1 - Introduction

Summary

The burgeoning literature in comparative constitutional has not devoted sufficient attention to the constitutional functions of political parties, nor has it systematically explored the constitutional law of electoral design. This volume examines the constitutional treatment of parties and elections both as a matter of constitutional theory and from the perspective of historical and contemporary practice. To this end, it draws together a series of contributions from a diverse range of scholars working in distinct disciplines. Political scientists tend to treat political parties as their key object of study, while comparative constitutional lawyers have largely ignored them, preferring to focus on other institutional question. What follows brings each perspective into conversation with the other.

A central function of a written constitution is to set forth “rules about rules,” otherwise known as metarules, that create a structure for political life. Constitutions do so by providing clear institutional pathways, instructions and limits that guide the subsequent creation of ordinary law and regulations, as well as by shaping the exercise of official discretion. Because these metarules are durable, they do not need to be recreated repeatedly – hence enabling democratic choice over the many policy questions of ordinary quotidian politics. In a democratic society, among the most important constitutional metarules are those that concern the principal offices of state, typically specifying the “who,” the “how.” and the “when” of selection of their occupants. Because of the centrality of elections to modern democratic practice, the constitution’s core task of institutional specification, at least in the context of a democratic order, almost always entails close attention to how electoral contestation will unfold, what spatial units of representation will be used, how preferences will be translated into outcomes, the duration of office-holding, and the terms or conditions concerning when and how transfer of power will (or must) occur.¹

In modern accounts of how democracies work and thrive, these who, how, and when questions are considered in light of the ambition to create durable democratic rotation through broad popular participation in meaningfully competitive elections. The practical operation of elections, in turn, usually depend upon the existence of political parties.² The latter typically (albeit not inevitably) play a central role in providing legible choices and hence lowering information costs for voters (which, Tarunabh Khaitan in Chapter 4 argues, is one of four key democracy-related costs that parties reduce). The party label conveys information and allows voters to both punish incumbent regimes and meaningfully anticipate the effects of voting for a neophyte candidate. In their contest over the possession of representative political offices, political parties hence critically provide an epistemic structure for political life that endures over time and between different elements of government. They give electoral politics temporal and geographic cohesion and,. as Khaitan puts it, “grease the wheels of representative democracy.”³

It is not just that parties are central to the functioning of constitutions. The choice of constitutional arrangements conversely shapes parties in vital ways. That is, a constitution’s design creates and distributes the incentives for party formation and organization. Parties emerge endogenously within a particular constitutional order. Hence, they tend to be tailored to those regimes (although this does not necessarily mean that their incentives align with the goal of regime preservation). When designing offices and electoral institutions, then, constitution-makers are also calibrating the incentive structures of political life. By shaping the rules of politics, they are crafting its core players.

Less frequently noted, but no less important, is the role that political parties play in promoting a constitution’s endurance. Parties emerge in ways that are adapted to specific constitutional arrangements. A change to those arrangements can be costly for a party, inducing a sort of Burkean orientation toward institutional rules. Relatedly, a party’s collective leadership typically has a longer time horizon than a particular leader. Because the party (if not the leader) often has an investment in the endurance of the constitution’s metarules as such, it may act to check a potential strongman or -woman. This offers a subtle friction against a leader’s aggrandizing aspirations toward one-person rule. Further, a constitution’s stable metarules for democratic competition ideally offer losing parties grounds to anticipate that they could have a shot again at political power through the ballot box, so that they need not resort to short-term violence. Parties hence have good reasons to keep in place constitutional protections (such as free speech rights) for opposition actors. This effect may be especially acute in federalized regimes, where a party might be both in power and in opposition at different levels of government.

Despite political parties’ important constitutional functions, the burgeoning literature in comparative constitutional studies contains relatively few contributions that systematically address parties’ constitutional functions, or that explore systematically the constitutional law of electoral design more generally. Writing of the British constitutional order, Vernon Bogdanor once noted that “it is perhaps because the law has been so late in recognizing political parties that constitutional lawyers and other writers on the constitution have taken insufficient note of the fact that parties are so central to our constitutional arrangements.”⁴ The same can be said more generally of comparative constitutional scholarship across the globe.

To be sure, some of these questions have received attention in the political science literature. Of particular importance is work on the ways in which constitutional design and the embedded choice of electoral systems shape party formation. Much of this work turns on a dichotomous distinction between first-past-the-post district elections and proportional representation, charting the effect of this binary choice on party systems⁵ and the consequent quality of democratic rule.⁶ Further, a debate dating back several decades has fleshed out in illuminating detail the relative merits of presidential, semi-presidential, parliamentary, and – more recently – semi-parliamentary systems.⁷ Yet much of this scholarly discussion has proceeded at a fairly high level of generality. It offers judgments on stark, binary choices based on analyses of acontextual big-N datasets. It is often penned by political scientists, who are not focused on the granular legal details of the constitutional metarules. While a useful starting point, this literature leaves unaddressed many questions of singular importance: What kinds of details of election and party system–design should be addressed in constitutions? When, in contrast, should those questions be left to the democratic process itself through statutes or regulation? When should rules be nationally uniform, and when can they be left to local or substate actors? How can the constitutional design of election systems (and hence of party systems) facilitate the larger ends of democratic rotation and endurance? What trade-offs exist in the design of electoral systems? How do choices concerning electoral systems and legislative-executive relations interact with choices in the design of central government or vertical choices of federalism? Are there independent, nonpartisan, judicial and “fourth branch” guarantor institutions (such as boundary commissions and electoral commissions) that should be assigned or allowed a place in electoral system design? If so, what is their role? Has the recent era of democratic backsliding, which has seen the employment of novel, subtler, and more incremental tools to undermine democracy, changed the answers to these questions? Finally, and most foundationally, is a transhistorical or historically durable, albeit modest, constitutional “theory” of elections and parties wise or possible? Or are those constitutional questions irredeemably anchored in contingent, local, and temporally situated dynamics?

This volume is an effort to address some of these, and related, puzzles. Its chapters examine the constitutional treatment of parties and elections both as a matter of theory and from the perspective of nationally and regionally specific historical and contemporary practice. To this end, it draws together a series of contributions from a diverse range of scholars working in distinct disciplines. This approach aims to take advantage of the perspectives distinctive to each discipline. Political scientists tend to treat political parties as their key object of study, while comparative constitutional lawyers have largely ignored them, preferring to focus on other institutional questions. What follows brings each perspective into conversation with the other. The characteristic disciplinary positivism of political scientists and the typical normative optimism of constitutional lawyers, we hope, create the dialectic necessary for significant intellectual progress on the topic.

This introduction frames the conceptual challenges of constitutional design that all the contributors to this volume grapple with, as well as providing some more general context. It canvasses, in particular, various models of constitutional treatment of elections and parties before introducing the contributions that follow.

1.1 Constitutions to Facilitate Politics

Like any competitive game, democratic contestation requires certain ground rules. One needs to know, at the most basic level, which offices of state that candidates can vie for; what qualifications are required to run and to vote for such offices; when elections are to be held; who and how votes are to be counted, and in what geographic units; and how potential disputes about eligibility and accuracy will be resolved. This list, to be sure, is illustrative, not exclusive. Yet with increasing frequency, many of these questions are resolved in constitutional text because of a sense among constitutional scholars and designers that a core constitutional function is to provide an enduring structure for democratic politics.

As a technology of specifically democratic governance, written constitutions are attractive because of two typical characteristics: their supremacy and their entrenchment. Constitutions are not just normatively superior to statutes, regulations, or executive actions. In many jurisdictions, they are the highest legal norm, superseding even international law. While most constitutions generate norms that are legal in character, not all constitutional rules are enforceable in a court. This quality of “non-justiciability,” as it is known in the technical literature, does not mean those rules cannot be both efficacious and also fundamental to the governance of the state.⁸ Rather, constitutionalization endows a normative legitimation to certain rules that may give them force even without a judicial sanction. To that end, appeals are often made to the hypothetical authority of a pouvoir constituent. In the political theology of democratic constitutions, the “people” have chosen this particular set of metarules. Supremacy facilitates legitimacy. Alternatively, constitutional supremacy can be defended on the ground that its abandonment would quickly give way to Hobbesian anarchy. On this view, supremacy is a response to the brute demands of realpolitik.

Entrenchment is the idea that constitutional norms are typically harder to amend than ordinary law. An organic document can hence take certain issues off the table or make some elements of the status quo ante more difficult to change.⁹ Removing issues from ordinary politics is useful when participants in a democracy believe that their core interests are at stake and such that leaving questions open to democratic revision creates intolerable uncertainty. This logic often underwrites the creation of constitutional rights. Alternatively, there may be a fear that leaving a question to change by simple legislative majorities will create perilous instability.¹⁰

Even where norms are not entrenched legally, political practice can engender de facto entrenchment by inducing beliefs and attitudes in key political actors. Consider here the Australian convention that the Governor General must assent to every Bill duly passed by both Houses of Parliament. This norm persists despite clear text to the contrary in section 58 of the Constitution, which leaves the matter to her discretion.¹¹ Such rules are politically entrenched in the sense that changing them would typically attract considerable public resistance and so demand the expenditure of significant political capital. As a result of this grip on political elites’ collective imagination, they can also provide a stable basis for political life.¹² Judicially-enforced constitutional norms are therefore not the sole means of stabilizing democratic constitutions.

At the same time, the idea of entrenchment should not be taken too literally or too far. It is easy to assume, especially in a field dominated by the study of US practice, that entrenchment in constitutional text will induce insurmountable difficulties in amending the constitution. The actual practice of constitutionalism around the world is quite different. The American model of rigid entrenchment is a distinct outlier. In general, constitutional entrenchment makes change difficult but not politically impossible. Its modal mechanism, as Adem Abebe (Chapter 6) explores, is one that frustrates the ruling party’s ability to make a constitutional change unilaterally, while allowing for constitutional change that enjoys broader political consensus from the ruling party as well as from sections of the opposition. On the other hand, in many jurisdictions there are likely to be many sub-constitutional arrangements that are not legally or constitutionally entrenched but are nonetheless very difficult or de facto impossible to change because of widely shared beliefs or attitudes.

Even without supremacy or an extreme form of entrenchment, constitutional texts can facilitate democratic politics by providing simple rules for coordination. Without some designation of who holds legislative power, for how long, and to what ends, citizens and residents cannot answer the basic question of what is valid law. Without knowing how a head of state is chosen, conflict over who has the role might be destabilizing. And without rules for the timing and composition for a first session of parliament, as well as other “first period” solutions,¹³ a polity might be trapped in an infinite regress trying to determine how to open official proceedings. Simple coordination on the names, nature, and timing of offices, as well as how officeholders are selected, is essential for any durable system of collective decision-making.

One might think that the basic democratic design of the political system should always be entrenched beyond revision through ordinary democratic processes. A government that can use its transient turn in power to completely change the rules of political competition would be sorely tempted to make its temporary grip on power more durable, even permanent. This creates worries in terms of both moral hazard and a related dynamic of adverse selection. Moral hazard occurs when sitting politicians try to change the rules to retain their office and insulate themselves from electoral challenge. The very possibility of such a move creates in turn the problem of adverse selection: In effect, the electoral system would create greater payoffs for anti-democrats than democrats, drawing bad actors into the system and squeezing out those who would rather compete on fair terms. Constitutionalizing basic elements of electoral choice beyond easy change helps tamp down on these complementary risks.

All this is to say that some choices need to be entrenched in a constitution to enable democratic politics. But which ones, to what extent, and in how much detail? The entrenchment frame suggests that constitutions might want to be very comprehensive in detailing the rules of political competition. A constitutional designer may think that the risk of leaving any metarules to the whims of self-interested politicians is simply not worth the candle.

Yet there are countervailing considerations, even setting aside the dubious idea that we can distinguish selfless constitutional designers and self-serving politicos. For one thing, we might not be sure we have the right answer at the time of constitution-making to some specific design question – say, the most desirable spatial dimensions of representational units. The “optimal” geographic units, or voting rules, might also change over time (say, as the polity enlarges to include groups previously denied recognition as full citizens, or as technological change leads to different patterns of economic activity). Entrenchment may make intuitive sense when it comes to fundamental rights on the theory that these embody enduring moral interests for which we think there are always limited justifications for infringement. But the same is rarely thought to be so when it comes to the design of political institutions. There may be no obviously decisive normative case for having, say, 200 or 300 members of the legislature, or holding elections on the Saturday of the third or fourth week of May. Nor have political scientists identified a universal “correct” answer to the choice between majoritarian, district-based elections, and multimember, proportional representation systems. Even the well-debated issue of whether presidentialism, parliamentarism, semi-presidentialism, or semi-parliamentarism is better remains up for grabs.¹⁴

For many of these questions, the “optimal” choice is likely to turn on some mix of the durable feature of the polity (say, geography, demography, history or education) and some of its more transient features. Even if there was an obvious normatively preferable option for a given context, it is possible that political constraints and path dependencies require a constitution’s framers to settle for sub-par alternatives. Or new circumstances – say the emergence of digital social media platforms – can render design choices that were normatively optimal in the past perverse or ineffective. As technological, social, or economic conditions change, the demos can (and, often, should) decide to modify its view, either through its elected representatives or by acting directly (through, say, a referendum) or in concert with its representatives. This counsels for a more limited extent of constitutionalization or a general open and viable route to constitutional amendment.

This is not the only possible pathway. A constitution can also create “variable speeds” of amendment difficulty. It can assign issues where less certainty obtains to a “lower” level of entrenchment than (say) fundamental rights.¹⁵ There are many design possibilities. But the core point here is that that maximal entrenchment is not always desirable. There is some unavoidable trade-off between the need for rigidity and flexibility in constitutional design.

One might also assume that only the very important issues ought to be put into the constitution. But again this is not obviously desirable. Very trivial issues, such as the choice of the first or the second Sunday of December for an election, might also be constitutionalized precisely because there is no strong reason for a particular choice. Further, the gains from the coordination enabled by a durable settlement (and the absence of conflict or opportunities for gamesmanship) may be greater than the risk of spending scarce political energy on it later on. As Louis Brandeis put it, sometimes it is more important that an issue be settled than it be settled correctly.¹⁶ By deciding a divisive issue once and for all, a constitutional designer preserves political attention for more important tasks down the road. To use a metaphor introduced by Hannah Arendt, constitutions set the furniture that we can then use to engage in democratic debate and decision.¹⁷

Very important issues, on the other hand, should be put into the constitution precisely because not doing so risks overwhelming and distorting ordinary politics. If each incoming government could completely overhaul critical elements of the political system unilaterally, this would create the aforementioned moral-hazard and adverse-selection problems. When the stakes of political competition are too high, there is little incentive to give up power upon electoral loss, such that electoral rotation is undermined. By lowering the political stakes, constitutionalization can help mitigate this problem, making democracy more durable. Of course, the status quo preserved by this strategy may well be sufficiently morally objectionable, such that any pro-democracy gain seems insufficient. The antebellum American experience with slavery exemplifies this possibility. Constitutional rigidity is only as valuable as the substantive merit of the system thereby preserved.

An intermediate path between constitutional silence and comprehensive constitutionalization of the metarules is to direct the legislature to adopt rules within a broad constitutionally prescribed framework without fully specifying their content. In this vein, so-called by-law clauses are frequently used for issues related to the political system and election.¹⁸ A constitution may use by-law clauses to establish a broad framework for elections and political parties, while leaving regulatory details to be determined by a specific (sub-constitutional) statute, which the legislature must enact post-ratification.¹⁹ Roughly one in ten constitutions in force, for example, leave the question of the immunity of legislators from either civil or criminal process to ordinary law. Similarly, 15 percent of constitutions in force explicitly leave to ordinary law the process for replacing individual legislators who are no longer able to serve. A further 14 percent leave legislator salaries to ordinary law.²⁰ Virtually every democratic country has a sub-constitutional electoral law that provides for the details of election management. And the enactment of these laws is often directed by the constitution itself. Kenya’s 2010 constitution includes a long list of provisions about parties, requiring that they have a national, non-ethnic character, and that they be internally democratic and refrain from violence. It then requires an electoral law to be passed by parliament within a year and that a code of conduct for political parties be produced by the Independent Electoral and Boundaries Commission.²¹

As a matter of observed practice, constitutional designers are increasingly choosing to weigh in on these questions. As Ginsburg and Versteeg show in Chapter 5, constitutional texts now say more and more about parties and elections, including by providing for new electoral institutions such as commissions and electoral management bodies as part of the “fourth branch” of regulatory (“guarantor”) institutions.²² Constitutions have also consolidated parties’ positions as central actors in the democratic process, even as their actual role as agents of representation is weakening in many democracies.²³ Even when using by-law clauses, constitutions increasingly limit the discretion of ordinary legislators by allowing judicial review of whether the legislature has created (say) an electoral commission with the constitutionally mandated degree of independence.²⁴ This paradox of greater substantive regulation in an era in which democracies overall have experienced a decline in performance reflects a core challenge of our moment.

1.2 Challenges

In his classic book, Party Government, the American political scientist E. E. Schattschneider wrote that “political parties created democracy and … modern democracy is unthinkable save in terms of parties.”²⁵ Elections, by definition, create winners and losers. A losing party must have some incentive to return to the fray rather than overturn the chessboard with protests or violence. It should be incentivized instead to jettison unpopular leaders and policies or to restrain figures that might be tempted to remain beyond their elected term. So nudged, well-functioning parties make an enduring system of representation possible, not least by creating a vehicle for carrying forward the interests of defeated politicians over the horizon of electoral loss.²⁶

In recent decades, however, some democracies have been suffering from what some have called a “crisis of representation,” involving a sharp decline in the strength of relationships between a party’s rank-and-file and its leaders. The classic account of such a decline of parties is the late Peter Mair’s analysis of European politics. Mair showed how Europe’s political elites remodeled themselves as a homogeneous professional class, withdrawing into state institutions that offered relative stability against fickle voters’ preferences. At the same time, non-democratic agencies and practices proliferated and gained credibility – not least among them the European Union itself.²⁷ Other accounts suggest that globalization has given us an economy with a mobile cosmopolitan elite pitted against a larger, far less wealthy group that is locally bound and economically precarious.²⁸ Under such conditions, traditional parties are no longer able or willing to screen out anti-institutionalist autocrats; nor are they able to provide effective representation. This has led to calls, by Frances McCall Rosenbluth and Ian Shapiro most prominently, for a concerted effort to revitalize parties.²⁹ Kim Lane Scheppele argues that “The Party’s Over,” at least the traditional mainstream ones, in that such parties are in decline as social and representative institutions even as they remain necessary for a healthy democracy.³⁰ She claims that the main reason for the collapse of mainstream parties is that the traditional left-right economic axis of politics has been superimposed by a cross-cutting cosmopolitan-nativist axis, but parties have not yet managed to reorganize themselves sufficiently along the new axes. In the terms Khaitan uses in Chapter 4, this development leaves salient voter types without a partisan home, thus breaching the “party system optimization principle” (one of four principles he argues constitutions should account for in relation to party systems).

A crisis of parties (or, better, of party systems) has metastasized in many countries into a more pluralist and less containable crisis of constitutional democracy. Instead of facilitating the predictable ebb and flow of representation that makes democracy possible, some parties have themselves become a vehicle of democratic backsliding.³¹ In some cases, they have repudiated the principle of rotation in office, leading to what some of us have called partisan degradation of democracy.³² Such parties aim to replace the constitutional entrenchment of uncertainty with the certainty of partisan entrenchment.

Party-led degradation can unfold in a couple of different ways. In some instances, an insurgent party takes advantage of a capsizing system of previously hegemonic actors to stake out a new claim and lock up political office. In other cases, a formerly mainstream party can be hijacked by demagogic populists, who then seek to fuse it with institutions of the state. We observe examples of both phenomena around the world today. In the first category, Jair Bolsonaro’s Social Liberal party in Brazil had won less than 1 percent of votes in the legislative contest prior to his successful presidential run. The Philippines’ Rodrigo Duterte rode to power in 2016 under the banner of the Partido Demokratiko Pilipino–Lakas ng Bayan (PDP-Laban), which held only one legislative seat at the time of his nomination. Examples of the second category include Donald Trump’s ascent within the Republican Party in the United States and (arguably) Narendra Modi’s rise in India’s BJP.³³ In Chapter 2, Richard Pildes and Samuel Issacharoff explicate how it is that minorities or individuals within parties can execute a take-over, facilitated by a series of poor regulatory choices.

In either of these scenarios, the ultimate result is that parties cannot restrain their anti-democratic leaders. The direction of representation is reversed. The party is either founded as, or becomes a vehicle of, concentrated personal power, rather than as a vehicle of deliberation and coordinated group action. Such dynamics are common in weak democracies generally.³⁴ But they are a relatively new feature in more established democracies. Once an autocratic leader personalizes the party, and hence makes contests over political office a matter of personality rather than policy, elections can take on an all-or-nothing quality. They can be framed as existential contests that can undermine the fabric of society itself. Democratic continuity will then be at stake.

How a constitutional order should respond to threats from political parties is therefore one of the difficult normative questions of our time. Answering it requires attention to the contributions of both positive social science and normative political theory. Even if one cannot get to an “ought” from an “is,” the real-world plausibility of adopting normative solutions should inform discussion. Addressing these challenges, it is inevitable that we seek to unearth lessons from existing models of democratic politics. It is to these models that we now turn.

1.3 Models of Constitutionalizing Elections and Parties

Over time, two different constitutional models have emerged for regulating politics with an eye to preserving its democratic cast. These could be called respectively first-order and second-order regulation. By “model,” we here mean a configuration of several institutional features that tend to co-occur, either by convention or by internal logic. These models can provide a template for jurisdictions to develop bespoke institutional designs.

A first-order – or command-and-control – regulatory model seeks to identify rogue parties that threaten democracy in order to ban or punish them. Second-order regulatory models, on the other hand, emphasize the importance of “background competitive structures” that shape decision-making, instead of seeking to police behavior directly through first-order commands directed at political actors.³⁵ A constitution organized along these lines seeks to craft these background competitive structures so as to make it difficult for anti-democratic parties to acquire or wield political power. These two models are not necessarily alternatives. Rather, they can exist simultaneously in the same jurisdiction.

First-order regulatory models can take one of two forms: (i) substantive regulation to protect democratic competition and democratic values and (ii) non-substantive regulation to ensure partisan non-domination. The most well-known model of substantively regulating anti-democratic parties is that of so-called militant democracy, developed by Karl Loewenstein in the 1930s.³⁶ Loewenstein was addressing the immediate problem of European fascist parties that had taken power through legal and democratic means during the interwar period. A modern encapsulation of their approach to democracy is the phrase “one man, one vote, one time.” To inoculate political systems against such threats, Lowenstein recommended prima facie undemocratic measures. These included bans of political parties with antidemocratic programs; hate speech restrictions; and “eternity clauses,” such as those discussed by Silvia Suteu in Chapter 12. All of these instruments are supposed to be deployed to protect the democratic basis of the state from constitutional amendments; and limits on political speech. (Some authors include term limits for high executive offices under the “militant democracy” rubric too.)³⁷ Several of Loewenstein’s ideas were adopted in Germany’s 1949 Basic Law as guarantees of a “free democratic order.”³⁸

The literature on militant democracy is large and growing, as is the popularity of its institutional recommendations.³⁹ Ginsburg and Versteeg, in Chapter 5, document the expansion of these clauses in constitutional texts. As a result of militant democracy’s diffusion, roughly 30 percent of constitutions in force either prohibit a specific party or a type of party program. In Africa, several countries ban parties of an ethnic or regional character.⁴⁰ Poland’s constitution bans parties “whose programs are based upon totalitarian methods and activities of Nazism, fascism and communism.”⁴¹ Militant democracy is not restricted to schemes in which it is subject to explicit constitutional embrace . Some constitutional courts have gotten into the action through the invention of the doctrine of “unconstitutional constitutional amendments.”⁴² Such judicial moves can be facilitated by eternity clauses, but their deployment frequently goes well beyond constitutional texts.⁴³ This phenomenon highlights a technocratic implication of militant democracy: some institution – typically the apex court – must be put in position to evaluate who can participate in democratic politics.

Militant democracy provisions have their advocates. But in recent years, the risks associated with their adoption have also come increasingly to the fore.⁴⁴ Militant democracy tools are attractive to prevent the rise of undemocratic parties. But once such a party has taken power, the tools of militant democracy can be used against those they were designed to protect. The Hun Sen regime in Cambodia, for example, has used hate speech and anti-terrorism laws to harass and ban political opposition. A concern with the “abusive” use of party-bans to undermine rather than uphold democracy is particularly important in an era of democratic backsliding.

One variant has been put forward by Jan-Werner Müller called “soft militant democracy.” This does not rely on the heavy-handed tool of party ban. Instead, it limits undemocratic parties’ effective reach by making “life for the party difficult.”⁴⁵ He offers the example of Israel, which does not ban racist parties, or those that deny the State’s “Jewish and Democratic” character. Rather, Israel prohibits such groups from registering for Knesset elections. India also does not allow certain kinds of speech appealing to religion and ethnicity during election campaigns, even though it is generally legal at other times.⁴⁶ (Of course, it might be questioned whether either Israeli or Indian law have been successful in excluding anti-system parties from political office.) Denying parties access to public financing or broadcasting time is another option. Application of anti-discrimination norms to political parties is yet another such “soft” substantive intervention. In Germany, in a recent case concerning the National Democratic Party, a far-right group with an anti-democratic program, the Constitutional Court refused to ban the party but permitted some less dramatic measures to be taken against it.⁴⁷ In this way, a democratic system allows undemocratic speech, and so avoids the risks and potential blowback of censorship, while penalizing politicians who refuse to moderate their appeals in conformity with inclusive democratic norms.

It could also be argued that intraparty democracy helps to protect democratic governance more broadly. Some constitutions, like Germany’s, require that political parties maintain internally democratic processes for leadership and candidate selection.⁴⁸ The argument is that internal democracy is a device for discouraging personalism or other autocratic forms of organization. It could also ensure epistemic openness, so parties change policy positions as their ordinary members change their minds. At the same time, it is unclear as an empirical matter whether a substantive legal requirement of intraparty democracy is indeed causally related to the creation of a more democratic polity. One could even postulate, instead, that stringent intra-democracy rules within parties may damage the quality of a democracy. More extreme base voters who tend to vote in intraparty elections are often more likely to prefer extremist candidates than the party establishment, driving a party toward anti-democratic stances.⁴⁹ The effectiveness of a legal requirement of intraparty democracy as an effective tool of soft militant democracy is therefore yet to be established.

All militant democracy models of constitutionalizing power ultimately depend on the availability of “neutral” observers to identify and discipline undemocratic parties. This has given rise to critique of militant democracy as elitist.⁵⁰ Stopp and Popp-Madsen argue that militant democratic tools rest upon a “depoliticizing, elitist, and exclusionary understanding of politics, relying on handing power to unelected and potentially unaccountable technocrats or jurists.”⁵¹ This critique has special force in an era of populism. Technocratic and elite institutions are suspect. The use of tools viewed as elitist to discipline popular movements risks adding fuel to the fire of antidemocratic populism. Theorizing in this vein, Tom Daly and Brian Christopher Jones in Chapter 7 argue that party bans can sometimes be counterproductive and potentially make the banned party even more appealing, since the party can style itself a martyr or can reincarnate itself with a new label, offering a brand of politics that they call “far-right lite.”

Could we avoid or reduce the danger parties pose to democracy without requiring a (typically unelected) body to reach controversial judgments about which party and which candidate is sufficiently democratic? This is what the non-substantive first-order regulation seeks to achieve by focusing on securing partisan non-domination – that is, a system in which no party or parties have systemic or structural advantages that allow them to dominate other parties. Although there is considerable debate on the definition of what constitutes a dominant party or a dominant party system, the literature tends to reserve the term for a party that wins a (sufficiently large) threshold number of seats/votes (usually a majority or a supermajority of seats and a plurality or majority of votes), over a sufficiently long and unbroken period of time (such as three continuous election cycles), with a divided, dispersed, or negligible opposition.⁵² Systems in which one or more parties dominate others may be classified based on (i) the number of dominant parties, (ii) the extent of their domination over the political system, and (iii) the mode of political domination.

Along the first dimension, depending on the number of dominant parties, a system may be a monopoly (i.e., dominated by one party), a duopoly (a domination of two parties over all others), or an oligopoly (domination of a few parties over all others). Along the second dimension, systems can vary based on the degree to which they permit or enable partisan domination. Indeed, the most extreme degree of domination is a total prohibition on the existence of opposition parties (i.e., a de facto and also de jure one-party system). Even when this is not the case, a lesser degree of domination can be achieved if law or official institutions sufficiently favor the dominant party, whether directly or indirectly. This may be done, for example, by affording some parties better or sole access to state or corporate funds,⁵³ facilitating or tolerating its use of civil or criminal regulation against political rivals for partisan reasons, partisan capture of unelected offices of the state, and so on. Cutting across these possibilities is a third dimension – that is, the modalities of domination. This concerns whether hegemonic parties secure their domination by de jure and de facto means (where the domination is baked into the laws of the country and coercively enforced in practice) or only de facto tools (where the domination exists in practice but is not required by the law).⁵⁴

A de jure (and de facto) monopoly arises where only one party is allowed to exist. This is the most extreme form of partisan domination. In other monopolies, smaller allied parties may be allowed to operate but not allowed to hold or compete for effective political power. The People’s Republic of China, for example, has eight parties. But the “leading role” of the Communist Party is constitutionalized and has always been beyond question. Today, only five countries, all communist, retain a formally one-party or hegemonic party model. The favored “vanguard” parties are the Worker’s Party of Korea, the Lao People’s Revolutionary Party, and the Communist Parties of China, Cuba, and Vietnam. Historically, the single-party model is not uncommon. It is found in fifty-two other historical constitutions, or roughly 8 percent of all texts ever written. Typically, the favored party is a national liberation front that led the struggle for independence from colonialism or civil war.⁵⁵

Sometimes a single-party monopoly arises because of a breakdown in the initial transition to democracy. This is common when a national liberation movement itself becomes the new dominant party. The ZANU-PF in Zimbabwe, and to a lesser extent the African National Congress in South Africa, became dominant forces that locked up politics over time. Both eventually became corrupt. Sujit Choudhry has identified some characteristic tendencies of such systems.⁵⁶ These include the use of public resources to distort electoral competition, attempts to fragment opposition parties to minimize credible alternatives, and the subordination of parliamentary members to extra-parliamentary leadership. He argues that courts in such conditions should actively seek to constrain such tendencies . In the US context, Richard Pildes and Samuel Issacharoff have characterized the United States as an effective duopoly. The two dominant parties have created de jure structural barriers that prevent the successful emergence of third parties.⁵⁷ Of course, what one makes of a monopoly or a duopoly turns on the degree of intraparty democratic fluidity, among other matters.

To address partisan domination, non-substantive first-order regulation often focuses on preventing what Issacharoff and Pildes call “partisan lock-ups.”⁵⁸ This directs attention to the architecture of democratic competition, not the substantive content of political programs or political speech. Such machinery must be insulated from party influence – not as a means to institutionalize substantive choices about who can compete but rather to prevent capture. A common way of ensuring non-domination is an independent electoral commission that can guarantee free and fair elections.⁵⁹ Another alternative, explored by Aziz Huq in Chapter 11, is the exercise of judicial review over election laws and election results. Huq’s chapter examines the choice between courts and fourth-branch bodies as supervisors of the democratic process.

Concern over partisan capture can also motivate rules about campaign finance and advertising. For example, the Israeli Supreme Court struck down an amendment to the country’s election law that reduced advertising time for all parties, while increasing it for incumbent Knesset members, on the ground that the change advantaged those parties that had already succeeded in securing seats.⁶⁰ Similarly, the Indian constitution was amended to make floor-crossing by legislators difficult, by requiring party-switching by a minimum threshold of one-third of a party members for it to be legitimate, in the face of experience with financially dominant parties simply bribing opposition members to change affiliation, so as to make up any shortfall in the dominant group’s legislative majority.⁶¹

Just because constitutions do not substantively regulate parties directly does not necessarily mean they are indifferent to the role of parties in upholding and threatening democracy. For instance, second-order regulatory models do not seek to constrain behavior of parties by direct command-and-control methods. Rather, they seek to organize the institutional ecosystem of politics to reward or facilitate good behavior while punishing or frustrating undesirable conduct. Of course, many of the institutional choices embedded in a constitution shape the subsequent behavior of political actors. A representative democracy must choose some kind of electoral system. Its constitution cannot realistically be wholly agnostic on the matter. Yet, as we have noted, the distinctive features of any system necessarily shape the behavior of politicians differently. Even a system that wants to adopt direct (rather than representative) democracy will need to make institutional choices concerning selection of the executive, the mode of decision-making, and so on.⁶² All of these institutional design choices influence the nature of the party system that comes about. The question is not so much whether to have second-order regulation but what type of second-order regulation is likely to encourage a party system conducive to a durable democratic order.

The US constitution is an early example of attempted second-order regulation of the party system. The American founders had a “keen terror” of parties, as interests that were partial rather than concerned with the country as a whole.⁶³ Madison saw the institution of representative (as opposed to direct) democracy in a large but federated country as the key means of preventing a “faction” emerging that would be “adverse to the rights of other citizens, or to the permanent and aggregate interests of the community.”⁶⁴ He claimed that, first, the principle of representation would ensure that “if a faction consists of less than a majority, relief is supplied by the republican principle, which enables the majority to defeat its sinister views by regular vote.”⁶⁵ Second, he reasoned, the process of representation itself would reduce the risk of factionalism inherent in a direct or “pure democracy [that is] … a society consisting of a small number of citizens, who assemble and administer the government in person, can admit of no cure for the mischiefs of faction.”⁶⁶ In a representative democracy in a large enough country, Madison reasoned, “it will be more difficult for unworthy candidates to practice with success the vicious arts by which elections are too often carried; and the suffrages of the people being more free, will be more likely to centre in men who possess the most attractive merit and the most diffusive and established characters.”⁶⁷ For Madison, the mechanisms of selecting representatives for the national government, coupled with the diluting effect of the large geographic scope of the new United States, offered second-order regulatory mechanisms that disarmed prophylactically “factions,” or anti-system parties.

Madison admitted that there were faults in this system. He noted that the large size of the country might “render the representatives too little acquainted with all their local circumstances and lesser interests.” But this cost could be mitigated by a federal structure, he contended, such that “the great and aggregate interests being referred to the national, the local and particular to the State legislatures.”⁶⁸

Drafters’ imaginations are inevitably limited. The Madisonian hope that mere representative government in a large country would suffice to prevent the emergence of antisystemic factions has plainly not materialized in the United States. American elections are managed at the state level, and the federal courts have been historically reluctant to constitutionalize them.⁶⁹ Only in the 1960s did the national government and the federal (i.e., national) courts move to address minority voter suppression in Southern states, and much of that work has been undone by the US Supreme Court in recent years through its reading down of the Voting Rights Act of 1965.⁷⁰ Partisan gerrymandering remains widespread and judicially tolerated.⁷¹ In effect, Madison’s dream of candidate filtration through careful institutional design has collapsed as a consequence of partisan workarounds.

Yet Madison’s was a relatively simplistic second-order regulatory model. Far more sophisticated tools have since been experimented with or proposed in other jurisdictions. These include:

• – compulsory voting rules, including in primaries, to ensure that parties are forced to account for the interests of those voters least likely to vote;⁷²

• – ranked-choice or preferential voting systems that incentivize parties to build broad-based coalitions in order to win enough second-ranked votes to win elections,⁷³ combined at times with multimember districting;⁷⁴

• – independent nonpartisan election management bodies and boundary commissions;⁷⁵

• – bicameralism to reduce ruling-party domination (along with anti-deadlock rules);⁷⁶

• – federalism to ensure multiple centers of partisan power;⁷⁷

• – parliamentarism or semi-parliamentarism, which tend to empower parties relative to their leaders, rather than presidentialism, in which parties tend to be weaker;⁷⁸

• – staggered elections to mitigate the impact of partisan “waves”;⁷⁹

• – a regime of opposition rights and powers that resists winner-take-all dynamics;⁸⁰ and

• – effective campaign finance rules to reduce the impact of the monetary might of the ruling party.⁸¹

This list, while long and no doubt controversial, is just illustrative rather than exhaustive. These and other second-order regulatory models are informed by the belief that party-systems and institutional design are mutually constitutive. Even as parties scramble institutional rules, institutional arrangements (along with political culture, context, and path dependencies) influence the arc of a party system, as well as the number and nature of its parties.

Before we end this discussion on the two models for regulating parties, it is worth noting the emergence of yet another approach to the problem that parties pose to democracy. Unlike the two regulatory models discussed above, this “direct democracy” approach takes parties to be an irredeemable threat to democratic order. It seeks to bypass them, either for a set of crucial decisions or for all political decisions. This back-to-the-people agenda has found some influential scholarly backers in recent years.⁸² They envisage a party-less utopia characterized by direct political participation of citizens via mechanisms such as sortition, referendums, citizens’ assemblies, and virtual voting. While these approaches recognize the dangers that current parties can pose to democracy, they often ignore the irreplaceable service political associations provide in facilitating democracy under conditions of imperfect information and limited resources.⁸³ It is possible, as Madison feared, that the anti-democratic risks are greater, rather than lesser, in a direct democracy, were such a thing even possible and sustainable over time in any reasonably large state.

***

The chapters in this volume address these challenges in myriad ways. The volume begins with consideration of constitutional theory writ large. Fragmented and concentrated power is the theme of Chapter 2 by Samuel Issacharoff and Richard Pildes, which revisits some of the terrain of their important 1998 article “Politics as Markets.”⁸⁴ Democracy, on their view, requires a strategy to avoid either of these tail outcomes. Starting with a history of recent electoral regulations in the United States, they show how reforms and court decisions have weakened party leadership. The result is a fragmented party landscape in which minority candidates can with some regularity triumph in general elections. They pithily label this phenomenon the “tyranny of the minority over the majority” and explain why and how it became a central threat to democratic legitimacy. They canvass and assess some recent proposals to improve the situation, including ranked-choice voting, and repealing “sore-loser” laws that empower more extreme primary electorates to screen out more centrist potential winners.

In Chapter 3, Rosalind Dixon and David Landau draw on their previous, influential scholarship concerning the concept of “abusive constitutional borrowing,” Here, they catalog a whole range of constitutional provisions such as voting rights, electoral quota, and establishment of dedicated electoral machinery, among others. They further examine the discourse and practice of election monitoring. In each case, they demonstrate how pro-democratic norms and practices can be repurposed toward the undermining of democracy. Canvassing the causes of abusive practices, they go on to propose a series of solutions to reinforce constitutional democracy.

In Chapter 4, Khaitan lays out his own specifically constitutional theory of parties. Drawing on what he calls the Janus-faced character of parties as entities that mediate the public-private divide in complex ways, he shows how well-functioning parties can reduce information costs for voters and facilitate politics. He then draws upon this ideal type to provide a framework for thinking about pathological parties. He derives from this four principles of constitutional design. Constitutions, on his view, should provide maximum autonomy for parties within constraints necessitated by the exercise of state power; they should seek to structure a party system that represents major groups of voters; they should separate party and state; and they should discourage factionalization of parties.

The volume then turns to empirical inquiries into the same issues. Tom Ginsburg and Mila Versteeg (Chapter 5) document a massive increase in the propensity of written constitutions to regulate elections, voting, and parties, along with instruments of direct democracy, over the course of the twentieth century. The steepest increase has been in provisions associated with parties. Employing statistical analysis, they show that at least some of these provisions are negatively associated with democracy. Pairing two recent case studies in Kenya and Thailand, moreover, they show how constitutionalization can be a double-edged sword. In Kenya, regulatory provisions were adopted to empower democratic politics and prevent lock-ups. But under Thailand’s authoritarian-drafted document, militant democracy provisions were weaponized to undermine new parties.

The volume then turns to a series of investigations into more specific legal or constitutional means of safeguarding democratic stability. A number of these chapters offer case studies from generally under-examined regions. Focusing on Africa, for example, Adem Abebe (Chapter 6) surveys the continent’s approaches to regulation electoral systems and political parties, with a focus on how they have performed during the recent era of democratic backsliding. He develops the critical role of opposition parties in a democracy and the importance of constitutions in protecting their ability to function as a prophylactic against the winner-take-all politics that dominate the region, by drawing upon a rich set of examples that are not usually accounted for in the literature.

Tom Gerald Daly and Brian Jones (Chapter 7) look closely at political parties as threats to liberal democracies. While political parties are central to democracy, their very importance can therefore easily pose a threat to such a system given their unique positions, making it ever more important that they are properly governed. Usefully taxonomizing traditional approaches to dealing with unconstitutional parties as being legal, constitutional, or political, the authors explain how they are all inadequate to the current threat. They go on to consider novel approaches that leverage multiple branches of government, including quasi-judicial entities, to ensure the decision-making power over political parties does not fall to one sole person or entity. They also examine emerging international mechanisms and stronger controls on electoral manipulation, concluding that more international involvement and cooperation will be necessary to make use of these options.

In Chapter 8, Yasmin Dawood examines the value of a constitutionalized right to vote. She demonstrates that this right is best understood as a multidimensional and complex one, involving plural constitutional and legal elements. Yet these aspects are also dependent on a broader institutional infrastructure. While recent literature has suggested that the right to vote is not, on average, efficacious in the sense of leading to greater implementation, Dawood argues that it is normatively valuable nonetheless. At a minimum, the right has an expressive function. Even if this can sometimes be abused by autocrats, the right to vote remains a core feature of constitutional democracy.

In Chapter 9, Stephen Gardbaum examines the role of different types of political systems in addressing another set of trade-offs. He begins by stipulating democracies’ need for effective, responsive, stable, and accountable government, as well as legislative bodies that are both representative and deliberative. He argues that not all of these values can be maximized at the same time. Regime types in the form of presidential, parliamentary, and semi-presidential systems provide different balances, conditioned by the operation of party systems. Here, he looks not just at the number of parties but whether they suffer the pathologies of being “polarized, hyper-partisan, and/or fragmented.” Such pathologies undermine the ability to achieve the core values. He also considers recent proposals for “semi-parliamentarism” as a solution to the problems, arguing that its core “solution” of modified bicameralism is more generally feasible than proponents have suggested. By introducing some of these design features into presidential and semi-presidential forms of government, Gaudbaum suggests, constitutional design might begin to ameliorate some of the pathologies of contemporary government associated with both fragmented and concentrated power.

Turning to specific institutions used to safeguard the quality of the democratic process, Aziz Huq (Chapter 10) analyzes the role that courts can play in maintaining or undermining the democratic character of a regime. He canvasses the different roles that apex courts in particular (as opposed to the wider array of judicial bodies lower down in a hierarchy of adjudicative instruments) have been called upon to play in the constitutional law of elections and parties with increasing regularity. He considers further whether other nonpartisan bodies, such as election commissions, can play the same function at lesser costs. Comparing the emergence of fourth-branch bodies to the possibilities of judicial review to preserve democratic stability, he expresses some skepticism about the temptation to favor either courts or fourth-branch as a categorical matter, as opposed to making more situated, contextually dependent judgments.

In Chapter 11, Elizabeth Reese offers case studies from Native American tribal nations in the United States. Their constitutional governance has evolved from a complex mix of traditional institutions, federally incentivized formal documents and more recent rounds of reforms and amendments driven by the tribes themselves. In the case of the Cherokee Nation, for example, contestation over who holds power has led to more than one constitutional crisis, but constitutional design has interestingly led to an escape from political gridlock. In the case of the Citizen Potawatomi tribe, the adoption of at-large voting by those outside the physical state of Oklahoma led to a change in dynamics. In both cases, constitutional institutions have managed to support continuity of political identify in conditions of ongoing colonial confrontation but also to resolve specific institutional challenges of governance that recur in other settings. Reese’s chapter also makes a general case for a more central role for Native American constitutions in the field of comparative constitutional law.

In Chapter 12, Silvia Suteu takes up a quite different mechanism that is often the focus of democratic preservation, namely unamendable “eternity” clauses that stipulate the permanence of the core commitments of the constitution. She examines how these devices can shape the field of electoral competition, using two examples: party bans and the protection of parliamentary mandates. In her view, a normative evaluation of those tools is inseparable from the political context in which decisions are being made. Questions of the party system, the electoral balance of power, and the availability of courts and other institutions are critical for any normative assessment. This nuanced approach is a helpful corrective to the literature, which tends to be generally optimistic about eternity clauses.

In Chapter 13, on this theme of specific institutional choice, Yvonne Tew examines another “neutral” institution that has an effect on political competition: the constitutional monarchy of Malaysia. Although the crown is nominally above politics, there have been recent moments in which the monarch has played a decisive role in government formation and in both undergirding and undermining democratic structure. The role of neutral third parties has been underexplored in the literature to date in reference to courts and fourth-branch bodies. Tew’s intervention usefully opens up a vista of fresh possibilities.

1.4 Conclusion

Together, the chapters in this volume offer a comprehensive, interdisciplinary account of the challenges of constitutional regulation of parties and elections. Any constitutional system contains a whole series of metarules that structure politics, most directly the rules regulating parties and their activities, and the electoral system which indirectly shapes the party system. In addition, there is also a structure of third-party decision-makers to draw boundaries, administer voting and resolve disputes, and a set of rights that express values and condition mobilization. The very complexity of the interaction of all these elements makes it difficult to draw conclusions strong predictive judgments about the effects of any particular institutional feature on its own. There are often too many moving parts to be confident of what one marginal change will do.

Further, the enduring analytic and conceptual challenges raised by Madison and his colleagues two centuries ago remain as pressing as ever. Self-interested political actors will endeavor to manipulate rules to their advantage, and so an enduring constitutional settlement requires not only some attention to deep structures but also to new institutional innovations that address the ever-changing challenges of democratic societies. This volume, we hope, goes some way toward showing how recent experience and theory can help us understand, and perhaps try to meet, those challenges at a moment when democracies around the globe are under particular strain.