Coalition Governance Institutions in Parlamentary Democracies - Strom

Coalition governance institutions in parliamentary democracies

Kaare Strøm
Department of Comparative Politics, University of Bergen
Christiesgate 15, N-5007 Bergen, Norway
and Department of Political Science, University of California, San Diego
9500 Gilman Drive, La Jolla, CA 92093-0521, USA
(kstrom@ucsd.edu)

Wolfgang C. Müller
Institut für Staats- und Politikwissenschaft, Universität Wien
Hohenstaufengasse 9, A-1010 Wien, Austria
(wolfgang.mueller@univie.ac.at)

Prepared for presentation in Workshop 13 Institutions at the Joint Sessions of Workshops of the European Consortium for Political Research, Mannheim, March 26-31, 1999. We thank Torbjörn Bergman and Elisabeth Gerber for comments and Magnus Blomgren and Scott Kastner for valuable research assistance.

Introduction

Coalition governance has long been at most a marginal concern in the literature on cabinet coalitions. In many ways, the cabinet coalitions literature resembles the romantic Hollywood films of the 1950s. Much is made of the courtship process and "who gets whom," whereas relatively little light is shed on how such alliances actually work. It is relatively easy to report the external parameters of the bargaining process (e.g., its duration), as well as its result in partisan composition and portfolio allocation. It is much more difficult to establish the fine detail of governance and implementation under coalition rule. Too often, studies of cabinet coalitions have therefore satisfied themselves with a recording of the formal membership, and perhaps the duration, of these governments. The recent theoretical literature, however, reflects a greatly enhanced interest in the process of governance once a cabinet coalition has been formed. As of yet, however, there is little agreement as to how governance can be understood, or what the most important governance institutions are.

Governance problems in multiparty parliamentary democracies

Coalition governance involves two problems that have received extensive attention in the rational choice literature, though much less in the coalition-specific field, namely bargaining and delegation. In this paper, we shall address the nature of governance problems in multiparty parliamentary democracies. We begin by examining the significance of delegation and bargaining in parliamentary systems. We then discuss some generic problems that coalition agreements entail. Arguing that parties in coalition governments face complex and severe bargaining problems that lead them to devise more or less incomplete agreements, we lay out some of the options facing them. We show that coalition agreements may take a variety of forms and range from relatively tight to very loose accords. With the help of transaction cost analysis, we develop expectations as to what sorts of governance institutions will be chosen under what conditions. Finally, we introduce a novel set of data on coalition agreements and governance institutions in 13 European countries. These data have been collected by a large cross-national set of researchers. Though our confrontation with these data is as of yet preliminary, they should ultimately be a most valuable and unprecedented source of data on coalition governance.

Parliamentarism and coalition politics

Despite the large literature that has evolved on democratic regimes, there is no universally accepted or unproblematic definition of parliamentary government or democracy (see Lijphart 1992; Strøm 1999). In this paper, we shall define parliamentary government as a system in which the prime minister and his or her cabinet are accountable to, and can be dismissed by, any majority of members of the national parliament (or in the case of bicameral parliaments, the lower house). Parliamentary democracy implies that government policy making is controlled by a legislative majority. When and where this form of government first emerged, in 19th century Britain, its growth coincided with the development of the classic British two-party system (Cox 1987). In its original form, parliamentary government was majoritarian, or Westminsterian, in Lijphart's (1984) influential conception. The belief in the unfettered rule by the popularly elected majority is key to the tradition of parliamentary government. The Westminster tradition is at heart a tradition of parliamentary supremacy. As Verney observes, "the political activities of parliamentary systems have their focal point in parliament. Heads of state, governments, elected representatives, political parties, interest groups, and electorates all acknowledge its supremacy" (Verney 1992: 46).

But parliament is not a player, but rather a forum or arena in which individual representatives meet and interact. With a number of caveats, parliaments typically operate by majority decision rules (Rasch 1995). In order to examine governance under parliamentary government, we therefore have to identify the critical players that constitute such majorities. What, then, are the units (the players) of which coalitions are composed? As Laver and Schofield (1990) point out, this question can in principle be answered in many different ways; yet in practice the literature almost always identifies them as parliamentary political parties. Explicitly or implicitly, these are typically assumed to be unitary actors. It is by no means a foregone conclusion that unitary parties should be so analytically privileged. Rational choice theories are fundamentally committed to the assumption of methodological individualism, to the principle that social events ultimately can only be adequately explained as the results of individual choices and actions, and that every collective phenomenon should be reducible to such analysis. By this logic, which underlies the most important research tradition in the field, any satisfactory account of cabinet governance would have to cast its analysis in terms of the actions and motivations of individuals. Political parties, of course, are not (or extremely rarely) individuals, but rather groups or organizations, aggregates to which we cannot legitimately ascribe motivations or rationality (see Arrow 1951).

Party politics

Nonetheless, coalition theorists brazenly and virtually unanimously stipulate cohesive political parties as their unit of analysis. Their reasons are in part pragmatic: The explicit analysis of all feasible coalitions in a parliament of, say, 300 or 400 members, would be a task of such daunting complexity as to practically preclude any possibility of tractable theoretical analysis (Laver 1989). Besides, empirical knowledge of the real world of cabinet politics firmly suggests that party unity is the rule much rather than the exception. It seems reasonable to suggest that the parliamentary representatives, and more generally the leaders, of a political party have a strong collective interest in cohesive behavior vis-a-vis other parties. They may have a collective action problem in maintaining a united front, but the empirical records amply document that they seem able to solve this problem most of the time (Laver and Schofield 1990).

Interparty bargaining

With two and only two cohesive parties, one of these would always have a legislative majority, and the problem of interparty coalition governance would simply not occur. With the spread of parliamentary government to Continental Europe, however, and the introduction there of proportional representation, multiparty politics has become the 20th-century norm in parliamentary democracies. In multiparty parliamentary democracies, the possibility always exists that no party alone will command a parliamentary majority. Such minority situations (Strøm 1990a), then, require coalition-building of some sort. Coalitions in minority situations could be purely legislative alliances, in which a minority government seeks support from day to day or from issue to issue among its companions in parliament. More commonly, however, coalition-building involves executive offices as well, so that the parties that form the parliamentary majority also share control of the cabinet and the executive branch. Indeed, that possibility has become the rule, rather than the exception, in most parliamentary systems.

Coalition governments manifest some sort of bargain struck between their participating parties. Political parties, or more specifically their leaders, are likely to have different preferences over a broad range of issues, driven at least in part by their motivation to secure for themselves policy benefits, office benefits, and electoral advantage (Strøm 1990b). All of these goods are likely to be scarce, so that not all demands can be satisfied simultaneously. Government offices and other benefits are finite and often more or less constant-sum. Ministerial portfolios are typically unique and indivisible. If the prime ministership, for example, is awarded to one party, it is necessarily lost for all others. Electoral advantage, to the extent that party leaders care foremost about vote shares, is even more strictly constant-sum. Ultimately, one party's gain has to be someone else's loss. In some cases, all incumbent parties may be able to gain at the expense of the opposition, but much more commonly, coalition parties are engaged in some sort of mutual competition for votes (Rose and Mackie 1983). And, finally, while the parties that make up a coalition may have more or less compatible policy preferences, it is hardly ever the case that all of their policy preferences can be realized simultaneously.

Problems: uncertainty and enforcement

Such conflicts of interest within governing coalitions generate a need for interparty bargaining and compromise. Coalition studies have traditionally given little attention either to the process of coalition bargaining, or to its results. The only notable exception is to be found in the study of cabinet portfolios, which of course are one manifestation of the agreements reached by negotiating parties. We see this as an important lacuna in the study of democratic governance in parliamentary systems.

Limited information

Bargaining over government formation can take place whenever there are players (in our case, political party leaders) that can all make themselves better off by reaching an agreement, and when there are multiple such feasible agreements that may be reached. In cabinet politics, these agreements presumably concern the set of political offices to be filled, perquisites and favors to be distributed, and public policy to be conducted. While the prime ministership is typically a discrete prize that can be given to only one party in each coalition, there may be a huge number of ways in which the remaining portfolios could be allocated, and the number of possible budget allocations are virtually infinite. Needless to say, the relevant parties typically have distinctly different preferences over these possible solutions. In many cases, there are multiple such solutions that fill the requirement of meeting every negotiating party's reservation level, the value it could derive without any agreement. Indeed, it is probably fair to assume that coalition bargaining does not occur in the absence of such incentives. Nevertheless, coalition bargaining does not always succeed. In many cases, efforts to put together coalition governments fail. In other cases, coalitions are successfully put together, but then fall apart in disagreement, or simply fail to yield the benefits many of the participants may have expected. The frequency of bargaining failure of one sort or another forms an important background to our scrutiny of coalition agreements.

Enforcement problems

Coalition bargaining is no easy matter. This is partly because of problems that are generic to all forms of political bargaining, and partly due to factors that are more parochial and have to do with particular problems of cabinet governance. Briefly, many of the problems inherent in coalition bargaining stem from the salient features of such situations, namely (1) limited information, and (2) lack of enforceability. Let us consider these issues in turn.

As in the case of the life of a child, the contingencies that may arise during the tenure of a government are exceedingly difficult to predict with any kind of accuracy. Consequently, the task of writing a complete contract for such a coalition is also very challenging. In most countries, even the duration of a legislative term is not certain at the time a coalition agreement is reached. Parliamentary dissolution powers may leave this at the discretion of agents that will not be parties to any coalition agreement, such as, for example, a separately elected president. The fact that parliament may thus be dissolved, and the coalition potentially terminated, under circumstances beyond the control of the contracting parties further complicates bargaining. More generally, the demands placed upon national governments are so complex and unpredictable as to make such agreements a daunting task even under the best of circumstances. This is, of course, reflected in the fact that the formation of governments involves a great deal of delegation of power to agents, such as cabinet ministers and their subordinates, with highly specialized tasks. It is precisely in recognition of the fact that the principals (here, the coalition parties) cannot hope to gain enough information to efficiently instruct their agents (here, cabinet members and other appointees) that such delegation occurs.

One reason that political parties involved in coalition bargaining so often fail to reach their goals is that their agreements are difficult to enforce, and that the parties themselves recognize this problem. Parties engaging in coalition agreements face an obvious enforcement problem. Unlike most business contracts, there are no courts to enforce coalition agreements. In Shepsle's terms, coalition agreements cannot be based on exogenous third-party enforcement but must instead rely on endogenous compliance (Shepsle 1996).

A particularly tricky problem in coalition bargaining can be the sequence in which the parties secure their gains. If the payoffs to all are evenly distributed over the anticipated reign of the government, then each will be most likely to have a constant incentive to keep the coalition going. If, however, the payoffs are very unevenly distributed over this period, the party which has already received a large part of its share (e.g., a major piece of legislation) early on has little incentive to fulfill its part of the deal later. This problem of non-simultaneity of exchange (Weingast and Marshall 1988) may prevent the enforcement of agreements and thus destabilize coalitions. Consequently, parties may try to avoid such misalignments of incentives whenever they can.

Institutional solutions

The bargaining problems that plague parties engaged in coalition bargaining are thus serious and profound, though not necessarily beyond any reasonable solution. There are both informal and formal (institutional) sources of solutions to such bargaining problems. Sometimes bargaining problems can be solved by informal means. This could happen for several reasons. Either the preferences of the parties are so well aligned as to facilitate a successful bargain. Alternatively, the anticipation either of repeat interactions in the same arena, or of other interactions with the same players in different arenas, may induce party leaders to honor otherwise unenforceable agreements and thus underpin successful negotiations (see, e.g., Shepsle 1996: 229-231; also Axelrod 1984; Frank 1988; Kreps 1990).

Yet, parties that enter a coalition government cannot always hope that their preferences will be sufficiently well aligned, or the shadow of future interactions sufficiently long, to induce cooperation. Hence, they cannot rest assured that whatever agreements they reach will be respected.