Chapter 2, conference committees are frequently employed to reconcile intercameral differences. Conference committees are important because, in most cases, they are the last stage where legislative compromises are worked out and legislation is forged. The outcome of conference committee deliberations is then introduced to the parent chambers for approval under closed rule, that is, without possibility of amendment. Consequently, a strategic conference committee can select, from among all the proposals that the parent chambers prefer over the status quo, the one that is closest to its own preferences. The parent chambers may dislike one provision or another of the compromise bill, but they cannot alter the specifics; they can only reject the whole bill and start the legislative process again. Consequently, if a bicameral legislature resolves its differences through a conference committee, the compromise elaborated in this committee usually becomes the law.

In all countries where the institution exists, except for the United States, the committee is composed of an equal number of members from each chamber, and it decides as a unicameral institution. In the U.S., the size of the conference committee varies as does the number of delegates from each house. Decisions are made by the so-called unit rule, that is, by concurrent majorities of each chamber's delegation. In countries where committees decide as unicameral institutions, the decision-making rule is by majority, with the exception of Japan, which requires qualified majorities.

To analyze the U.S. conference committee, we refer the reader to Chapter 3, since the unit decision-making rule is the same as bicameral decisions. However, before proceeding, we point out one aspect of the unit rule that we will refer to subsequently.

In the next four chapters we empirically investigate the predictions developed in Part II and compare our findings with alternative theories of bicameralism outlined in Chapter 1. In Chapters 6 and 7, we draw on data from the French Fifth Republic. The prominent position of France is attributable to the complexity of its institutions: the two houses may be equal or, upon decision of the government, unequal; the deliberation may involve a conference committee or not; the navette may conclude after a single round or last indefinitely. With the exception of the United States and the European Union, no other country or institution has bicameral procedures that vary so extensively. Thus, France presents an opportunity to test various dimensions of our models and to evaluate institutional features of the bicameral systems described in Chapter 2 within a single country, so that we can hold a variety of conditions constant.

Chapter 8 relies more heavily on data from other countries in order to capture the variations in conference committee composition, decisionmaking rules, and constraints imposed by the parent chambers – the characteristics of conference committees that are important independent variables in our theoretical models. The tests presented here are far from exhaustive. We strongly believe our models should undergo additional evaluation employing data from a wide variety of countries. Nonetheless, the evidence presented here helps to establish the plausibility of the models developed in Part II.

In Part III, we employ both statistical analysis and case studies to evaluate our theoretical arguments. Undoubtedly, it would be helpful to extend statistical analyses further than we have done in this book.

The goal of Part I is to present the diversity of bicameral institutions over time and among countries, as well as the diversity of theoretical justifications employed to promote them. Bicameral institutions have been adopted by class societies and by federal states, by republican polities and by unitary political systems. They have been used to maintain the status quo, to amalgamate the preferences of different constituencies, and to improve legislation, and have been justified in all of these terms. The diversity of bicameral legislatures extends to the institutional devices employed to resolve differences between the two legislative chambers, the navette (shuttle) system, conference committees, joint sessions of two chambers, one-chamber decisions; these are rules that may be applied individually or in combination. In a word, bicameral institutions are protean. They change forms and functions, and the arguments that justify or explain them follow these transformations.

This part of the book is divided into two chapters. Chapter 1 focuses on the history of bicameral diversity and presents the different theoretical justifications for bicameralism, while Chapter 2 provides information about the geographic dimension, the institutions regulating bicameral interactions in different countries today.

Chapter 1 begins with the earliest known political institutions that divided legislative deliberations between two or more chambers and follows the institutional evolution up to the modern period. At the same time, we trace the intellectual debate about the merits of this institutional structure. We describe the reasons proffered by various politicians and political theorists to justify the existence of two legislative bodies. The historical presentation emphasizes the dual evolution of institutional structures and intellectual debates.

Conference committees are frequently employed to resolve disagreements between chambers that remain after one or more rounds of the navette. In Chapter 5, we pointed out that the importance of conference committees lies in their ability to make proposals to the parent chambers under closed rule, that is, without amendments. Consequently the details of a bicameral compromise are worked out in the conference committee, without possibility of new input from the parent chambers. Delegating the power of agenda setting to the conference committee presents the parent chambers with a serious danger, a “runaway conference” in U.S. terminology (Longley and Oleszek 1989: 4–5). The runaway conference is a conference committee that proposes compromise positions that either differ from the common positions of the chambers or exclude common positions of the chambers.

To avoid this danger, the parent chambers have two ways to rein in conference committees. The first is the explicit, restrictive, and credible specification of the set of acceptable solutions. Where the lower and upper house versions of the bill follow the same structure and disagreements are located at specific points, the conference committee may be restricted to discussing only those aspects remaining in disagreement and to locating a compromise somewhere between the positions of the two parent chambers. Conversely, where the two versions of the bill differ widely, sharing only the topic of legislation, the leeway of the conference committee expands to the maximum. In Chapter 5, we noted that since the compromise must be within the bicameral restrictions, reducing the space contained within bicameral restrictions will reduce the freedom of choice of the committee.

This chapter is designed to test the model developed in Chapter 4. Consequently, we retain the assumption that bicameral negotiations are driven by impatience to reach agreement. We operationalize impatience in terms of the strength and breadth of the governing political coalition. The reader should suspend judgment about the adequacy of our choice until the following chapter, where we take time to examine the impatience assumption critically.

The complete information model developed in Chapter 4 connects different institutional features of the navette system with outcomes. The incomplete information model of Chapter 4 predicts that under conditions of one-sided incomplete information, the number of negotiating rounds in bicameral legislatures increases with one house's uncertainty about the other house's impatience (time discount factor). In more common political terminology, relations between the chambers should be more acrimonious under conditions of uncertainty; it will take longer for the two houses to reach agreement.

We test these predictions with data from the French legislature under the Fifth Republic. France represents a natural test of the model of one sided incomplete information because the composition of the Senate remained relatively constant while the composition of the National Assembly varied widely – from a Gaullist to a Socialist majority, by way of a centrist–Gaullist coalition. We argue first that the composition of the National Assembly affects the impatience of that legislative body to reach agreement. We argue second that the changing composition of the National Assembly introduces uncertainty about the impatience of the legislative body in a nonlinear fashion.