Introduction: On the Nature of Parliament(s)

The term parliament is often used to describe a particular type of contemporary democracy, namely those in which elected representatives choose the leader of the executive branch and can replace that leader in various circumstances. When used in this way, parliament and parliamentary governance are synonymous. They are typified by the British, Swedish, Danish, and Dutch systems of government. However, this unnecessarily restricts the term parliament to a relatively small subset of systems of governance, most of them relatively new, in which the selection of policymakers is organized in a particular way.

Prior to the 19th century, all these European governments had representative bodies that were considered parliaments but that had relatively little authority. Prior to 1800, these and other European parliaments were relatively large representative bodies that had some ability to propose new policies and veto proposals for new taxes. They were for the most part multi-cameral, and for the most part their members were unelected. Three-chamber parliaments were commonplace, with one body representing nobles, another the hierarchy of the national church, and another relatively wealthy commoners. Four-chamber parliaments also existed in which wealthy commoners were divided between urban and rural chambers, as in Sweden. Two-chamber parliaments also existed. The one in Great Britain combined the noble and church chambers into a single chamber, one that was initially dominated by church leaders (the House of Lords).

For most of this period, the noble and church chambers were far more important than the commoner chamber(s). In the early days of European parliaments, nobles often had their own armies. Together, these were larger than the king’s, and discussions in the noble chambers played a significant role in national military strategies. As the military was centralized, the nobles typically financed a significant part of that military and had veto power over new taxes. Partly because of this, but also because nobles were educated and participated in their own international networks, essentially all senior posts in the executive and military were from the church and noble chambers until the late 19th century. History makes it clear that the association between parliaments and democracy is a relatively recent phenomenon and one that is by no means universal.

If we characterize “parliament” with its European history in mind, the term parliament should be used to describe any relatively large representative body that is formally part of a nation’s government and has nontrivial authority over a significant subset of public policies. Defined in that way, the term parliament would include both the representative chambers of contemporary parliamentary governments and the representative assemblies of most authoritarian regimes. The English House of Lords and People’s Congress of China, for example, should be regarded as parliamentary assemblies, although their members are not selected through competitive elections and have relatively little influence over government policies.

This use of the term parliament implies that several natural programs of research exist concerning parliaments. First, there are the static questions often focused on in political science and political economy. What formal authorities do they have? How are their memberships determined? How do such bodies actually make decisions? How do these factors influence public policies? Second, there are the dynamic questions. Why do parliaments exist? How do their authority, selection, and decision-making procedures change through time? Is there any tendency for the institutions of parliament to converge on particular procedures and architectures?

All of these questions have been explored to some degree in political and historical research, although the first has attracted by far the most attention among contemporary researchers.

Origins of Parliament

Governments with representative bodies are essentially as old as civilization itself. Tribal societies were often ruled by a chief and council of wise men that made decisions through deliberations of one kind or another. As governmental authority expanded, this template was modified, and the “chief’s” authority tended to increase, in some cases to the point where the chief alone could be regarded as the fount of governance. However, even in such cases, influential advisory councils and assemblies of regional administrators were commonplace. There is evidence of such systems in Mesopotamia, China, and Egypt, as well as in ancient Greece. Systems of governance based on what has been called the “king and council” template are historically commonplace, which requires some explanation. There are several possibilities, but most likely the use of this template for policymaking advances the purposes of both those who create governing organizations and those who subsequently occupy positions of authority within them.¹

All organizations confront a variety of PDE (prisoner dilemma with exit) and coordination problems. Standing procedures for making policy decisions have to address those problems to function and in order to survive must refine their solutions as circumstances change. The latter is why governance in addition to rules are necessary.

What might be called “government,” “management, “collective decision-making,” or “rule making” benefits from specialization like most other tasks. This implies that only a subset of an organization’s members are likely to participate directly in policymaking. It is this subset that should be considered “the government.”

Members of such rule making or “ruling” groups tend to have a comparative advantage at planning. They may be a bit more informed and more skilled at assessing the relative merits of alternatives (which is to say “wise men”—men with what Aristotle termed “practical wisdom”). They may also be more interested in participating in such decisions and so devote their energies to being included in decision-making bodies (which is to say, they take steps to become leaders or “men of action”). In most cases, those “in charge” exhibit both a relatively strong desire to influence policy and backgrounds that tend to generate superior knowledge of the affairs of government. One rarely stumbles into positions of authority, except those that are hereditary.

Why a group of policymakers organize themselves into systems of governance based on the “king and council” is less clear. Congleton (2011, 2001) argues that this template solves a variety of informational, incentive, and secession problems. A committee, council, or parliament will normally have a greater pool of knowledge and experience to draw on than any single individual and so can both offer useful advice and make more informed general decisions than any single individual is likely to be able to do. However, members of such deliberative bodies may free ride on many of the decisions and tasks required for the active management of their organization’s operations. A chief or chief executive has better incentives to undertake day-to-day management than a group, either because he or she may directly benefit from good management and be harmed by poor management, or because he or she is held responsible for his or her management efforts.

The king and council template combines the informational advantages of group deliberations with the better incentives of one-person rule (e.g., what economists refer to as residual claimantcy). Secession problems can be solved in a variety of ways under this template. For example, the council may choose a chief’s successor or do so when the chief fails to do so, and vice versa. In addition, the distribution of authority between the king and council can be adjusted at a variety of margins, which allows the organization (here the government) to take advantage of the relative skills of members of the rule-making bodies. Authority can be shifted between the king and council and among members of the council (parliament) to reward persons for their relative contributions (or usefulness) to the government. Some persons or groups may provide more resources or more critical support to the governing organization than others, and such persons and groups may be rewarded with relatively greater authority in a council or parliament, or in the executive.

All these properties tend to increase the long-run survivorship prospects of governments based on the king and council template. Their relative success, in turn, would draw that template to the attention of persons forming or reforming other governmental organizations. These practical considerations imply that robust governing organizations tend to share various properties in the long run, which at least partly accounts for similarities in the core architecture of contemporary national governments and many private organizations. Most private organizations also have councils or parliaments of one kind or another and chief executives (Congleton, 2001, 2011, 2013).

Reforms of Parliament

In the long run, governments can undergo a broad range of reforms, while retaining the same basic template for governance. With respect to parliaments, the potential for reform is enormous. Members may be selected in different ways. The number of chambers and their internal deliberation and decision-making procedures can be modified. Organized groups within the chambers of parliaments—such as political parties or committees—may possess more or less formal and informal authority. Parliament itself may have more or less influence over the policies adopted and the personnel of the executive branch such as the chief executive and cabinet members.

Nonetheless, in order for governments to make decisions, most of its deliberative and selective procedures have to be relatively stable. The procedures and personnel for making policy decisions must be more stable than the policies chosen, if a government can be said to exist.

This stability is generated in several ways. First, at least some potential reformers doubtless recognize the value of stable procedures for making rules. Stable rules tend to reduce conflict and facilitate decision-making. Second, the persons with the authority to change the procedures of policymaking often directly benefit from the existing rules. They typically benefit from the distribution of income, authority, deference, and status associated with the existing rules. Moreover, it is under such rules that most members of durable governments have risen to positions of authority, and changing the rules may undermine their claims to the fruits of high office. Third, risk aversion combined with uncertainties about the effects of relatively complex reforms tends to generate a status quo bias, especially with respect to major reform. Fourth, formal and informal amendment procedures tend to make reforms of the existing procedures of governance more difficult than reforms of the policies adopted using those procedures. All these factors tend to generate a predisposition to retain existing procedures and to modify them incrementally rather than whole cloth.

The result is greater institutional stability and path dependency than one might anticipate given the wide range of governing institutions that are conceptually possible. The United Kingdom, for example, has had a parliament with a House of Lords and House of Commons since the 14th century (although the names for the two chambers emerged somewhat later). Over the course of several centuries, the balance of authority between the king and parliament and within parliament fluctuated to a substantial degree, changing radically in the 18th and 19th centuries. By the 20th century, parliament had become the dominant branch of government and the House of Commons the dominant chamber of the Parliament. This was essentially the opposite distribution of authority that the English government had in the 14th and 15th centuries.

The first suffrage laws for the House of Commons were adopted during the first half of the 15th century. These rules were remarkably stable and not substantially reformed until the 19th century. After a series of reforms in the 19th century, the House of Commons came to be elected via far broader suffrage than it had been during its first several centuries. Universal adult suffrage, however, was not the rule until after World War I (Morgan, 1984).

Many other parliamentary systems underwent similarly wide fluctuations in the distribution of authority in the period before 1800 and subsequent shifts in authority and suffrage in the 19th century favoring their parliaments and democracy. Political parties emerged throughout most of Europe in the 19th century during the period in which suffrage and parliamentary authority expanded (Congleton, 2011). Contemporary Western parliaments thus took their present forms gradually during the past two centuries with noticeable differences emerging every few decades during the late 19th and early 20th centuries.²

Newer parliaments may change more rapidly than the ancient European ones, because their standing routines for governance are less matters of habit and tradition, but nonetheless reforms tend to be relatively modest. Most African countries, for example, have systems of governance with internal architectures based on their post–World War II independence constitutions, albeit with considerably different balances of authority than many anticipated at the time of independence (Congleton & Yoo, 2017).

Effects of the Institutions of Democratic Parliaments on Public Policy

Parliamentary Decision-Making: Coalitions and Committees

Parliamentary history provides a good deal of evidence that the internal organization of parliament affects parliamentary decisions. One aspect of the internal organization of parliament is its cameralism, another is the formal and informal organization of members of parliaments into factions and committees. In the early days, factions within the chambers tended to reflect informal groups that shared family or regional interests. In most contemporary parliaments, factions are largely determined by membership in formal political organizations referred to as parties. Other formal subdivisions within parliament reflect advantages of specialization, as various committees emerge and are staffed by parliamentary leaders. All such internal divisions have effects on parliaments by affecting the details of the policy proposals voted on, the most likely majority coalitions, and the stability of those coalitions. Some parts of the internal organization of parliaments are consequences of constitutional law, but many are the consequences of choices made by the members of parliament themselves.

Election law has indirect effects on the internal organization of parliament through effects on the number of political parties likely to be represented. District-based majoritarian systems of representation tend to have two or three parties in parliament. Electoral systems based on proportional representation tend to have more parties, which reflect the electoral thresholds required for seats in parliament, as explained by Duverger’s (1963) law. The lower the threshold, the more parties tend to be represented and the less likely any of them will hold a majority of the seats in parliament (Grofman & Lijphart, 2003). Although there is a sense in which every party is itself a coalition, parties are organized coalitions with their own systems of government, conditional rewards, and punishments. Coalitions among persons without such formal organizational structures tend to be less robust and more likely to break down. The latter may be good for democracy and equity, as Buchanan (1954) argues, but it makes governing more difficult for those in positions of authority.

Riker (1962) developed a theory of coalitions, which explained why most majorities in parliament include about 50% of the membership rather than 90%. Riker’s minimal winning coalition rule may be violated when voters are organized into relatively large blocks of voters (Kreppel & Tsebelis, 1999; Kreppel, 2000) or when coalitions are relatively frail and a small supermajority is necessary to align incentives within the coalition (Diermeier & Feddersen, 1998; Diermeier, Eraslan, & Merlo, 2006). A relatively small majority minimizes the concessions that have to be made to others to get one’s preferred policies adopted.

One problem with coalition-based governance is that most coalitions are inherently unstable according to most rational choice models. There is not stable solution to dividing the pie contests settled by majority rule when there are large numbers of parties. The threat of breakdown undermines the stability required for a coalition to pass legislation or form a government in parliamentary systems. Stability is increased when each party’s ideal vector of policies (its platform) can be mapped into a single-dimensioned issue space. This tends to reduce the number of plausible coalition partners and so makes alliances more stable (Kreppel & Tsebelis, 1999).

Other institutions of parliament provide additional stability by making coalitional agreements more enforceable. In addition to the stability that may be generated by parties and ideology, Weingast and Marshall (1988) and Shepsle and Weingast (1981, 1987) suggest that the committee structures through which bills are developed and the rules through which proposals can be amended prior to chamber-wide votes tend to stabilize both policies and governing coalitions. The former makes vote trading possible within coalitions through decision-making procedures that make reneging on promises less likely to be advantageous. Such institutions may be prerequisites to democratic governance, although they remain poorly understood and underappreciated.

Parliaments in Authoritarian Systems

Most contemporary research on parliaments focuses on systems in which elected representatives occupy most seats in parliament, and the parliament has dominant or nearly dominant authority over public policy decisions. Such parliamentary systems are relatively new in history. Historically, parliaments were far more often secondary than primary centers of policymaking authority. This older balance of authority remains commonplace today. As in the medieval systems, elections for parliament often take place in contemporary authoritarian systems, but the candidate slates and party competition are restricted in various ways, as in Iran and China. In some cases, “king-dominated” systems simply continued in place, albeit with various modifications. In others, a coup might have ended the electoral basis of the executive and rebalanced what would otherwise have been a more liberal constitutional system in favor of the “chief executive.”

This is not to say that parliaments have no formal authority in such systems, nor is it to say that their de facto authority is necessarily less than their de jure authority. However, authoritarian parliaments clearly have less authority than those in liberal regimes and so have a smaller role in policy formation. The parliaments of authoritarian states have attracted less attention from political scientists and political economists than democratic regimes, partly for that reason.

Research on authoritarian regimes has also been made difficult by the lack of public records and transparency that often characterizes such regimes. Nonetheless, there is a good deal of both historical and statistical analysis of authoritarian regimes at what might be called the macro level. Parliaments are common in authoritarian states. Authoritarian regimes tend to have smaller public sectors and relatively poorer typical (median) citizens than ones with parliaments grounded on competitive elections and possessing greater authority, unless they have extensive mineral wealth.

Conclusion: What Do We Know about Parliaments That Aristotle Did Not?

When assessing what one has learned after a project, it is often useful and humbling to revisit Aristotle’s Politics (1981). This is not because Aristotle had “it” all worked out, but because he and his colleagues had worked a good deal out about the politics and institutions of classical Greece that generalized to the rest of human history. The Politics reveals that political science is possible—which is to say that general principles can be developed that can explain many of the political facts around us—and also that it is possible for a single study to make some progress toward identifying such principles.

When one reads the Politics with contemporary research in mind one cannot help but see that substantial progress had already been made in 330 bce on issues that are still being researched. With respect to parliaments per se, translators do not use the word parliament when discussing Aristotle’s analysis of representative bodies. Nonetheless, there were many assemblies that can be regarded as parliaments in the sense used here. Aristotle regarded such systems to be forms of aristocracy and recognized that the rules through which representatives are selected had effects on public policies. For example, he distinguished between assemblies based on wealth, virtue, and votes. He argued that assemblies based on virtue and the votes of middle-class voters tended to generate the most robust parliamentary governments. He also discussed how different rules for citizenship (with its associated rights to participate in politics) can affect outcomes but did not systematically explore how alternative voting rules or divisions of authority within such representative bodies or within divided governments tended to affect policies. Such concerns emerged much later, and systematic treatments not until well into the 19th century.

New areas of research emerged in part because the scale and scope of governance increased in the 18th and 19th centuries. A city-state in Aristotle’s day would be only a large town in today’s world. The greater scale and scope of contemporary politics doubtless account for the rise in parties, the decline in direct democracy, and much of the internal specialization of contemporary parliaments. The rise of ideological parties in the 19th centuries changed the nature of politics within and without parliaments. Additional new topics for analysis emerged as relatively large nation-states came to be grounded in elections with broader suffrage.

Investigations undertaken before World War II also generally lacked the mathematical models and statistical methods that contemporary work often relies on to develop and test hypotheses about factors that may account for the political facts at a given time and place. The quantitative analyses made possible by the new methods allow old questions to be addressed more thoroughly and often with greater precision than possible in the past. Together they allow relatively objective conclusions about what is general and what is not to be reached on both ancient subjects and ones never analyzed before.

As a consequence, a good deal of recent progress has been made in understanding how parliaments operate and how they affect policies, although both parliaments themselves and the policy issues decided have become more complex through time. Of course, there is much left to be done. Both historical and contemporary parliaments can be examined more carefully. Most statistical studies are flawed in one way or another, because of data limitations and modeling problems. Most models abstract from institutions that doubtless matter. Relatively little work has been undertaken on the effects that parliaments have on authoritarian policies and on the dynamics of parliamentary reform. Although much progress has been made since Aristotle’s time, there is still much left to do.