Part II - Constitutionalization

4 Political Parties in Constitutional Theory

Tarunabh Khaitan

4.1 Introduction

Political parties appear to be in crisis. Lazy clichés in popular culture routinely stereotype politicians and parties as a group to be self-serving, elitist, and corrupt. The recent wave of democratic deconsolidation in several established democracies has been accompanied – perhaps caused – by the collapse, authoritarian takeover, or external capture of mainstream political parties, the partisan capture of state institutions, and a rise in hyper-nationalistic and exclusionary partisan rhetoric.¹ This chapter forms part of a larger ongoing project in defence of parties, politicians, and politics, one that is examining relatively ignored non-judicial phenomena in constitutional scholarship, such as incremental party-state fusion,² rights and powers of opposition parties,³ second chambers and semi-parliamentarism,⁴ political directives as constitutional norms,⁵ anti-plutocratic constitutional norms,⁶ and the so-called fourth-branch guarantor institutions.⁷

Political parties and party systems have long been the central institution for analysis in political science. A significant part of the political science literature is taxonomical, based on relationship between parties, their leaders, their members, the electorate, the state, and electoral systems.⁸ This literature has taught us about cadre and mass-based parties,⁹ catch-all parties,¹⁰ electoral professional parties,¹¹ cartel parties,¹² and market-oriented parties.¹³ Some of this literature has also looked at the relationship between party systems and certain institutional arrangements – such as Duverger’s law that a first-past-the-post system is likely to result in a two-party system,¹⁴ and the correlation of strong and weak parties with parliamentary and presidential systems respectively.¹⁵ The first set of scholarship shows just how dynamic, adaptable, and indispensable parties have been to numerous challenges democracy has thrown their way. The latter set shows – albeit at a very high degree of generality and typically in a monocausal fashion – how parties respond not only to the broader politico-social and economic developments but also to certain features of constitutional design. Other disciplines that have examined parties closely, although not to the same degree as political science, include political history¹⁶ and political theory.¹⁷

While political parties have long been a central object of study in political science, Anglophone constitutional law and theory scholars have, until recently, largely ignored this key democratic institution.¹⁸ The little attention constitutional scholars have given to parties either concern the jurisdictionally specific legal regulation of parties and elections¹⁹ or (rarely) the impact of electoral or party systems on specific policy outcomes.²⁰ Comparative constitutional scholars and constitutional theorists have largely been silent due to the influence of the American and the British constitutional traditions on the field, which, unlike their European continental counterparts, are largely silent on political parties. This silence is mainly a feature of big-C constitutional codes in the Anglophone world. Small-c constitutional statutes, conventions, and judicial precedents in these states do, admittedly, engage extensively with political parties.²¹ But the large-C textual silence is nonetheless indicative of the low level of salience this key constitutional institution has been given, both in constitutional practice and constitutional scholarship. Madison, echoing Rousseau,²² was famously hostile to the emergence of political parties as ‘actuated by some common impulse of passion, or of interest, adverse to the rights of other citizens, or to the permanent and aggregate interests of the community’.²³ That he would go on to found one of the two first political parties in the United States is, of course, another matter.²⁴ Contrast Madison’s ideological hostility to parties with Burke’s more affirmative opinion that ‘a party is a body of men united for promoting by their joint endeavours the national interest upon some particular principle in which they are all agreed’.²⁵ Sadly, this Burkean optimism about parties fails to inform constitutional theory despite the British influence on the field because structural features of the organic constitutional development in the United Kingdom never produced a big-C constitutional code.

Big-C codes typically design key state institutions in a democracy. Parcelling off considerations about political parties to small-c statutes and conventions has the effect that the shape of the party system becomes an afterthought, left to be regulated by small-c statutes while taking the design of key state institutions as a given. As this chapter argues, however, bringing parties to the forefront of the constitutional imagination has very important implications for how we ought to think of fundamental institutions and offices of the state. Furthermore, big-C constitutional change tends to require the buy-in of opposition parties, whereas small-c changes can usually be made by the ruling party/coalition alone. It is simply bad design to let one of the competing players unilaterally change the rules of the game. It is no surprise that continental big-C codes, led by Germany after the Second World War, are far more explicit in their attention to parties and their relationship with democracy. Even so, the Anglophone silence is mimicked in comparative constitutional studies scholarship, dominated as it is by American constitutional discourses, including the latter’s attendant pathologies.²⁶ It is almost impossible to properly understand the functioning of different institutional arrangements without close attention to the party system in which they operate.²⁷ Constitutional scholarship that confines itself to institutional analysis alone, without understanding how they are conditioned by political parties, is looking at a seriously distorted picture of constitutional practice.

To some extent, however, big-C constitutional silence on political parties is also explained by the serious conceptual and functional challenges they pose to (liberal democratic) constitutionalism. Conceptually, as the primary holders of real public power in most democracies (at least when in government), political parties should be apt for constitutional regulation. But, as the chief vehicle for organising democratic will, the level of autonomy that political parties enjoy is one of the markers of the health of a democracy. Like private clubs, their membership is voluntary and often informal. The control they have over their members is limited.²⁸ Their role as an intermediary between the state and its people make them insufficiently public to be burdened with the normal duties of state institutions, and inadequately private to be treated on par with rights-bearing citizens.²⁹ Functionally, parties organise popular will and thereby make democratic functioning possible. But as wielders of concentrated state power, they have also emerged as a key threat to liberal democracy (alongside the military and the very rich).³⁰ Constitutions therefore need to walk a fine line between preserving their ability to organise and channel popular will while reducing the threat they pose to democratic governance. Scheppele is right in exhorting constitutional democracies ‘to find ways to support and regulate’ parties, for – she says – ‘the secret to democratic self-preservation may rest in the realization that the party isn’t over yet’.³¹

In this chapter, Section 4.2 will first provide an idealised functional account of political parties and party systems. The idealised account presented in Section 4.2 clarifies what parties do when they function as they should function in a healthy party system of a representative democracy. As such, this is a conceptual and normative, rather than a historical, project. The normative account is acontextual, but the principles offered are pitched at a sufficiently high level of generality to be worth taking into consideration when making all-things-considered decisions in a wide variety of contexts, but not so high and abstract that anything goes. I will argue that parties are difficult to regulate constitutionally because of their Janus-faced public-private character. The key function they perform, when functioning as they ought to function, is to facilitate a mutually responsive relationship between public policy and popular opinion by acting as an intermediary between a state and its people. When they perform this function effectively, political parties significantly reduce four key information and transaction costs that would otherwise make democratic governance impossible: political participation costs, voters’ information costs, policy packaging costs, and ally prediction costs. For critics of this methodological approach who might be worried that an idealised account is too far removed from how parties in fact function in the real world, the practical payoff is that it helps us identify pathological parties and party-systems and diagnose their particular ills. Even as pathological parties abound, especially in recent years, relatively healthy parties have existed and continue to exist around the world. Furthermore, if we started discounting norms simply because they may never be fully realised, our moral universe would become seriously impoverished indeed. I will therefore use this idealised account to ground four principles that can help diagnose the health of a party system and, consequently, principles that constitutions should seek to optimise in relation to political parties and party systems, with a view to avoiding, diagnosing, curing, or mitigating systemic pathologies.

Thus, I argue in Section 4.3 that democratic states (and their constitutions) should respect and optimise four distinct, and sometimes conflicting, political principles in relation to political parties:

1. i. They should guarantee maximum autonomy for the formation, organisation, and operation of political parties, moderated by the restrictions necessitated by their purpose of winning (a share in) state power (for fixed terms) in competitive elections by acting as intermediaries between the state and its people (the ‘purposive autonomy principle’).

2. ii. They should try to optimise the party system such that the total number of serious political parties is large enough to broadly represent every major ‘voter type’, but not so large that the information costs on judicious voters are too high (the ‘party system optimality principle’).

3. iii. They should ensure a separation of parties and the state (the ‘party-state separation principle’).

4. iv. They should discourage the factionalisation of political parties (the ‘anti-faction principle’).

These political principles are drawn from the value of democracy itself. They are likely to bring real world political parties and party systems closer to their idealised form as described in Section 4.2, thereby improving and deepening democratic governance. As such, they should – alongside other relevant political and constitutional norms – inform fundamental constitutional design choices. Retrofitting the regulation of parties through the small-c constitution after key design choices have already been made in the big-C code is, therefore, a mistake. Big-C constitutional silence on parties is as much a regulatory choice as any other and carries significant risks of unintended consequences. In other words, big-C constitutions – as the chief organisational tool for public power in democracies – simply do not have the option of remaining agnostic about the nature and functioning of political parties. The question is not so much whether to regulate parties but why and how.

Another important caveat – especially to the more lawyerly readers – is that not every desirable norm can be converted into law, nor should it be. Some of the implications of the arguments in this chapter are legal, many are matters of political morality. Saying that a norm is a matter of political, rather than legal, morality is not to suggest that there is nothing we can do about its enforceability. Constitutions can make it more likely that certain political norms will be complied with – these are also design choices, they simply transcend the law. In particular, different electoral systems, legislative design, and executive-legislative relations fare differently with respect to the principles defended in Section 4.3. These matters of constitutional architecture have serious implications for the health of a regime’s parties and its party system and are largely settled outside the courts. The aims of this chapter are largely theoretical and conceptual. A more detailed account of what practical implications might follow if constitutions are to take these principles seriously have been canvassed in a different article.³²

4.2 Parties: An (Idealised) Functional Account

In this section, I will argue that political parties, when they function as political parties ought to function, perform the key democratic function of acting as an intermediary between the state and its people in a representative democracy. Two particular features make this intermediary function of parties unique: the bidirectionality of their intermediation and the plenary character of political parties. A party system with healthy functional parties incurs lower levels of four key information and transaction costs: political participation costs, voters’ information costs, policy packaging costs, and ally prediction costs. Keeping these costs low makes a representative democracy viable as a mode of governance. Clear recognition of these features of a healthy party system with functional parties allows us to distinguish them from pathological party systems and diseased parties.

4.2.1 Parties as Intermediaries

A democracy, by definition, requires the rule of (all) the people who constitute a polity. Systems that systemically engage in comprehensive exclusion or suppression of discrete groups of voters, therefore, are diseased democracies, or not democracies at all. If democracy was the only legitimacy criterion for a political regime, we would vest all decision-making powers directly in the people. But democracy is not the only value we care about. Most of us would consider a law requiring the enslavement or genocide of any group thoroughly illegitimate, howsoever democratic its pedigree might be. Less dramatically, rule of law (legality) values – such as fairness, consistency, efficiency, impartiality, non-retroactivity, and generality – invite us to distinguish between rule-making and rule-application.³³ We will put the enduring scholarly debates about the legitimacy, sharpness, and feasibility of the distinction aside for the purposes of this chapter, but note that constitutional practice in liberal democracies has, broadly, come to accept that while the rule-making aspect of political power must be largely vested in institutions that represent the people, rule-application is, on the whole, best left to experts of some description. With regard to rule-making, there are several ways of securing the normative ideal of a popular regime (i.e., a regime that is ‘of the people’ but not necessarily one that is all the rage with the electorate). Direct democracy usually imposes high transaction costs, especially in large and complex societies, although even simple democracies have need for at least some political offices (and committees) that can represent and act in the name of the people and procedures that determine how they may act validly.³⁴ Ancient Athenians used sortition – selection by lot – to fill political offices; in our times, election by universal franchise is the more common method. Political parties have emerged as the main – and arguably indispensable – vehicle for facilitating representative democratic elections in even moderately sized complex modern states.³⁵ What follows is an idealised functional account of political parties in a democracy. Real-world parties will no doubt fall short of these ideals, but the fact that we may never fully achieve an ideal is not a reason to give up on trying to realise them as far as possible.

The chief function of political parties is to act as intermediaries between the state and its people. This claim does not presuppose a specific type of party organisation: I use the term ‘intermediary’ in a loose sense here to be compatible with a varying range of intensity in the relationship between the party and the people.³⁶ What matters is that parties have a threshold level of communicative relationship with the people. State officers and institutions are typically too removed from the people to access popular opinions directly, and ordinary civil society organisations are usually too removed from the state to influence state policies. Exceptions no doubt exist: in systems in which individual legislators represent sufficiently small constituencies, they can have a direct relationship with their constituents; similarly, many policy influencers, such as wealthy corporations, lobbyists, thinktanks, and powerful media houses, often have a significant influence on state policy. Yet political parties are a very special type of intermediary between the state and its people for two reasons: the bidirectionality and the plenary character of their intermediary function.

4.2.1.1 Bidirectionality of Parties

Mediation by parties is bidirectional, in as much as they simultaneously perform both functions of accessing popular opinion and shaping state policy. They are embedded in the structures and institutions of the state but also (at least ideally) have direct access to the people. This simultaneity is essential to the democratic legitimation that parties alone can provide to rule-making state institutions and offices. In general, the state functions through offices and institutions: these are modes of corporate action that are defined by a measure of formalisation of their processes, purposes, and modes of operation. This formalisation is typically necessary for satisfying various virtues (rightly) associated with the state: impartiality, rationality, fairness, legality, and so on. But formalisation imposes a cost – it reduces the ability of offices and institutions to (informally) connect with the people and build authentic interpersonal relationships of mutual understanding and dialogue. Healthy parties, on the other hand, despite their internal institutionalised structures, tend to retain the nimble flexibility and informality of civil society organisations – at least in their local units. This measure of informality allows them to perform their key coordinating function: to imbibe and influence popular opinion, on the one hand, and to formulate and justify their proposed policy package, on the other. The relationship between popular opinions and policy packages is mutually responsive – in a well-functioning democracy, they respond to each other and form a feedback loop. The central task of political parties is to facilitate this responsive relationship between popular opinion and policy.³⁷ Sometimes, they absorb popular opinions and translate them into policy proposals. At other times, they articulate policy proposals and mould public opinion to get behind them.

While Sartori accepts this dual function, ‘grant[ing] that parties are a two-way communication channel’, he insists that ‘the conclusion does not follow that parties are a transmission channel downward to the same extent that they are a transmission belt upward’.³⁸ In part, his reluctance to idealise the bidirectional communicative purpose is based on his characterisation of what I have described as the justification of public policy to the people as ‘manipulation’ of popular opinion.³⁹ This pejorative characterisation under-appreciates the democratic need for the state to justify its policies to the people, including to those who disagree with the government, and the critical role that (healthy) parties play in demanding, articulating, and challenging such justification.

4.2.1.2 Plenary Character of Parties

The second special feature of the mediation role that well-functioning parties play between the state and its people is their plenary character. In heterogenous societies, the values as well as the interests of the people are likely to be diverse. Value pluralism as well as interest pluralism pose a huge challenge to the ability of the state to frame public policy that would be broadly acceptable to its people. The multitude of ways in which different values and interests may combine is so staggeringly large that any complex society faces the potential problem of being left with most of its population being perennially disgruntled. Parties (when they function well) perform a significant legitimation function for the state by coalescing around distinct families of values – often described as ideology – and aggregate the diverse interests of (all) the people of a state into a coherent policy package more-or-less compatible with their ideology. The policy package need not be internally coherent – it often involves the weighing of various interests, preferences, and values. It may entail a multitude of compromises that seek to bridge the gap between the ideal and the feasible and must frequently cater to logically opposed interests, values, and preferences.

The internal contradictions of the policy package of a well-functioning political party notwithstanding, the party can claim that its mediation has a plenary character in three distinct senses: first, it mimics the plenary nature of governance, which is at least potentially concerned with all issues affecting human flourishing (as well as interests of non-human beings). No state can decide to have a policy only on healthcare, for example. Even its silence or inaction on all other matters will amount to a policy decision, which it would be well-advised to adopt deliberately rather than inadvertently. As the drivers of governments-in-waiting, governance parties (that seek to capture high executive offices, as opposed to influence parties that primarily focus on policy impact) come up with policies on a wide range of issues, drawing upon their interaction with the people, and then seek to sell them politically to the people as a package. In doing so, they persuade their supporters to accede to certain compromises made with their own values, interests, and preferences so long as the overall policy package remains attractive to them. These policy platforms also make the opportunity costs of their policy packages transparent to voters, who are better able to prioritise their preferences in the context of resource constraint.

The policy package of a healthy party is also plenary in a second sense: it is one that is designed by putting the interests of all the people on the scales. I will shed further light on this feature when discussing the anti-faction principle. For now, it will suffice to note that a party need not – and cannot – commit itself to all the mutually incompatible values that the multitude of people in a state adhere to, nor can it sincerely claim to have aggregated the preferences of all the people in even a moderately diverse society. But what a non-factional party can – and should – do is to sincerely seek to aggregate the interests of all the people in its polity. No policy package can cater to all these interests equally, and some interests will of necessity be compromised or sacrificed for others. But parties should consider the well-being of all the people: any political group that a priori dismisses the interests of any section of the population as either irrelevant to its policy considerations or worse, meriting its hostility, is no longer committed to the rule of all the people and is basically a faction rather than a party.

Third, parties have a plenary character in as much as they are more likely than most other political actors in electoral democracies to have long-term horizons and therefore are likely to care more about the interests of the future people. Unlike the naturally limited lifespans of individual politicians, parties can – and usually seek to – endure over a long time. Sure, parties care about winning the next election, but that is not their only goal. It is sometimes rational for parties to prioritise ideological victories, organisation-building, performance in future elections, etc. over winning the next election only because parties have longer-term time horizons. As Rosenbluth and Shapiro correctly state, ‘parties have reputations that outlive those of individual politicians, and to the extent that they must represent a wide view of societal interests, they are more capable of delivering desired outcomes than any amount of direct democracy, and more trustworthy than even the most appealing individual politician’.⁴⁰ Indeed, parties frequently outlast constitutions, even several constitutions. This endurance feature adds a temporal dimension to the inclusive plenary character of parties and makes healthy parties much more likely than parties dominated by single individuals to attend to issues such as climate change that will disproportionately affect non-voting children and future people.

4.2.2 Key Costs Reduced by Parties

In providing this uniquely bidirectional and plenary mediation between the state and its people, political parties (in efficient multi-partisan systems)⁴¹ reduce key information and transaction costs for both, making democracy possible. For Sartori, ‘Parties make for a “system” … only when they are parts (in the plural); and a party system is precisely the system of interactions resulting from inter-party competition. That is, the system in question bears on the relatedness of parties to each other, on how each party is a function (in the mathematical sense) of the other parties and reacts, competitively or otherwise, to the other parties.’⁴²

Parties are able to reduce the costs I am about to discuss mainly in well-functioning party systems. One-party ‘systems’, therefore, obviously fail to achieve a reduction in the costs that makes democracy possible. Multi-partisan systems – defined by the nature of the parties they have and the nature of the interaction between them – may be more or less efficient at reducing these costs. Other things being equal, constitutions will deepen democracy if they make their party systems and parties more efficient at reducing the following costs.

4.2.2.1 Political Participation Cost

First, healthy parties in efficient party systems reduce the transaction costs of political participation for citizens (political participation costs). Even in a smallish party-less democracy, an ordinary citizen acting on her own would almost certainly need to take up political engagement as a full-time occupation to have any hope of making a modicum of difference to state policy. Sooner or later, she will have to invent something that looks like a political party to enable some political engagement by citizens who do not wish to become full-time politicians. Parties also reduce the transaction costs of political participation for citizens – not only for partisans, but also for non-partisan citizens – who, in a well-functioning pluralistic democracy, are likely to find some party that reflects their values and priorities most closely and could therefore be their first port of call when raising a matter of political concern.⁴³

High political participation costs can be debilitating for a democracy. Imagine a society without shoemakers. In such a society, one would have to make one’s own shoes or do without them. By reducing the political participation costs, parties offer a similar service of specialisation to citizens – they can custom-build their own mode of political participation, but without parties this is going to be expensive, futile, or both. One of the biggest limitations of recent innovations such as citizens’ assemblies and other party-less sortition-based mechanisms aimed at enhancing political participation is that even as they facilitate political participation,⁴⁴ they are not likely to do so for most citizens for the issues they care about most. A person who is really interested in the question of climate change will not feel sufficiently included without any political party that incorporates the issue on its political agenda, even if this citizen is selected through sortition to participate in an assembly to determine whether abortion should be legalised.

Furthermore, because of their temporally plenary character, parties alone can provide a modicum of representation to future electorates, who are otherwise entirely unable to participate in a democracy. Note that, by definition, the political participation cost has to be affordable by all the people, if the regime is to count as a democracy. Any group that is permanently excluded from the political process because their participation cost is too high changes the very character of the regime. Thus, systems that systemically exclude or suppress voter types, gerrymander constituency boundaries to make their access to power very difficult, or permit political domination by the wealthy through inadequate regulation of campaign finance impose very high political participation costs. This is why, in a healthy party system, parties will be accessible to all persons without discrimination and have a fair chance of winning (subject to the principles explained in Section 4.3); if a salient voter-type finds itself without a political party representing it, the costs of creating a new party to provide such representation will also be relatively low in a healthy system.

4.2.2.2 Voters’ Information Cost

Secondly, parties reduce information costs. In constituencies whose large size is typical of contemporary states, voters tend to lack personal knowledge of electoral candidates. Given modern population sizes and the predominance of gesellschaft relationships outside small kinship and friendship circles, it is usually not feasible to have constituencies so small that most voters are personally sufficiently acquainted with all candidates. Parties reduce the information costs for voters because party affiliations of different candidates provide them with a significant amount of broadly accurate proxy information about their political views and agendas, thereby reducing their voters’ information costs. Party systems that get rid of individual candidates entirely and allow voters to vote directly for parties may be particularly efficient at keeping this cost low, assuming that the number of parties in the system is small enough and their policy platforms transparent enough for a reasonably diligent voter to make informed choices without too much effort. But voter education can be difficult in closed list voting systems because it is not just parties that supply proxy information for their candidates; local candidates also educate voters about their parties. A system where both parties and local candidates (in single-member or multi-member constituencies) matter is likely to reduce voters’ information costs to the greatest extent.

It is easy to take the lowering of voters’ information cost by parties for granted, but in doing so we imperil democracy itself. When electing a government, the act of voting is not merely self-regarding but also other regarding. Given the serious implications my vote can have on the lives of others, it is reasonable to suppose that the right to vote also entails a (moral) duty to vote judiciously. Casting a vote is more like a judge deciding a case (albeit without the duty to give public reasons) than choosing what to eat from a restaurant menu. While a voter has a duty at least to discharge her democratic function with due diligence (if not also with an attitude of care towards her compatriots),⁴⁵ it is also incumbent upon the state to ensure that the diligence burden on a citizen trying to vote judiciously is not too onerous. One implication of this is that the voters’ information costs should be kept relatively low.

4.2.2.3 Policy Packaging Cost

Third, parties also reduce information costs for democratic state institutions by revealing to them what combination of policies will be acceptable to what proportion of the people. All parties that campaign on policy packages provide this information to state institutions, whether they win or lose. And winning parties, in addition, inform state institutions about the particular policy packaging that a large proportion of – if not a majority of – the people are willing to at least tolerate. This information can be generated and revealed, and state policy be legitimised, only through the bidirectional and plenary character of the mediating function that parties perform. Let us label these information and transaction costs as policy packaging costs.

The importance of lowering the policy packaging costs for state institutions should be obvious. It allows them to be responsive to the people in a way that is likely to cause the least amount of disgruntlement among those who disagree with individual policy proposals. Bundling different proposals into a single policy platform, while also rejecting alternative policies, makes the opportunity costs of policies clear to voters. They know not only what they are getting but also what they cannot get if they choose a certain set of policies.⁴⁶ This is key to the legitimation of the state’s policies to its people. The fact that the policy packages are themselves framed bidirectionally by political parties drawing upon their proximity to popular opinion, and then justified to the people, buttresses the perceived as well as actual democratic legitimacy of partisan state institutions. Without parties, there is simply no efficient way for state institutions to gather this information: opinion polls can tell us the level of popular support for particular issues, but the unity that packaging brings to a stack of proposals is only possible through the iterative platform and the legitimising glue that only a political party can supply. In other words, in a party-less state, the regime-legitimising policy packaging costs are infinite.

4.2.2.4 Ally Prediction Cost

Finally, parties reduce information costs for other political parties as well as for state officers and institutions by indicating to them which office-holders are likely to be persuadable political allies, whose support can be taken for granted, and who are likely to oppose certain policy proposals. Moreover, when they are reasonably disciplined, parties permit the identification of key leaders whose support will likely translate into the support of a predictable number of legislators and what it might take to secure their support. By aggregating and publicising political leanings, parties reduce the information costs associated with discovering whether another political actor is a political friend or foe and the consequent transaction costs in making political decisions (ally prediction costs).

Ally prediction costs are also of considerable significance in a democracy. Without parties, any decision-maker will need to seek the individual consent of every representative. Because policy packaging will also be absent, the ensuing political bargains will need to be simultaneous, numerous, and mutually compatible. Every individual representative in such a system will have the incentive to maximise the concessions she can extract for her support, whereas the policy initiator must make these concessions until she has the necessary support – she will, after all, be working with little knowledge of potential allies given the high ally prediction costs. The number of possible veto-players will be too high, and thus, state policy will tend towards preserving the status quo. Democratic decision-making under such scenarios can become extremely difficult.

We can therefore see the important role parties can play in keeping these four key information and transaction costs low. Without them, these costs will be too high to permit the smooth functioning of a democracy . Recent proposals for directly democratic citizen’s assemblies (especially when conceived as decision-making rather than advisory bodies) grossly underestimate – or ignore – the substantial hurdle that these costs pose and the role that political parties play in reducing them and making them affordable.⁴⁷ Parties are therefore the key vehicle for a responsive interaction between public opinion and public policy, the very essence of democratic governance. Apart from policy responsiveness, the intermediary role that parties play makes them ideal training grounds for tomorrow’s leaders, for public education and debate on civics, and as a vehicle for social solidarity and camaraderie. The training role is often underestimated in our popular culture that does not appreciate that politics is a skilled profession, like every other, one that typically requires a politician to regularly meet political opponents who vehemently oppose and criticise them, to interact with a wide cross-section of society that includes people from backgrounds very different from their own, to be nimble with compromises to get things done, to be able to withstand the emotional burden of constant public scrutiny, to learn to speak pithily and engagingly, to respond quickly to changing circumstances. These virtues do not come easily to most of us; parties provide a forum for acquiring and honing these key political skills and a community for solidaristic support and mentoring from party colleagues that these difficult engagements often necessitate.

With respect to citizens, the acts of raising a concern with the local party representative, becoming a party member, campaigning in an election, or joining a political protest organised by her party transform a legal subject into a performative citizen. Healthy parties can translate an individual’s grievances into common causes, allowing her to see, for example, that the absence of a girls’ toilet in her daughter’s school is a broader concern that implicates patriarchy and affects women and girls across the country.

In this section, we have seen that political parties are essential to the proper functioning of representative democracies in sufficiently large and complex polities. They act as intermediaries between the state and its people, on the one hand, transmitting popular opinion to state institutions that typically lack the ability to gauge it directly and, on the other hand, formulating state policies and justifying them to the people. This dual role gives them a Janus-faced public-private character – they need to operate as a private association proximate to the people in order to access popular opinions and justify state policies. They also simultaneously need to be embedded in (but not fused with) the institutional structures of the state to transmit popular opinions back to them and to help them formulate policies, which in turn they will help justify to the people.

Healthy parties in well-functioning party systems therefore grease the wheels of representative democracy by reducing the following information and transaction costs: the political participation costs and the voters’ information costs for the people and the policy packaging costs and ally prediction costs for state institutions. This functional account of political parties has highlighted their indispensability to the effective functioning of a representative democracy. This alone should suffice to alert us that lazy dismissals such as ‘all politicians are corrupt’ and ‘all parties play dirty’ imperils democracy itself. While a discursive defence of the importance of parties in a democracy is crucial, the state and its constitution too can lend a helping hand. They can bolster the ability of individual parties to perform their important mediating function between the state and its people as well as structure the party system in ways that facilitate rather than hinder the ability of parties to deepen democracy. They are more likely to do this if they deliberately consider certain guiding principles in constitutional design.

4.3 Constitutional Principles in Relation to Political Parties

Section 4.2 offered an idealised account of what parties do in a well-functioning democratic system. This idealised account is helpful in distinguishing between parties and party systems that function reasonably well from those that are pathological. It also helps us diagnose the particular ailments that afflict a party or a party-system. It should be clear that parties that fail to perform their intermediary function appropriately and effectively are bad for democracy. A party system can tolerate a few malfunctioning parties and still be healthy, so long as most of the governance parties are sound. In this section, we will turn our attention to certain normative goals that constitutions ought to adopt in relation to parties that will make it more likely that parties and party systems are healthy or that – if there are diseased parties in the system – the system can still tolerate or mitigate their ill effects or even nurse them to better health. To do so, I will explain four principles that constitutions should adopt in relation to parties: the purposive autonomy principle, the party system optimality principle, the party-state separation principle, and the anti-faction principle. Readers should note that for reasons of space, this already ambitious chapter does not take the logical next step of outlining more concrete prescriptions that might follow from the adoption of these principles. Some of that discussion has been undertaken elsewhere.⁴⁸ In this section, I will argue that democratic states should, through their constitutions, consider and seek to optimise four distinct, and sometimes conflicting, political principles in relation to political parties:

1. i. They should guarantee maximum autonomy for the formation, organisation, and operation of political parties, moderated by the restrictions necessitated by their purpose of winning (a share in) state power (for fixed terms) in competitive elections by acting as intermediaries between the state and its people (the ‘purposive autonomy principle’).

2. ii. They should try to optimise the party system such that the total number of serious political parties is large enough to broadly represent every major ‘voter type’, but not so large that the information costs on judicious voters are too high (the ‘party system optimality principle’).

3. iii. They should ensure a separation of parties and the state (the ‘party-state separation principle’).

4. iv. They should discourage the factionalisation of political parties (the ‘anti-faction principle’).

As constitutional principles, they primarily address the state. However, because the ruling party/coalition, which typically holds the reins to state power, is unlikely to be particularly interested in respecting them – and may even have a strong self-interest in breaching some of them – these principles need to be constitutionalised as pre-commitments by the state. I hasten to add two caveats to this proposal: first, I do not take constitutionalisation to necessarily entail judicialisation.⁴⁹ In fact, sometimes it may be neither necessary nor desirable to express a constitutional principle as a constitutional norm directly regulating constitutional actors, let alone as a legal norm. Instead, establishing an institutional arrangement that is most likely to uphold that principle – what may be termed ‘second order’ regulation – may well be the most optimal design solution.⁵⁰

Second, a norm can be ‘constitutionalised’ in multiple ways, its inclusion in a big-C constitutional code being only one of them. Other modes of constitutionalisation include judicial interpretation, quasi-constitutional statutes, and constitutional conventions. The proposed principles should ideally be reflected – at least at a broad level – in the big-C constitutional code so that the institutional arrangements of the state are framed alongside its party system rather than ex ante. The finer details will, obviously, need to be left to the small-c statutes, conventions, and caselaw. The key determinant in each context should, in the main, be feasibility and effectiveness in light of path dependencies and all-things-considered judgments. The following sub-sections will explain each of these principles in turn.

4.3.1 The Purposive Autonomy Principle

4.3.1.1 The Public Purpose of Private Parties

States and their constitutions must seek to support (existing and future) political parties in the performance of their bidirectional and plenary intermediary role between the state and its people, so that they are best placed to reduce the four key information and transaction costs that make democratic governance difficult. In order to do so, parties must remain simultaneously private as well as public – this duality is their unique strength and also the reason for the complexity in regulating them. Liberal constitutionalism has long adhered to a controversial public-private divide. This divide is premised on diametrically opposed default assumptions about the regulation of public and private actors. Private actors are granted the autonomy to do whatever they like, unless there are very good, and special, reasons for regulating their actions. Public actors, on the other hand, may not do anything at all unless they have good, and constitutionally permitted, reasons for doing something. This distinction has long been criticised, with some scholars calling for its abolition,⁵¹ whereas others have argued that the distinction may be preserved while the line between the public and the private should be drawn differently.⁵² In a previous work, I have argued that the distinction is best understood as a spectrum rather than two discrete boxes.⁵³ The spectrum is both actor-sensitive and action-sensitive and primarily tracks interpersonal power and other-regarding functions. At the public end of the spectrum stands the all-powerful state enacting a general criminal or tax statute in its legislature. At the other end, a natural individual person lost in deep thought in her bedroom is paradigmatically private. When she acts as the manager of a large firm the next morning, she has moved away from this private-most end and become more public. Some areas of law – such as discrimination law – have come to terms with the idea that the public-private divide is a spectrum.⁵⁴

Constitutional law is yet to follow this trend. It continues, on the whole, to draw the line sharply, vesting private actors with constitutional rights and burdening public actors with constitutional duties.⁵⁵ This structural limitation is an important hurdle that must be overcome if constitutions are to properly regulate political parties without destroying their public-private duality. Treating them as just another state institution is likely to seriously compromise their ability to engage with the people directly. While constitutions must be careful about over-regulating political parties lest they destroy their private character, they should also worry about constitutional silences and under-regulation that fails to acknowledge their publicness. A fit-for-purpose constitutional scheme for political parties will pay attention to three dimensions: (i) subject to the principles discussed in this chapter, it will grant them maximum autonomy; (ii) it will vest in them the necessary rights, powers, and entitlements that will enable them to better discharge their functions; and (iii) it will impose only those duties on parties that are necessary to preserve their public character. Is there such a happy regulatory middle that would preserve their privateness while demanding that they be sufficiently public at the same time?

To locate that regulatory middle, we need to point out with greater precision what precisely makes parties public. The private dimension of parties demands maximum autonomy for the formation and operation of political parties. But their public character demands a recognition of their purposive dimension: unlike natural individuals, political parties in a representative democracy cannot be allowed to choose their purpose with complete freedom. What makes them a political party in a democratic party system is their public purpose of participating in competitive elections – with other parties – in order to secure a measure of control of the levers of state power for fixed periods of time and to do so by acting as intermediaries between the state and the people. This purpose is definitional of what a political party in a democracy is. It is specified at a high level of generality, being compatible with an extremely wide range of more specific purposes that parties may have. But it is incompatible with certain purposes: such as instituting a single party state; making elections insufficiently competitive; barring or making it difficult for (other) parties to connect or communicate with the people; and so on.

4.3.1.2 The Autonomy of Parties

To respect their private dimension, constitutions should guarantee maximum autonomy for the formation (from scratch or by splitting an existing party), organisation, and operation of political parties, moderated by the restrictions necessitated by their public purpose of winning (a share in) state power (for fixed terms) in competitive elections by acting as intermediaries between the state and its people. Hence the purposive autonomy principle. Simply put, the principle permits significant autonomy to parties (and partisans) but seeks to ensure that they are committed to the purpose of being but one player in a multi-party democracy. The principle requires that parties should be relatively easy to form and disband and to enter or leave. The main barriers to their success should be political, not legal. New parties or opposition parties must not be locked out of political competition through high entry barriers.⁵⁶ High access barriers are not only bad for new parties but also for established parties – without an alternative political outlet, strong political forces that are excluded from partisan expression would eventually seek to capture an established party or upend the system itself. The autonomy of political parties is key to keeping the political participation costs in a democracy low.

Translating the need to protect their purposive autonomy into particular rights and duties requires further specification beyond the scope of this chapter. Full specification cannot, obviously, happen without catering to the peculiarities of a particular political context. In general, parties may need the whole suite of civil and political rights that citizens ordinarily have access to in a liberal democracy; sometimes they may even need special protections of their autonomy over and above what citizens are guaranteed. They are likely to need access to all the usual fundamental freedoms to enable them to be formed, to contact and connect with the people and mobilise them, to campaign and express political views, to contest elections, to raise funds, and so on. Without these freedoms, a political party may be woefully inept at reducing key democratic costs. Furthermore, a range of ‘fourth branch’ guarantor institutions that can effectively guarantee norms such as free and fair elections, probity, transparency, fair boundaries delimitation, campaign finance regulation, and a host of other norms that the ruling party/alliance may have reasons to want to undermine are usually necessitated by the purposive autonomy principle.⁵⁷ These constitutional protections of party autonomy are also necessary to insulate opposition parties from any self-interested targeting by the ruling party/coalition: not least because the autonomy of all (serious) parties matters, such that the principle of inter-party equity is embedded in the purposive autonomy principle.

4.3.1.3 Public Entitlements for Serious Parties

While their privateness demands the protection of their autonomy, their public purpose may entitle them to special privileges and powers, as well as make them fit for bearing special duties that are inapplicable to natural individuals. Public entitlements, such as (limited) state funding for political campaigns or immunity from defamation laws for political speeches, can help secure a level playing field between political parties and enable many of them to discharge their democratic functions effectively.⁵⁸ This would be especially useful for smaller parties in a hegemonic or a predominant party system.⁵⁹ Even if the predominant party is itself healthy, the party system isn’t. Thus, any provision of state benefits has to avoid cartelisation. In order to reduce the political participation costs of all voters, a measure of state support to less powerful parties would usually be necessary. That said, it may be permissible to restrict these rewards to serious political parties. Seriousness is a measure of the party’s intention to play the requisite intermediary role between the state and its people – a serious political party seeks a role in the governance of the state, or an influence in state policies, or both, primarily by winning elections and shaping political discourse. Note that both governance parties and influence parties are ‘serious’ in the sense I intend here. On the other hand, a party that doesn’t campaign or put up candidates for elections hardly has a public dimension worth worrying about. Similarly, joke parties (such as the British Monster Raving Loony Party) and a number of single-issue parties (such as the Australian Help End Marijuana Prohibition Party) often don’t even intend to win – and tend to be themselves surprised when they do win – even if their ability to focus the spotlight on an ignored issue can often be valuable for a polity. Such non-serious parties can largely be regulated as private clubs left to enjoy the autonomy that other political parties enjoy, but without the public benefits or burdens until they have proven their political seriousness.

While performance in past elections is a reasonable proxy of seriousness, it is generally a good idea to base it on aggregate performance in the three or four previous elections taken together, rather than the most recent election alone. For example, equitable state funding may be provided to every party that secured at least (say) 5 per cent of the popular vote share in any of the last three election cycles, or to every party that had a presence in the legislature in at least one of the last three elections. Where these seriousness thresholds are fixed is necessarily arbitrary and context-specific – any reasonably low threshold that has not been weaponised against particular parties and doesn’t squeeze out smaller, but serious, parties should normally suffice (any such weaponisation would, we will see, breach the party-state separation principle). Admittedly, one problem with measuring seriousness (a forward-looking phenomenon) through past electoral performance is that it cannot distinguish between serious and non-serious new parties. Since the purposive autonomy principle applies not only to existing parties but also to future ones, any forward-looking public support may need to be extended to all new parties willing to accept the public duties imposed on serious parties, with exclusions kicking in after they have participated in (say) two election cycles.⁶⁰ Note also that while seriousness of a political party is a relevant consideration in determining public funding, it does not extend to calibrating the extent of funding proportionate to party size. If anything, the kind of distinction US law draws between major and minor parties in order to supply greater benefits to the major parties are likely to be inimical to the principle of equity embedded in the purposive autonomy principle (as well as the party-state separation principle).⁶¹

4.3.1.4 Public Duties of Parties

The public purpose of parties invites not only special entitlements but also some public duties. We will discuss some of them under the other principles to follow. But some duties flow directly from their role in reducing the political participation costs and the policy packaging costs. Parties are likely to lower the political participation costs only if there is a fair measure of transparency surrounding their core value commitments, internal institutional structures, decision-making processes, financial affairs, and credible – even if internal – enforcement mechanisms of their institutional commitments. Serious parties should, therefore, have the duty to publicise these details, so that voters and partisans can rely on them and hold parties accountable. Likewise, parties are likely to lower the policy packaging costs (as well as the other three democratic costs) only if they offer a more-or-less comprehensive policy package in their election manifestos. Serious parties may, therefore, be mandated to adopt and publicise (at least broad) policy positions on a number of the key issues of governance of the day (such as taxation, foreign policy, education, health care, immigration, and so on) in their election manifestos, preferably published before a fixed period prior to elections to give them sufficient public airing.

On the whole, the purposive autonomy principle seeks to preserve the public-private duality of political parties that is essential to their role in facilitating democratic governance. Supporting parties requires guaranteeing considerable autonomy to all existing and future parties. At the same time, at least serious political parties need affirmative state support and protections that enhance their abilities to reduce the four democratic costs. The purposive autonomy principle also justifies the imposition of the duty to adopt transparent party constitutions and plenary policy manifestos on such serious parties. But all public duties imposed on parties need to be justified with reference to their public purpose of being but one player in a multi-party democracy. If we consider Article 21(1) of the German Basic Law, for example, it is broadly a recognition of the purposive autonomy principle:

The political parties participate in the formation of the political will of the people. They may be freely established. Their internal organisation must conform to democratic principles. They must publicly account for their assets and of the sources and use of their funds as well as assets.⁶²

‘Broadly’, because I am doubtful that inner-party democracy – mandated by the third clause above – can be justified by the purposive autonomy principle. It is by no means obvious that internally democratic parties are better at reducing the key democratic costs, not to mention the pragmatic difficulties in determining what suffices as an internally democratic party at an age of relatively loose and myriad ways of associating with a party.⁶³ If anything, base-voter-dominated party primaries have emerged as a key centrifugal force in US politics and the concomitant threat that force poses to democracy in the country.⁶⁴ Constitutions should be slow to mandate inner-party democracy or regulate how parties discipline their members. Many courts have enforced fundamental rights claims by ordinary voters and party members against political parties and their leadership.⁶⁵ Doing so has clear, and often adverse and unintended, consequences for the purposive autonomy of political parties. The implication is not that parties should be allowed to treat their members in any manner they wish; it is rather that state regulation of the relationship between parties and their members should cross a high bar of purposive justification that takes their hybrid public-private character seriously.

4.3.1.5 Preference for Second-Order Regulation

The need to protect purposive autonomy of parties dictates not only a cautious approach to imposing duties on parties but also how any duties may be imposed. Duty-imposing norms should be crafted so as to not destroy the dual character of parties. In general, and subject to their effectiveness in a given context, three broad regulatory criteria should govern design possibilities for duty-imposing norms with respect to political parties:

• – Political enforcement and self-regulation are better than judicial enforcement,⁶⁶

• – Nudges are better than command-and-control,⁶⁷ and

• – Carrots are better than sticks.⁶⁸

These criteria are partial to ‘second-order regulation’, which emphasise the importance of ‘background competitive structures’ that shape decision-making, rather than seeking to police behaviour directly through first-order commands.⁶⁹ Note that all background structures shape the behaviour of actors – the question is not so much whether to have second-order regulation but what type of second-order regulation is worth having. For example, a democracy has to choose some electoral system, and each system shapes the behaviour of politicians differently. In fact, the choice of the electoral system (majoritarian or proportionate, ranked or unranked, at large or constituency-based), the nature of executive-legislature relations (parliamentary, presidential, semi-presidential, semi-parliamentary), the number of legislative chambers, and the degree of centralisation or federation are all institutional choices that can significantly impact the nature of the party system in a polity. For example, majoritarian, cumulative, approval, and ranked voting systems are likely to incentivise centripetal parties, whereas proportionate representation is more conducive to factions. Parliamentary systems are likely to encourage collective party leadership, whereas presidential systems could encourage individual-centric parties. Bicameralism and federalism are probably better suited to protecting opposition rights than unicameralism and centralisation. The directionality of these precise connections is beside the point – if one concedes that a significant connection exists, in whatever direction, then many regulatory objectives in relation to parties can be achieved by the right combination of institutional design of state bodies. This indirect, second-order, regulation is generally more conducive to party autonomy than first-order legal regulation.

I am not suggesting that rule-based, judicially enforced, command-and-control first-order regulation will never be appropriate. Sometimes it will be. The party-state separation principle will normally require stringent, often legal, norms to protect the separation of the state from the ruling party/coalition. Most aspects of private law, such as the law of torts or contracts, should apply to parties, unless their public character demands an exception to be made. Anti-discrimination norms that support the anti-faction principle are usually already calibrated to justify judicial intervention against hybrid public-private actors⁷⁰ and will therefore be appropriate for judicial enforcement against parties in most jurisdictions. Consider Smith v Allwright, where the United States Supreme Court first determined that all-white primaries for the selection of Democratic Party candidates were so well-integrated in the electoral system of Texas that they amounted to state action, before applying the anti-discrimination principle to the Party and prohibiting all-white primaries.⁷¹ This move may have been doctrinally necessary because of the quirks of the American legal system, but morally speaking, there is no reason why anti-discrimination norms should not apply to parties qua parties, that is, as hybrid public-private bodies, without needing to characterise them as state institutions. In general, however, where feasible and effective, second-order regulation is more likely to preserve the dual character of political parties. Extreme measures such as party bans are best avoided. Courts understandably struggle to condemn even the most clearly anti-democratic parties when the outcome is as far-reaching as a total ban.⁷² What’s worse, even when successfully deployed, party bans may be ineffective, or even counter-productive.⁷³ Having said that, second-order regulations would sometimes fail to achieve their desired purpose. While an independent guarantor institution like a boundaries commission may well ensure fair constituency determination in most cases,⁷⁴ even the best design can fail. If a captured boundaries commission produces a gerrymandered electoral map, there may well be strong reasons for first-order judicial intervention.⁷⁵

The purposive autonomy principle is a meta-principle that dictates how constitutions should approach political party regulation. It demands a move away from a binary dichotomy that liberal constitutionalism is used to: rights for private persons, duties on public bodies. Instead, it draws attention to the need for a more nuanced approach, one that considers each right and duty in terms of its appropriateness for the hybrid character of political parties. The discussion above is not meant to be too directive: in constitutional practice, context matters a lot. Even in a given context, there may be many different ways of satisfying the purposive autonomy principle. The main point is that instead of trying to fit political parties into a ready-made template designed either for private individuals or for state bodies, an a la carte regulatory framework that works for them is required. The three following principles may be understood as facets of the purposive autonomy principle but merit separate discussion because of the important bearing they have on the constitutional regulation of parties.

4.3.2 The Party System Optimality Principle

Healthy parties tend to attract members, affiliates, and voters through their ideologies, rather than through clientelism or patronage. I understand ‘ideology’ in the sense that Converse explains it: a relatively wide-ranging belief system, which is relevant to political behaviour.⁷⁶ A belief system, in turn, is ‘a configuration of ideas and attitudes in which the elements are bound together by some form of constraint or functional interdependence’.⁷⁷ The centrality of an element in a belief system is a measure of the likelihood that a voter will change her party preference if her party’s stance regarding that element changes, rather than change her view on the element itself.⁷⁸ She may tolerate a party’s change of position on less central elements in a belief system but give up on her partisan loyalty if the party reneges on a more central element. For Gerring, the quality of being ‘bound together’ (which he calls ‘coherence’) has two corollaries: contrast (‘implying coherence vis-à-vis competing ideologies’) and stability (‘implying coherence through time’).⁷⁹ Thus, competing political ideologies straddle the same ideological axis and are relatively stable over time. While much penumbral content of political ideologies is malleable, their most central elements are likely to be most relevant to contrasting them with other ideologies and determining their stability over time. Not all ideologies matter politically. Chhibber and Verma argue that politically salient ideologies not only need to be competing and stable but also possess two further features to structure a party system:

First, there must be political, social, or economic elites with interests, vested or otherwise, who differentiate themselves on the basis of such ideas, offer resources to support the creation of a particular ideological position, and assist in the transmission of these ideas to the voters. And, second, the ideas thus transmitted, and the issues that embody them, must have the support of enough people.⁸⁰

Without sufficient elite support, parties are unlikely to get off the ground. Without enough support in the electorate, they are unlikely to become meaningful political players. Sartori’s classical account analysed party systems through this lens of salient ideological axes. His distinction between two-party systems, moderate pluralism systems, and polarised pluralism systems was based on two factors: the number of political parties in a system and the ideological distance between them.⁸¹ In this model, a two-party system had two large, ideologically centrist parties (i.e., very little contrast between their ideologies); a moderate pluralism system had multiple parties organised around two broadly centrist ideological coalitions; and a polarised pluralism system had multiple parties pursuing ideologies cutting across the traditional left-right axis seen in the other two systems. The ideological distance between the parties/coalitions in the two-party system and the moderate pluralism system was relatively modest because each pursued the median voter organised on a single – macroeconomic policy – axis. On this account, the first two centripetal systems usually yielded stable and effective governments, whereas polarised pluralism often resulted in political as well as democratic instability.

This classical account needs an important revision in our times. As Scheppele has argued, politics is no longer organised on a single left-right ideological axis in contemporary Western democracies. In the very least, a nativism-cosmopolitanism divide has strongly emerged as an additional, cross-cutting, axis for political alignment.⁸² Chhibber and Verma argue that independent India has always had two salient political axes: the politics of statism (‘the extent to which the state should dominate society, regulate social norms, and redistribute private property’) and the politics of accommodation (‘whether and how the state should accommodate the needs of various marginalised groups and protect minority rights from assertive majoritarian tendencies’).⁸³ These insights scramble the tidiness of Sartori’s single-axis classification. When two major ideological axes are salient to voters, there are at least four stable party types (and concomitant ‘voter types’) that can broadly capture the worldviews and political preferences of most voters in such systems: on Scheppele’s classification, for example, one should expect left-nativist parties,⁸⁴ right-nativist parties,⁸⁵ left-cosmopolitan parties,⁸⁶ and right-cosmopolitan parties.⁸⁷ With each new salient axis, new permutations give rise to the possibility of an even larger number of voter types in search of distinctive political representation.

These ideological axes must be salient: political parties offer package deals, and they cannot be expected to customise their policy offerings to all the individual preferences of every voter. We should therefore understand key voter types only in relation to the politically salient ideological axes of division in a given polity, which in turn are defined by the most central belief elements in their belief system (such as nativism or redistribution). A salient political division comes to define the political preferences of voters in such a deep way that it becomes difficult for the same party to simultaneously represent the preferences of distinct voter types effectively. Such parties may try to speak in multiple voices to mutually incompatible constituencies (in the short term). However, if the salient divisions are abiding, these parties are either likely to split (if the system tolerates multiple parties) or have one group eventually come to dominate or decimate the other within the party (if it doesn’t).

If the latter happens, and no party exists to cater to a particular voter type, the political participation costs of such voters will be extremely high, to the point that they may be totally excluded from representative politics. What is worse, even their efforts to start a party that caters to their voter type may be frustrated if the party system is structurally predisposed to a two-party system and does not facilitate the emergence of new parties. Their only remaining options will be either to capture one of the existing parties (and thereby deprive a different voter type of political representation) or try to change the party system, or even the political system entirely, from the outside. Such excluded voters become especially vulnerable to a quick-fix populist rhetoric.⁸⁸ Such powerful, but unspent, political force is extremely dangerous to democracies. Furthermore, the near impossibility of political alliances in a two-party democracy – except when party discipline is extremely weak – is also not necessarily a good thing for democracy. This is the reason why a two-party system is more likely to encourage compromise-resistant tribalism in politics, accentuating the winner-takes-all feature of majoritarian politics.

What follows is that in any system that has more than one salient political axis, a two-party system simply cannot approximate the broad political worldviews of major voter types. A regime will establish ‘the rule of the people’ only if it facilitates the representation of the preferences of every major voter type in its party system, with two caveats: first, as I will argue later in this chapter, it is legitimate – albeit sometimes unwise – to restrict the likelihood of political representation – or, at least, the likelihood of political success – of factional voters who do not accept that a democracy is the rule of all the people, even if a factional-inclusivist axis has become salient in that polity. Just as I cannot rely on my autonomy to sell my children or my (future) self into slavery, rule of all the people cannot be relied upon to transform a democracy into the rule of some of the people. Neither autonomy nor democracy apply to themselves in this self-harming manner.⁸⁹

The second caveat is that there is a feasibility limit to the total number of serious parties that a democracy can accommodate. It is true that the larger the number of distinctive parties in a system, the smaller the political participation costs are likely to be for a voter. In fact, if there is a party that mirrors every voter’s customised set of political preferences, political participation costs will be non-existent for every voter. Needless to say, such single member ‘parties’ won’t be parties in any meaningful sense. Furthermore, even as they reduce political participation costs, a large number of parties significantly increases voters’ information costs. A voter who has to go through a list of fifty candidates belonging to fifty different serious parties is able to make an informed choice only after putting in considerable effort to educate herself on the distinctive ideological commitments and political platforms of all these fifty parties. She might as well focus her research on the fifty individual candidates in such cases (which would not be any less daunting, in any case). Too many choices may not matter when the stakes are low – such as when one is ordering a meal from a restaurant’s menu – for one can make a reasonable choice having considered only the first five options. But when the stakes are as high as entrusting the government of one’s polity, the voters’ information costs must be reasonable enough to enable a judicious voter to consider the pros and cons of all candidates.

The sum of these concerns is the ‘party system optimality principle’: in contemporary democratic polities that divide along multiple salient axes, party systems should be optimised such that the total number of serious political parties is large enough to broadly represent every major ‘voter type’, but not so large that the information costs on judicious voters are too high. The choice of electoral system for the elected institutions and offices is probably the single most important regulatory tool to respect the party system optimality principle (although the party system can also become federalised in federal systems, since province-specific parties can emerge irrespective of the electoral system). Note that a polity can coherently adopt different electoral systems for different state institutions (especially in bicameral and federal systems), thus vastly expanding the regulatory permutations and combinations at its disposal.

4.3.3 The Party-State Separation Principle

One danger in approximating the democratic ideal through elections is the abiding possibility that we only manage to secure the rule of some of the people at any given point in time. The political exclusion of the losing parties – and potentially their voters – is more acute in winner-takes-all systems. We seek to solve this problem by temporally distinguishing between a state’s regime and its government: a regime can still call itself democratic if, over time, it allows different parties to win political power. So long as there are no permanent winners, the regime can still claim to be ruled by all the people. This is the ideal that legitimises many power-sharing arrangements in deeply divided societies.⁹⁰ It also generates our third political principle: that a state should seek to ensure a separation of the ruling party/coalition and the state, so as to allow a genuine hope for today’s losers to be tomorrow’s winners. We will call this the ‘party-state separation principle’. The basic argument is that if a party (usually the ruling party/coalition) becomes entrenched in the apparatus of the state, the political participation costs of the supporters of all other parties become insurmountable.

The party-state separation principle demands a recognition of a host of opposition rights: including a significant opposition voice – perhaps even a veto – in constitutional amendments and constitutional appointments. It requires the bureaucracy, police, prosecution, judiciary, and guarantor institutions to function in a non-partisan manner. The principle also demands equity in state benefits given to the ruling party/coalition and to opposition parties – making it structurally difficult for a given party or parties to win state offices breaches the party-state separation principle as much as making it more likely for a given party or parties to win or retain state power. Recognising this principle is especially important given the salience of the institutional ‘separation of powers’ principle in constitutional theory – given how partisan loyalties can scramble institutional separation, it is essential that the party-state separation principle is considered alongside the institutional separation of powers principle and given the same weight in constitutional thought. Note, however, that the party-state separation principle does not require that the opposition be given the right to veto ordinary (as opposed to constitutional) policy objectives of a legitimate and democratically elected government. A balance has to be struck between allowing the opposition to perform its constitutional duties without thwarting the government’s ability to govern.⁹¹

One way to appreciate the party-state separation principle is to think of it as a constraint imposed on the democratic rights of today’s people by those of a future people. Constitutional theory has frequently worried about ‘the dead hand of the past’ in the context of entrenched fundamental rights but rarely considered constraints generated by the future people’s right to democracy. Even if we assume that there are a people at a given point in time in a state, and that this people today unanimously wish to entrench the ruling party/coalition in the apparatus of the state, they cannot possibly obtain the consent of any future peoples to do so and must therefore lack this power. One might seek to respond to this claim by suggesting that it might at least be permissible to fuse a state and its ruling party/coalition for one generation, and the arrangement may be revisited by every future generation? The claim presupposes the all-too common, but mistaken, generational outlook to think of a ‘people’ in the temporal sense. A people – in its temporal sense – is not like The Doctor (from the famous British science-fiction show Doctor Who), who disintegrates and regenerates into a new version of him/herself at precise moments in time. Each version is the same eternal Doctor, but different, with clear and distinct temporal geneses and dissolutions. A people would be like The Doctor if everyone in a generation was born on the same day and died on the same day. In reality, a people – temporally – are better compared to a river. An ancient Greek philosopher, Heraclitus, is said to have posed the famous paradox: Can one step in the same river twice?⁹² The paradox lies in the fact that anyone stepping into any given point in the course of a river twice – the two attempts separated by some interval of time – will be stepping into different waters of the same river. Just like a constantly flowing river that receives new waters from its source and loses old waters to the sea, a people are constantly gaining new individuals by birth or immigration and losing old ones to death and emigration. Even the separation of a day, therefore, will suffice to separate the current people from a future one. As an ideational entity, they remain the same people over time; as a sum of their constituent individuals, they are constantly changing over time. Thus, even a single generational fusion of the ruling party/coalition and the state will be impermissible by the democratic ideal.

The party-state separation principle therefore requires that a state should preserve the genuine likelihood of different parties securing governmental power at different points in time. The transfer of power following elections should be peaceful, and the political opposition must be able to plausibly imagine itself as a government in waiting. It should therefore be hostile to a one-party system (where only one party is allowed to exist, de jure and de facto), or a hegemonic party system (where smaller parties are allowed to exist, but the system de facto and de jure favours a hegemonic party that remains in power) at all times.⁹³ It should even be hostile to the kind of two-party system in which the two parties operate like a cartel and make it structurally difficult for a third party to emerge.⁹⁴ Any such fusion of parties and the state is not only bad for democracy, it is also likely to make the regime unstable because any significant voter type without mainstream political representation is likely to find solace in anti-system parties. On the other hand, it is compatible with a predominant party system, where a single party or coalition de facto dominates all others, although de jure the system permits free and fair political competition and gives no structural advantage to the predominant party. That said, the purposive autonomy principle would still view a predominant party system as non-ideal and seek to enable opposition parties to rise and flourish in such a system. Even if the opposition does not win elections, a robust opposition is essential to check the political power of the ruling party/coalition (balanced against the need for effective government) and therefore to reduce the four democratic costs effectively. Recall that these costs remain high in a system with only one healthy political party. Hence the party-state separation principle.

4.3.4 The Anti-faction Principle

We can now consider the final principle. We must accept that an elected democratic government is unlikely to represent all the people of a state at any one given time, where representation is understood in terms of voters’ electoral preferences as expressed on the ballot. But it does not follow that we should also accept that such an under-representative government only needs to serve the interests of those it represents. An under-representative government can, and should, still aspire to serve the interests of all its people. This is not drawn from an agenda for ensuring minority rights under a majoritarian government. Basic rights concern themselves with only the most fundamental human interests. Governments should never breach them unjustifiably. But governments do a lot more than not breach rights – when they work well, they facilitate human flourishing. Parties, in their idealised sense, work towards the flourishing of all the people of their state; factions care only for a sub-section thereof.⁹⁵ Factions a priori exclude the interests of their disfavoured section of the people from even being considered when framing policies – if these interests are considered at all, it is with a view to hurt them rather than to advance them. Importantly, given our capacity to threaten the very survival of humanity, at least in our times, factions would include parties that do not count the interests of the future people as legitimate concerns for their political calculations.

Factions fail to reduce the policy packaging costs for state institutions. We have seen that one of the key functions of political parties is to package the interests of all voters based on the party’s value commitments. These policy packages are then tested in elections, and voters express their preferences for or against such packages, which information is then available to state institutions when framing policy. In the process, parties also translate any voter’s factional interests into a subset of the common good through their policy packaging function, thereby moderating them to make them compatible with the interests of other citizens. Factions fail to do so. They also increase the political participation costs of the excluded voters – it is one thing to not have every party reflect one’s voter type, quite another to have a party in a system not even consider one’s interests as legitimate and relevant alongside the interests of all others.

This distinction between a party and a faction has been long recognised in political theory.⁹⁶ As Sartori put it, ‘If a party is not a party capable of governing for the sake of the whole, that is, in view of a general interest, then it does not differ from a faction. Although a party only represents a part, this part must take a non-partial approach to the whole.’⁹⁷ Factions are concerned with the interests and well-being of only a sub-section of the people. Parties, even when they make claims on behalf of particular groups, ‘must transcend the language of particularity and re-articulate the claims they represent in such a way that their demand for a share in political power is justified to the entire people and not only to that particular group of individuals that chooses to associate with them’.⁹⁸ The point of the distinction is normative rather than taxonomical: ‘[It] is very likely that the empirical analysis of existing practices will show how parties and factions are often entangled, with different political agents exhibiting features of both, to a greater or lesser extent.’⁹⁹ It is important to note that the distinction attaches itself to the entity as a whole and not to its individual actions. A party may have distinct policies catering to the interests of different sub-sections of the people – it will be a faction only if, taken as a whole, its political ideology and its policy platform is not justifiable to all the people. Any attempt to distinguish real-world parties from factions too sharply is likely to fail. Having said that, Rosenblum is probably right when she suggests that, even as an empirical matter, ‘where it is an original identity, or at least not reducible to prior political identities, the “we” of partisanship is more inclusive than other political identities’.¹⁰⁰

‘Rule of the people’ demands not only that political power is exercised by the people’s representatives but also that it is exercised in the name of all the people. In the words of White and Ypi, ‘the very ideal of collective self-rule implies that power is considered legitimate to the extent that it is justified to the whole people’.¹⁰¹ It is this normative ideal that leads us to our final principle: that a state should seek to ensure that political parties do not operate as factions. We will call this the ‘anti-faction principle’ and amend White and Ypi’s formulation somewhat to suggest that it requires political parties to ensure that their policies are objectively justifiable (rather than subjectively justified) to all the people. The amendment is required because it may be that a party fails to even communicate, let alone actually justify, its policies to all the people. So long as its policies are justifiable to all of them, the anti-faction principle should be satisfied. The anti-faction principle, therefore, does not require parties to articulate their policies in Rawlsian ‘public reason’ terms.¹⁰² Furthermore, a justifiability-standard is more tolerant of parties strategically appealing to particular sub-sections of the people as a matter of electoral tactics – so long as their packaged policy platforms are justifiable to all the people. Other independent moral and political constraints no doubt exist – such strategic appeals should not demonise any other section of the people, for example.

Unlike the purposive autonomy principle, which frowns upon single-issue parties, the anti-faction principle – on its own – does not require parties to have a plenary policy package. An anti-corruption party is not a faction. The party’s size doesn’t matter either, even if the party seeks only to influence policy rather than to govern. A small Green Party is likely to be a party, since its environmental objectives are justifiable to all. Even a party whose entire policy platform is devoted to advantaging a single societal group may not necessarily be a faction. A Workers’ Party, a Dalit Party in India (for former ‘untouchable’ castes), or an African American Party in the United States can be parties, if they can justify the interests of their preferred groups by reference to the general interest (for example, that historically excluded groups have a greater claim on the state’s resources). Furthermore, parties are allowed to make ideological and policy mistakes – the anti-faction principle does not demand that their policies actually work. But it does demand sincerity and plausibility – some obviously unworkable or implausible policies may evidence a lack of sincerity. A party that continues to deny the human impact on the global environment and its potential implications for future people, despite all the evidence to the contrary, is probably a faction because it is refusing to consider the interests of the future people, and its policies are unlikely to be justifiable to them. The one exception to the sincerity and plausibility test is this: even if a ‘party’ sincerely believes that the only interests that count are the interests of a sub-section of the people rather than those of all the people, its sincere rejection of democracy as rule-of-all-the-people is not enough to dodge its characterisation as a faction.

The hardest case is perhaps that of a party that claims that the sub-set of the people it represents is a distinct people – the Scottish National Party is a good example. Clearly, a full account of factionalism requires an understanding of who count as ‘a people’, as distinct from ‘a nation’ and ‘a citizenry’. What may be relevant to the determination is whether the party at least represents all the people in the constituencies in which it puts up candidates for elections. If it does, it is less likely to be a faction, if it doesn’t, more so. Perhaps the difficulty of this example demonstrates to us that it may well be parties, rather than constitutions, that constitute ‘a people’. The shared experience of a people as an audience for a party’s ideological and policy platform, in whose name and for whose benefit the party seeks to act, may well be constitutive. The failure of Europe to emerge as a robust polity may be owed to the absence of effective pan-European parties. This people-constituting feature of parties may well be the chief reason why factions are so problematic – they seek to redefine who the people that constitute a polity are by excluding a component thereof. Such redefinition may well be justified sometimes, such as when there is a legitimate right for internal or external self-determination. But outside this context, a majoritarian faction that excludes a minority from its concern is effectively seeking to exclude it from the polity by reconstituting its people on narrower, sectarian, terms.

The absence of a bright line dividing parties from factions may especially bother legal scholars: their fears are unfounded. I have already argued that apart from the application of anti-discrimination laws to parties, the anti-faction principle is best implemented by thorough second-order regulation: by creating a constitutional architecture – especially through the choice of an appropriate electoral system – which makes centrifugal factional behaviour politically less rewarding. Ranked-choice voting, cumulative voting, approval voting, and other such models, for example, are known especially to reward centripetal parties.¹⁰³ Such a regulatory architecture need never require a court to decide whether a party before it is a genuine party or a faction. This is precisely why the anti-faction principle demands that states ‘encourage’ anti-factionalism, unlike the party-state separation, which they ought to ‘ensure’. It is best implemented through political nudges rather than legal command-and-control. While the anti-faction principle may not be necessary for satisfying a wholly procedural conception of a democracy, only a slightly thicker conception of democracy is needed to appreciate that, for example, the capture of the ruling party/coalition by a small set of wealthy donors transforms a democracy into a plutocracy.¹⁰⁴ The principle is compatible with even an interest-bargaining model of democracy – so long as the interests of all social groups matter in the bargain; it does not necessarily require an endorsement of the more demanding republican or deliberative conceptions of democracy.

4.4 Conclusion

In this chapter, I have provided an idealised account of the functions of a political party in a healthy democracy that can help diagnose the ailments of unhealthy parties and party-systems. That account emphasises their Janus-faced role as intermediaries between the state and its people, which they perform by lowering key information and transaction costs in a democracy. Parties are therefore simultaneously public and private. Party systems that successfully reduce political participation costs, voters’ information costs, policy packaging costs, and ally prediction costs grease the wheels of representative democracy and are indispensable to its smooth operation. In order to aid parties in performing their intermediary function well, constitutions should seek to optimise four key principles in relation to political parties. First, they ought to protect the purposive autonomy of parties and align their rights and duties closely to their hybrid public-private character. Second, constitutions should optimise the number of parties such that there are enough parties to represent every salient voter-type, but not so many that voters’ information costs become unaffordable. Third, constitutions should ensure the separation of the parties from the state so that no party is able to entrench itself in the institutions and offices of the state. Breach of this principle increases the political participation costs of the supporters of opposition parties. Finally, the anti-faction principle requires that constitutions should encourage parties to cater to the interests of all the people, rather than merely of a sub-section thereof. Factional parties increase the political participation costs of excluded minorities. They also make policy packaging difficult.

Needless to say, these principles are compatible with many different ways of designing a constitution, but they are not compatible with all of them. How they apply precisely will obviously depend on the context. Note also that these principles may sometimes pull in opposite directions. For example, the anti-faction principle can be in tension with the party system optimality principle if a salient voter type demands factional representation. It is also in tension with the party state separation principle inasmuch as it demands that factions are structurally disadvantaged in the design of the state’s political institutions. The goal is therefore to attempt their collective optimisation rather than any one principle’s maximisation. In a separate article,¹⁰⁵ I have shown one set of institutional arrangements that might achieve such optimisation.

Political parties are the life-blood of representative democracy. Proposals seeking to respond to democratic deconsolidation have seen them as the problem, unnecessary middlemen who should be cut out of the system entirely in favour of direct democracy. These proposals fail to acknowledge the important function that parties perform in lowering key information and transaction costs that all-but-the-smallest democracies must contend with. If democracy is to survive, political parties need to be supported and improved, not eliminated. Despite all our advances in democratic technology, Schattschneider’s famous claim that ‘modern democracy is unthinkable save in terms of parties’ remains as true as it ever was.¹⁰⁶ Hence the four political principles that I argue should inform constitutional designs of democracies. A party-less democracy, unless really small, is unlikely to remain democratic for very long.

5 The Constitutionalization of Parties and Politics

Tom Ginsburg and Mila Versteeg

5.1 Introduction

“War,” said Ross Perot, “has rules. Mud wrestling has rules. Politics has no rules.”¹ This view captures an increasingly common view of the political arena as one in which all bets are off, and any attempt to constrain the players is doomed to fail.

A moment’s thought, however, exposes how facile that view is, at least for democracies. The metaphor of politics as a contest that takes place in a public arena is an evocation of a repeated game, in which the rules are themselves constitutive of play. Democratic politics may be dirty, but it relies on a set of structures that provide for competition, and its continued maintenance depends on some consensus on these underlying rules.

While the idea that politics has rules is probably uncontroversial, few have noticed that, over the past decades, these rules are increasingly specified in the text of the constitution itself. This was not always the case. Constitutions used to be mostly silent on key issues like the regulation of political parties, voting rights, or the details of holding and administering elections. But over the past decades, many constitutions have come to regulate core aspects of the democratic process, including political parties, details of elections and the electoral system, and voting rights. To illustrate, according to our data, 83 percent of constitutions in force today regulate political parties, 71 percent give a court or electoral body the power to oversee elections, and 44 percent specify rules for electing the lower house. Indeed, we find a sharp rise in the constitutionalization of democracy in three related areas: (1) the regulation of parties, (2) rules relating to voting and direct democracy, and (3) rules relating to administering elections and their oversight.

This is a profound change, with potentially significant consequences. Constitutions are usually entrenched, meaning that they cannot be changed by ordinary democratic majorities, but instead require larger thresholds of legislative support for approval. Most constitutions also envision the practice of judicial review, meaning that courts can invalidate laws and regulations that contradict the constitution.² Constitutionalizing the rules relating to the democratic process, therefore, means that the rules of the democratic game will be harder to change and are subject to judicial oversight. This, in turn, implies a judicialization of democratic politics – core decisions about democracy are now made by constitutional and supreme courts.

Because of the nature of constitutions, the constitutionalization of democracy can aid democratic practices. After all, entrenching democratic rules and making them subject to judicial oversight means that ordinary majorities cannot tinker with them to create partisan advantage. Because of these qualities, constitutionalizing democracy can serve a hands-tying function; it allows constitution-makers to double down on their commitment to maintaining democracy.³ In addition, writing down specific rules can also provide clarity on the rules of the game.⁴ Such clarity is especially important when political conventions are weak, as is often the case during democratic transitions.

We find some evidence to support these ideas. We show how constitutional provisions protecting democracy served a clear hand-tying function in Kenya and that these same provisions appear to have helped prevent democratic backsliding. Looking at cross-national data, we find that constitutionalizing democracy is correlated with higher levels of democracy, which is consistent with the idea that constitutional rules on democracy can help protect democracy.

But we also add a note of caution. Constitutionalizing democracy gives an important role to high courts, in that they get to act as umpire over the rules of the democratic game. This might be a democracy-enhancing feature when courts are independent. But when courts are captured by the ruling coalition, they can interpret these same provisions in ways that are inconsistent with democracy.⁵ This risk is especially present for provisions that give substantial discretion to courts, such as bans on undemocratic parties. We illustrate this possibility with the case of Thailand, where the constitutional provisions banning certain kinds of parties has been deployed by the Constitutional Court to ban democratic parties. Overall, we conclude that, while promising, constitutionalizing the rules of democracy is not a panacea to prevent democratic erosion. Under some conditions, particularly when the military plays a strong role in politics, constitutional provisions may have the opposite effect from their stated purpose of protecting democracy.

5.2 The Constitutionalization of Democracy

One general trend in constitutional design over recent decades has been articulation: a growing propensity to regulate constitutionally with greater levels of specification. Constitutions written today tend to cover an ever-growing list of topics and deal with these topics in substantial detail. This trend has also been described as a move toward “constitutional codification,” or growing “constitutional specificity.”⁶ The trend toward constitutional articulation affects many different areas of constitutional law, including constitutional rights, judicial power, foreign policy, the separation of powers, and “fourth branch” institutions. Here, we explore the same with respect to some of the core ingredients of democracy: political parties, voting, and elections.

5.2.1 Data

To map and explore the constitutional articulation of core aspects of democracy, we draw on data from the Comparative Constitutions Project to select some two dozen variables relating to democracy. We only selected those that arguably reflect rules compatible with democracy. To illustrate, we do not include provisions that ban specific parties or declare a one-party state; even though these provisions surely deal with parties and elections. At the same time, we do include provisions relating to militant democracy that ban certain types of parties and empower certain institutions to ban them. While one might argue that such rules are undemocratic, their goal, at least in theory, is to protect democracy.⁷

A first set of variables relates to the role of political parties: whether (1) the constitution refers to political parties; (2) creates a right to form political parties; (3) bans certain types of political parties (but not specific parties); (4) whether the legislature is given the power to ban certain types of unconstitutional parties (but not specific parties); (5) whether the constitutional court, supreme court, or the judiciary is given the power to ban certain types of unconstitutional parties (but not specific parties); (6) whether an electoral court or electoral commission is given the power to ban certain types of unconstitutional parties (but not specific parties); (7) whether the constitution guarantees equality of political parties; and (8) whether the constitution specifies that political parties can initiate general legislation.

A second set of variables relate to voting and direct democracy: (9) whether the constitution makes claims of universal suffrage; (10) whether there are any restrictions placed on the right to vote; (11) whether the constitution makes voting mandatory; (12) whether the constitution prescribes electoral ballots ought to be secret; and (13) whether the constitution gives individuals the ability to propose legislative initiatives.

A third, related set of variables describes to the electoral systems and the mechanics of elections. They include: (14) whether the constitution establishes a voting threshold of a certain proportion of the votes for a party to be able to take a seat in parliament; (15) whether ordinary (constitutional or supreme) courts have the power to supervise election; (16) whether an electoral commission, electoral court, or both, exist to oversee the election; (17) whether the constitution makes arrangements for scheduling the elections; (18) whether a specialized body (and not the executive or legislature) establishes the shape and size of electoral districts; (19) whether the constitution has provisions on the public financing of campaigns; (20) whether the constitution has provisions setting limits on the money used for campaigns; (21) whether the constitution prescribes the election timing for executive and legislature (either same or different days); (22) whether the constitution specifies the electoral system for the lower house; and (23) whether the constitution specifies the electoral system for the upper house.

5.2.2 Trends

A cursory look at the data reveals a growing constitutionalization of core features of democracy. Figure 5.1 shows the average number of all twenty-three democracy provisions over time. It shows a gradual increase in the number of electoral provisions in national constitutions. In 1810, the average number of provisions was 0.3; today, it is 6.9. But of course, there is substantial variation across countries. Even today, some constitutions do not enumerate any provisions relating to democracy, such as the absolute monarchies of Saudi Arabia, Brunei, and the United Arab Emirates.

Figure 5.1 Number of democracy provisions on world map.

The US Constitution today includes only two democracy provisions – restrictions on abridging the right to vote under the Fifteenth, Nineteenth, and Twenty-Sixth amendments, which we code as being a claim to universal adult suffrage. But the US Constitution is notoriously silent on parties. In fact, the founders of the American republic sought to create a system of government that would protect liberty by retarding the formation of parties. As James Madison put it in Federalist 10, factions were a danger to popular government. These entities, composed of “some common impulse or passion, or of interest, adverse to the rights of other citizens,” were an evil to be avoided through careful institutional design.⁸ Despite Madison’s best intentions, however, parties emerged early in the Republic as useful mechanisms to coordinate behavior in the legislature. As American democracy evolved and expanded in the nineteenth century, the party system changed as well, but constitutional language did not keep up. The Constitution was amended to provide for nondiscrimination in the provision of rights to vote for racial minorities, and later women and youth, but parties remain absent from the text. The time, place, and manner of congressional elections remain in the hands of state legislatures, which, contrary to Madison’s expectations, have become hotbeds of partisan self-dealing. In other settings, such questions are increasingly taken out of the realm of partisan politics and placed into the constitution.

On the other end of the spectrum, Kenya, starting in 2010, is the country with the largest number of constitutional democracy provisions. Its constitution contains no fewer than fifteen such provisions. It is followed by Liberia and Thailand (from 2007–2012), which both have fourteen democracy provisions.

When depicting the same data on a world map, we can see some regional trends, with a notable concentration of a high number of such provisions in both Latin America and Africa, which are areas with a good deal of constitutional turnover.

We see the same trends if we look at each of these categories separately. Consider, first, political parties. At the turn of the twenthieth century, exactly two constitutions referred to political parties: those of Colombia and Greece.⁹ Today, 83 percent of the 193 national constitutions in force contain such a reference. As of 2020, 29 percent of constitutions prohibit certain parties or types of party programs, and some 13 percent provide that a court determine whether a political party is unconstitutional. These phenomena were unknown in 1900. Figure 5.2, Panel A shows the increase of provisions on political parties over time, a trend that accelerated after World War II.

Figure 5.2 Average number of provisions relating to parties, voting, and elections.

We observe the same basic trends for voting and direct democracy. Figure 5.2, Panel B depicts the average number of provisions relating to voting. It is worth noting that these provisions are older than the political party provisions. We see them start appearing in the 1800s and steadily increase in number over time.

Finally, we also see the same trend for features of the electoral systems. Figure 5.2, Panel C depicts the average number of these provisions and reveals how they have increased over time. One increasingly popular constitutional design choice is to establish some form of electoral commission. This is a distinct body (or several) established to manage tasks like boundary delimitation and election management. In 1900, only Colombia and Liechtenstein provided for an electoral commission or/and court to oversee elections, but today 71 percent of constitutions constitutionally establish such a commission or give courts the power to oversee elections. Another noteworthy trend is to clarify the electoral system for parliamentary elections. In 1900, the four constitutions that clearly specified a voting rule for the lower house had a simple version of plurality or majority rule. Because of significant innovation in the understanding and design of voting systems during the twentieth century, contemporary constitutions not only have to choose between majoritarian and proportional representation but will sometimes specify a remainder formula, identify special constituencies for particular groups, or create a complex mixed system involving both proportional representation and districts. The five constitutional articles describing Sri Lanka’s mixed method electoral system for the lower house are nearly as long as some entire national constitutions.¹⁰ They illustrate the tendency to constitutionalize the specifics of parliamentary elections.

5.2.3 Relationship with Democracy

Are these new constitutional features associated with regime type? The data reveals that these features can be found in democracies and autocracies alike, although they are somewhat more common in democracies.

We can observe this by analyzing the trends for democracies and autocracies separately. Figure 5.3 depicts the growing constitutionalization of democracy for autocratic and democratic regimes (we consider a country to be democratic if its polity2 democracy score is over 4 [on a scale from -10 to 10]). The dashed line denotes democracies; while the dotted denotes autocracies and solid captures the full sample. The first panel shows the trends for all democracy provisions, the second panel for parties, the third for voting, and the fourth for the electoral process. Overall, the graphs reveal that the trend toward constitutionalization is slightly more pronounced in democratic countries. By 2016, the average democracy constitutionally enumerated 7.7 of the variables we collected, while the average autocracy enumerated 6.5. But Figure 5.3 also reveals that this difference is mostly driven by constitutional provisions relating to the electoral process. By contrast, the second panel reveals that the constitutionalization of parties is just as common in democracies as autocracies.

Figure 5.3 Average number of provisions relating to parties, voting, and elections by regime type.

Is the constitutionalization of democracy associated with higher levels of democracy de facto? It is notoriously difficult to sort out causation with cross-national data, and what follows is merely an initial exploration of whether constitutional provisions on democracy correlate with democracy de facto.

To explore the relationship between de facto democracy and constitutional rules relating to democracy, we estimate a simple OLS panel regression with the well-known polity2 democracy scale as the dependent variable. The model includes country-fixed effects, year-fixed effects, and a linear time trend. With these, we control for non-time varying country characteristics, common trends, and global shocks. We also experiment with adding a standard set of control variables: GDP per capita, population size, and civil war. Robust standard errors clustered are clustered by country to account for serial correlation.

Table 5.1 reports results from this exercise. It reveals that correlation between constitutionally entrenched rules relating democracy and de facto democracy is positive and statistically significant. For all the models capturing all provisions (columns 1 and 2), adding one additional constitutional feature (out of twenty-three) on democracy is associated with an increase of 0.455 on the 21-point democracy scale. While we should not interpret these as causal relationships, they are suggestive that these constitutional features are associated with somewhat higher levels of democracy.

Table 5.1 Relationship with democracy

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
all provisions	0.455***	0.426***
	(0.100)	(0.122)
parties			0.585**	0.515**
			(0.252)	(0.259)
voting					0.925***	0.984**
					(0.333)	(0.439)
elections							0.707***	0.706***
							(0.180)	(0.225)
controls	no	yes	no	yes	no	yes	no	yes
country FE	yes	yes	yes	yes	yes	yes	yes	yes
year FE	yes	yes	yes	yes	yes	yes	yes	yes
time trend	yes	yes	yes	yes	yes	yes	yes	yes
Observations	12,849	8,712	12,858	8,721	12,893	8,756	12,889	8,752
R-squared	0.707	0.731	0.698	0.723	0.701	0.728	0.705	0.731

Note: the dependent variable is democracy. ***p < 0.01, **p < 0.05, *p < 0.1.

5.3 Why Constitutionalize Democracy? An Exploration

Hands-tying. What motivates drafters to regulate core aspects of democracy in the constitution? A likely reason is that constitutions are thought to serve as “pre-commitment devices.”¹¹ The key idea here is that, when certain rules or values are constitutionalized, they are placed outside of the reach of ordinary politics. Because constitutions are harder to change than ordinary laws, ordinary democratic majorities cannot change constitutional rules. Additionally, courts are usually empowered to enforce these rules and to invalidate laws that contradict them. This also means that when disputes over interpretation or application arise, courts can further clarify and enforce these rules. The combination of entrenched and justiciable rules makes these rules harder to undermine and allows drafters to tie the hands of future democratic majorities.

This hands-tying logic reflects a certain amount of distrust of those who will govern under the constitution. Such distrust is often present during democratic transitions, when democracy is not taken for granted and constitution-makers use the constitution to attempt to lock in their commitments to a liberal democratic order.¹²

Under such conditions, creating a new machinery to run elections and to oversee democratic competition will make good deal of sense. The constitutionalization of parties likewise can serve a hand-tying function. The idea that constitutions should regulate and protect political parties can be traced back at least to Hans Kelsen, who noted that it is crucial “to anchor political parties in the constitution and give legal form de facto to what they have long since become: organs forming the will of the state.”¹³ On the one hand, modern constitutions protect parties. They do so through a right to form political parties, which ensures that constitutional mechanisms are available to prevent existing players from monopolizing the machinery of politics. Party rights usually also offer protections for smaller parties, preventing the state from intervening in a party’s internal affairs or making party registration requirements too arduous.¹⁴ In some cases, it might also entail certain guarantees for opposition parties to be represented in the democratic process.¹⁵ Empirical evidence suggests that rights to form parties are among those rights that are most likely to be effectively enforced, in part because parties themselves use these rights to protect their interests.¹⁶

But constitutions do not only protect parties; they also regulate them. The most common approach is to ban parties that propagate certain viewpoints. The practice of requiring parties to be democratic became relatively common after World War II. The rise of Nazism through mechanisms of parliamentary democracy led to great angst, and in 1937, the German political scientist Karl Loewenstein coined the term “militant democracy.” Concerned with the inadequate democratic response to the rising threat of fascism, he called for a set of legislative and legal techniques that would allow democracy to defend itself against threats that emerge from within. “Constitutional scruples” he noted, “can no longer restrain from restrictions on democratic fundamentals, for the sake of ultimately preserving these very fundamentals.”¹⁷ Loewenstein went on to catalog techniques used by inter-war drafters that could constrain autocratic elements within society. But the prototype instrument of a militant democracy limits parties to those that are democratic. The design strategy of combining party rights with party bans, then, seems to reflect a simultaneous desire to recognize parties as important institutions of governance but also to set some boundaries on the ideological playing field.

Coordination. Another reason why constitutionalizing the rules of democracy can aid democratic practices is that constitutions can aid coordination around the rules of the game. If everyone has common understandings of the rules of politics, there is little additional advantage from writing them down. But there may be times when existing understandings break down, either because of exogenous change or because of the escalation of partisan competition. This can create incentives to clarify the rules of the game, so that parties can coordinate better in the future.

Coordination is especially important in the context of democratization. In such a context, new systems of elections and voting are often being set up from scratch. When starting from scratch, clarifying the basic rules of the game is particularly important as there are no preexisting political conventions to rely upon.

In a similar vein, coordination also becomes important when democratic political process become more open in character. The nineteenth century saw the rise of political parties around the world, be they the Colorados and Blancos in Uruguay, the Liberals and Farmers parties in Sweden, or the Liberals and Conservatives in Canada. What these party systems had in common at this time is that they are largely elite affairs, in which parties regulated their interactions through reciprocity. But industrialization brought demands for an expansion of the franchise. As mass-based parties emerged out of labor movements in the early twentieth century, there was a sharp rise in polarization, putting pressure on norms of reciprocity among parties. In such a context, it made sense to start clarifying the rules of the political game and the role of parties within it. As democratic politics became more open and polarized, having clear rules helped parties to coordinate and channel their competition.

The explanations of hands-tying and coordination are not mutually exclusive, and likely work together in specific cases. For example, in the context of democratization, constitutionally specifying rules of the democratic process can be motivated by a joint desire to pre-commit to multiparty democracy as well as to clarify basic rules. Indeed, it is likely no coincidence that we see the constitutionalization of democracy take off during the wave of constitution-making in the 1990s.

5.4 The Case of Kenya

We can see the importance of distrust of political elites in the constitution-making process undertaken in Kenya in the early 2000s, culminating in the Constitution of 2010. As noted above, this constitution is noteworthy because it has the largest number of provisions dealing with democracy of any constitution in force today. It is therefore worth probing the motivations behind this document and to explore how these provisions have been used in practice.

The Kenyan constitution-making process had followed an uneasy decade of democratization, during which longtime strongman Daniel Arap Moi retired in 2002. The prior text, adopted at independence in 1963, included an Electoral Commission, but had no mention of parties, and left the rules about voting to ordinary legislation. This document governed a rigged political process, and the Electoral Commission had no financial autonomy. Mass action in 1997 by civil society led to a constitutional amendment increasing the number of electoral commissioners and giving political parties more say in their selection. In addition, a Constitution of Kenya Review Act sought to “facilitate the comprehensive review of the Constitution by the people of Kenya.”¹⁸ This required the creation of the Constitution Review Commission (CKRC) to consult broadly and produce a draft. The draft would then be debated at a National Constitution Conference (NCC), made up of politicians and civil society, before going to parliament and the public for final approval.

Headed by famous civil society activist and scholar Yash Pal Ghai, the CKRC produced a draft after wide public participation. A summary of the public views presented to the CKRC included various criticisms of political parties: there were too many of them, they were too autocratic, and without internal democratic procedures; they were vehicles for self-interest; they engaged in violence and hooliganism; and they did not seek to advance the national interest.¹⁹ The solution, according to the report of the CKRC, was that “Political parties, as institutions of democratic and republican governance … should be regarded as constitutional organs that should be provided for in and regulated by the Constitution.”²⁰ Accordingly, the draft finally approved by the NCC in 2004 (known as the Bomas draft after the location of the NCC meetings) contained a major section on political parties, with rules, regulations, and provision for a code of conduct, as well as a fund for campaign expenditures, which included limits on party activities. Ghai and his colleagues sought to limit parties’ ability to engage in corruption and to steer their activities toward democratic representation. In addition, the draft also fixed a time for parliamentary elections and introduced a clear right to vote.

However, as the process had been designed, the parliament had a chance to modify the draft before it went to public referendum for approval. Attorney General Amos Wako, a consummate political insider, was able to take control and watered-down key provisions in a process negotiated by the governing political parties without the opposition. The political-parties section shrank from ten articles to just two. The executive branch was made stronger. Civil society was outraged at this perceived capture of the draft by politicians, and the Wako version was rejected by public referendum in 2005 by a resounding vote of 58 to 42 percent. Subsequent elections were marred by significant violence, reflecting the largely tribal organization of Kenya’s existing political parties. The country stood on the brink of unraveling, and the international community came in to broker a reset.

A second constitution-making process followed, in which initial drafting was done by a nonpartisan Committee of Experts, which included three foreigners. This body was charged with drawing the best of the earlier proposals into a new harmonized draft, which would presumably reflect a consensus. Its members were well aware of the need to reflect both public demand and elite consensus in this process.²¹ The dynamics were similar to the prior episode, in which the Committee proposed a draft with an extensive section on political parties, but this was then submitted to a Parliamentary Select Committee, which reduced the number of articles on parties from nine to two. The draft then went back to the Committee of Experts for presentation to the public. The draft was ultimately approved by a referendum in August 2010, with more than 2/3 of voters supporting adoption, in a context of great distrust of politicians.

The resulting document included several global innovations, including a whole chapter on leadership and integrity, and established good governance, integrity, transparency, and accountability as national values. More specifically, the document creates a body to manage elections and oversee politics – an Independent Electoral and Boundaries Commission – and requires public participation in the process of delimitation of districts.

Even though the number of provisions regulating parties were reduced by politicians, the final document regulates parties substantially. Notably, the legislature is instructed to enact legislation regulating political parties, and Article 91 contains myriad constitutional rules about parties. These include requirements that parties have a national character and that they not be founded on ethnic, linguistic, regional, or other lines; that they have internal democracy in terms of their governing bodies and leadership; that they abide by principles of good governance; respect human rights, including those of minorities and marginalized groups to participate in the process; observe a code of conduct for parties; and several others.²²

Overall, these new provisions can be best understood as efforts to tie the hands of politicians on behalf of the public, to constrain and channel political activity in beneficial ways. For parties, the provisions allowed for better coordination among themselves on the rules of the game: knowing that the rules were present, parties had some incentive to “play fair” under the new framework, rather than escalating competition in ways that violated the terms of the constitutional text.

Did the 2010 Constitution’s strategy of “spelling things out” strategy work? Although Kenya’s political class is among the world’s most venal – the first act of parliament after approving the draft of new Constitution was to vote a pay raise for members – the text seems to have had an impact on the political system. A recent attempt by then-President Uhuru Kenyatta and his former rival Raila Odinga to cartelize the political system by unilaterally creating seventy new parliamentary constituencies and a new prime minister position was challenged by voters and minor parties. Among other provisions, the proposal – known as the Building Bridges Initiative – restricted judicial review of the boundary delimitation process and utilized a provision on popular initiative to propose the reforms. The courts rejected the attempt as procedurally flawed and incompatible with the Constitution, thereby foiling a partisan lock-up.²³ The High Court judgment in particular emphasized the role of the scheme in undermining the political process and the independence of the Commission, thus infringing on the rights of smaller parties and undermining the basic scheme of the Constitution. In August 2022, Odinga lost the presidential election to William Ruto, and outgoing incumbent Kenyatta peacefully turned over power to the latter – the first orderly alternation under the new charter.

To conclude, the Kenya story illustrates the democratic politics of constitutionalization: pressures for greater democratization led to the introduction of new ideas for regulation and strengthening constitutional protections for regulators. Both dimensions of parties and electoral mechanics were constitutionalized. The actual process took multiple iterations and involved civil society, technocrats, and sitting politicians, all under the supervision of courts at the phase of implementation. A major attempt to capture the process by two large parties, begun in 2020, was decisively rejected by the courts, using the constitutional text. In short, Kenya’s democracy is far from perfect but no doubt in much better shape than when Moi stepped down two decades ago, and the constitutionalization of democracy helped to secure this.

5.5 A Note of Caution

One important implication of constitutionalizing core features of democracy is that it empowers courts to police the boundaries of democratic politics. This is by design: when democracy is constitutionalized, constitution-makers envision that a neutral umpire can enforce constitutional boundaries in the face of partisan conflicts. Thus, high courts can now call out violations and clarify the rules of the democratic game when there is ambiguity on what these rules require.

When courts are independent, constitutionalizing the rules of democracy might make a whole lot of sense to protect democracy. But the central role for the court also comes with a vulnerability, which is that these same constitutional provisions can be used to accomplish partisan goals when courts are not independent. When the rules of democratic politics are constitutional in nature, then controlling a court becomes a powerful tool that allows would-be autocrats to stack the deck in their favor. Not only can courts reinterpret constitutional rules on democracy, their rulings enjoy priority over ordinary laws, which means that ordinary democratic majorities cannot simply alter them.

The pay-off of controlling the court might be highest for the provisions that represent broad and flexible standards. It is more difficult (though not impossible) to reinterpret clear rules, such as the day on which elections are to be held.²⁴ But other provisions give more discretion to courts. The ban on undemocratic parties is a prime example: it leaves to the courts to judge which policies and programs are undemocratic. When courts are captured, constitutional bans on undemocratic parties can become a tool to lock in current configurations. For example, in 2017, the authoritarian government of Cambodia had the Supreme Court disband the country’s main opposition group, the National Rescue Party, leaving strongman Hun Sen in control for an upcoming election. Individual members of the party were subjected to a five-year ban from running for office, as Hun Sen is reputed to be grooming his son to succeed him.

Thus, when courts become central to the functioning of democratic politics, it raises the stakes of controlling them. In the recent wave of democratic erosion, one of the first moves of would-be authoritarians is to control courts. When core features of democracy are constitutionalized, this only adds to the incentives to control courts.

Another factor that may lead to abuse of these provisions is a powerful role for the military in politics. Militaries often see themselves as the guardians of a set of fundamental values in the constitutional order, be they secularism (as in Turkey), religion and property (as in Chile under Pinochet), or the territorial integrity of the country (as in Pakistan or Nigeria). This self-conception can lead them to closely monitor civilian politics, even in a democratic era. They often align with judiciaries, with whom they share a hierarchical organization and technocratic orientation.²⁵

There is a long history of military-aligned governments seeking to use the machinery of party bans to maintain power and shape politics. Perhaps the paradigm case is Turkey, which banned Islamist parties several times before the eventual success of Recep Tayip Erdogan. The case of Thailand illustrates how provisions constitutionalizing democracy can be a double-edged sword. As in Kenya, the Thai Constitutions of 1997 and 2007 contain a large number of provisions regulating democracy. But in the Thai case, the tools of constitutional regulation have been used to constrain democracy itself on behalf of military-bureaucratic institutions that are at the core of the political system.

Thailand is a country with a long history of alternating between corrupt civilian governments and military regimes, and in 1997, a new constitution was adopted to try to end this cycle. Known as the “People’s Constitution” because it came out of a popularly elected constitutional assembly, the 1997 Constitution was the most democratic in Thai history. It contained numerous rights, political reforms to facilitate electoral majorities, and a network of “post-political” institutions to constrain and clean up democratic politics. Central here was a new Constitutional Court. Under Section 63 of the Constitution, the Court was given the power to disband any party found to be seeking to gain power by illegal means or to overthrow the “democratic regime of government with the King as Head of State.” This is an old formula in Thailand, which seeks to balance the formal requirements of democracy with a semi-divine monarchy.²⁶ We do not know who pushed for the inclusion of these provisions, but it likely reflected a combination of those with a genuine motive to protect democracy along with elements of the conservative bureaucratic-military elite.

Populist billionaire Thaksin Shinawatra dominated subsequent elections and held the premiership from 2001 to 2006, but conservative forces resisted him. Polarization increased sharply, and by the mid-2000s competing political factions were engaged in open and violent struggle in the streets of Bangkok. The Constitutional Court had dismissed corruption charges against Thaksin to allow him to take power in 2001, and sought to maintain a neutral role, but the rising political turmoil led the country’s revered monarch to demand that courts address the problems. In 2006, a military coup sent Thaksin into exile. The Constitutional Court subsequently took up a case against Thaksin’s Thai Rak Thai party, eventually deciding that it was banned under Section 63 of the Constitution. Polarization had drawn the courts into one side of a partisan divide.

The 2007 Constitution, produced by technocrats working with the military government, drew on the basic structures of 1997, with multiple regulatory institutions to govern elections, and contained an elaborately specified electoral system. It retained much of the language of the 1997 constitution with regard to parties, including the power of the Constitutional Court to ban parties under Section 68.²⁷ Section 65 of the 2007 document added new language, worth quoting in full:

A person shall enjoy the liberty to unite and form a political party for the purpose of making political will of the people and carrying out political activities in fulfillment of such will through the democratic regime of government with the King as Head of the State as provided in this Constitution.

The internal organization, management and regulations of a political party shall be consistent with fundamental principles of the democratic regime of government with the King as Head of the State.

Members of the House of Representatives who are members of a political party, members of the Executive Committee of a political party, or members of a political party, as to the number prescribed by the organic law on political parties shall, if of the opinion that their political party’s resolution or regulation on any matter is contrary to the status and performance of duties of a member of the House of Representatives under this Constitution, or contrary to or inconsistent with fundamental principles of the democratic regime of government with the King as Head of the State, have the rights to refer it to the Constitutional Court for decision thereon.

The motive for this section seems to reflect a deep distrust of political parties on the part of military technocrats, turning party members into monitors. When a member finds that the party leadership is acting outside the bounds of Thailand’s democratic system, they can refer the case to the Constitutional Court. Section 68 goes on to provide the Constitutional Court with the power to disband unconstitutional parties for actions incompatible with the democratic regime with the King as Head of State. More broadly, the courts generally serve as guardians to limit the domain of politics, with power to strike resolutions of parties and channel their activities. One might describe this section as abusive from the outset – an attempt on the part of military leaders to control democratic politics.

The 2007 Constitution quickly became an instrument to limit democratic politics. Despite having left the country after his prior party was banned, Thaksin Shinawatra did not disappear as a political force. His allies reorganized Thai Rak Thai into a new People’s Power Party, but this too was banned by the Constitutional Court in December 2008. Another successor party, Pheu Thai, won subsequent elections, and Thaksin’s sister Yingluck took over as prime minister. Conflict continued, occasionally erupting in violence. While the courts had in the early years of Thaksin’s ascendance sought to play a neutral role, they gradually took on the task of enforcing the demands of Thailand’s traditional conservative elite against populist challengers. By 2014, a new coup sought to restrict the domain of elections and parties even further. Thaksin remained out of the country and seemed to have been defeated through the efforts of multiple party bans along with two military coups.

This meant that there was something of an opportunity for a fresh start in Thai politics. When the new Constitution of 2017 was adopted, several new parties formed to compete. A leading challenger was the Future Forward Party, which represented urban and young voters, and was led by a scion of a billionaire family, Thanathorn Juangroongruangkit. When this party proved to be fairly popular, however, it came to be seen as a threat to the party led by coup leader Prayuth Chan-Ocha. On relatively thin evidence, the Thai Constitutional Court found that Thanathorn had violated campaign laws and removed him from his seat in parliament in 2019. The next year, the Future Forward Party was disbanded by the Court. Its successor, the Move Forward Party that won an electoral victory in 2023, is currently facing a number of complaints asking for its dissolution.

The Thai story illustrates that constitutionalization of parties can be a two-edged sword and puts a great deal of faith in those who will make determinations of compliance with core norms. Even if adopted to protect democracy initially, constitutional provisions can be reinterpreted and repurposed. Party bans raise the stakes of controlling a constitutional adjudicator and increase the temptation to abuse the power. As always, when the institutions charged with protecting democracy become aligned with one faction in local politics, there is great potential for abuse.

5.6 Conclusion

Whether constitutionalizing democracy will serve a genuine hands-tying function or instead be subject to abuse will depend on context, and this can change over time within a single constitutional order.

In the case of Kenya, the ability of the courts to rely on constitutional principles and procedures was central to rebuffing an attempt by the two largest parties to cartelize the political system and to facilitating a transfer of presidential power in 2022. The constitutionalization of elections and parties has led to a vigorous jurisprudence, which simply would not have been possible absent the rules laid out in the 2010 Constitution . The broader positive correlation between constitutional regulation and measures of de facto democracy gives us reason to believe that this story is not confined to Kenya but that we might see similar dynamics at work in other cases such as Liberia and Nepal, which both have high levels of regulation.

However, constitutionalization is not a panacea. As our Thailand case study illustrates, the presence of provisions on elections, voting, and parties in a constitution can in practice be abused. In the wrong hands, constitutionalization might make things substantially worse for democracy. Hard limits on the political process are themselves a double-edged sword and may be subject to abuse by military-aligned governments in particular. This implies that, when considering whether or not to include constitutional provisions on the political process, democrats have to focus on who will be interpreting them down the road, and to choose the provisions least likely to be abused

6 Tackling Winner-Takes-All Politics in Africa Inclusive Governance through Constitutional Empowerment of Opposition Parties

Adem Kassie Abebe

6.1 Introduction

Constitutions set the fundamental ground rules for democratic contestation and therefore influence the nature and outcomes of the dispensation. It is no wonder that the ground rules can be a point of serious contest among political groups, with dominant groups at times seeking ways to alter or entrench rules that favour their chances of victory. To check this self-serving temptation, the broad recognition, internalisation, acceptance and entrenchment of the ground rules and their relative stability and insulation from undue alterations and manipulations is critical for the legitimacy of democratic political systems.

Central to these ground rules are constitutional provisions guaranteeing multi-party democracy, universal suffrage and the right to stand for elections, rules governing elections (for the executive and legislature) and those regulating the establishment, operation, and rights and obligations of political parties. In connection with these, constitutions in Africa also often provide for election management bodies and courts to ensure the integrity and protection of electoral rights and multi-party democracy, including constraints on self-serving and capricious constitutional reforms. The constitutional recognition of these fundamental aspects of constitutional democracy is founded on the realisation that constitutional designers cannot simply assume that fair rules would be adopted and maintained by players with direct interest in the outcome.

Virtually all African constitutions directly or indirectly guarantee multi-party democracy and provide for the right to vote of citizens – even the 2005 Constitution of Eswatini (Swaziland), Africa’s only absolute monarchy, recognises the right to vote of all adult citizens (article 85).¹ Moreover, while the African Charter on Human and Peoples’ Rights (African Charter), does not specifically mention the right to vote, it guarantees the right to political participation (article 13), which has been interpreted to include the right to vote and stand for elections. The 2007 African Charter on Democracy, Elections and Governance (ACDEG) specifically recognises the right to participation through universal suffrage (article 4(2)).

To be sure, the right to vote has generated debates and litigation, particularly in relation to the right of prisoners to vote, which is increasingly recognised across the continent, on many occasions as a result of court decisions.² African constitutions also recognise the right of citizens to stand for elections, which often comes with more constraints in the form of citizenship, residency and other requirements. Notably, many constitutions in Africa still ban independent candidates, especially for presidential elections,³ despite the fact that the African Court on Human and Peoples’ Rights has found the ban on independent candidates incompatible with the African Charter.⁴ Some constitutions in Africa also ban dual citizens or citizens of acquired nationality from running for certain high offices (e.g., 2016 Constitution of Cote d’Ivoire, article 55; 2010 Constitution of Angola, 110(2)), and in Muslim majority countries, candidates, particularly presidential candidates, are required to be Muslims (e.g., 2014 Constitution of Tunisia, article 74).

Perhaps more critical and contentious are provisions regarding the applicable electoral system and recognition and regulation of political parties. In general, African constitutions provide for key aspects of the electoral system for presidential elections. The detailed constitutional regulation of the electoral system for legislative elections is less prevalent, especially in Francophone Africa, where key aspects of the electoral system are left for legislative regulation. The recognition and regulation of political parties shows similar pattern to legislative election rules.

This contribution seeks to provide a broad overview of the constitutional regulation of the electoral system (Section 6.2) and political parties (Section 6.3) in African constitutions to identity any discernible patterns. Notably, it discusses how the relevant rules fared in times of democratic backsliding, using the example of Benin, one of the notable stories of democratic progress in Francophone Africa, which has recently witnessed downgrades in its democratic ranking (Section 6.4). The chapter also briefly discusses winner-takes-all politics (Section 6.5) as a fundamental scourge of both stability and democratisation in Africa and explores the importance of constitutional rules in ameliorating the challenge, with some examples. The Section 6.6 concludes.

6.2 Overview of Electoral Systems in Africa

The large majority of African countries have established presidential and semi-presidential systems of government, with only a handful of countries adopting a parliamentary system. Most Anglophone African countries have established presidential systems, with the exception of some, such as Namibia and Tanzania, which establish semi-presidential systems (president-parliamentary), and Botswana and South Africa, which effectively have a parliamentary system where the position of president and prime minister are fused in a unified presidential office. In contrast, Francophone and Lusophone African countries have predominantly established semi-presidential systems. The only formally parliamentary systems in Africa are Ethiopia, Lesotho (under a king) and Mauritius. While Morocco has a powerful government led by a prime minister akin to parliamentary systems, the king exercises tremendous powers and is therefore not seen as parliamentary.⁵

The rules governing electoral rules to the executive and legislature, as well as the institutions charged with managing elections, show similarities across the systems of government. This section briefly discusses the applicable rules regarding presidential and legislative elections. While discussing electoral rules, it is important to note the existence of supranational constraints on substantive, procedural and institutional aspects. Notably, the ECOWAS Protocol on Democracy and Governance prohibits members states from altering electoral rules within six months of elections, except through broad agreement of political forces (article 2(1)). Substantively, a panoply of instruments, notably the African Charter and the ACDEG, provide additional external constraint on the content and change of electoral rules.

6.2.1 Electing Presidents

The applicable rules regarding presidents (both directly elected or otherwise) are largely regulated at the constitutional level across Africa (as well as around the world)⁶ and generally not controversial. In countries with directly elected presidents, the rule is either plurality (first-past-the-post) or absolute majority (with a second round when necessary), with the latter dominating since the 1990s. Reforms have occurred both ways, with, for example, The Gambia (2001) and the Democratic Republic of Congo (2011) moving from absolute majority to plurality, while Togo (2019) shifted to the two-round system (which it had abolished in 2002) . Kenya and Nigeria provide for a regional distribution of votes for winning presidential elections with a view to encourage inter-group/regional political formations. Accordingly, in Nigeria, a candidate must win both a plurality of the votes nationwide and at least 25 per cent of the votes in at least two-thirds of the thirty-six federal states (Constitution of Nigeria, 1999, section 134), and in Kenya a candidate must win an absolute majority nationwide and at least 25 per cent of the votes in at least half of the forty-seven counties (Constitution of Kenya, 2010, section 138). In both cases, if no candidate secures a win in the first round, a runoff is held where the regional distribution rule no longer applies.

In general, constitutions in Anglophone African countries include elaborate provisions governing presidential elections, largely avoiding serious controversies around the rules. An exception is Malawi. The Malawian constitution is not clear on whether the applicable rule is absolute majority or plurality. In practice, plurality was accepted as the rule, particularly following a decision of the highest court of the country.⁷ Nevertheless, in 2020, the Supreme Court invalidated the May 2019 elections and crucially found that the applicable rule was an absolute majority, and not a plurality, effectively reversing its standing jurisprudence. In the absence of constitutional rules regarding the run-off election, the Court ordered parliament to clarify the applicable rules through reforms. Such reforms are yet to be enacted, and fortunately, the rerun election provided an opportunity for the second and third opposition candidates in the invalidated election to form an alliance, which secured a decisive absolute majority against the incumbent (who has been declared a winner of the invalidated election with a plurality of just over 38 per cent of the votes).

The constitutional rules for presidential elections are also generally broadly defined in constitutions of Francophone African countries, although the level of detail is comparably less than in Anglophone counterparts. Partly because of the relative constitutional minimalism, the rules in Francophone countries have been controversial in some cases. As indicted in Togo, the two-round system was restored in 2019 after persistent opposition demands and political instability. Perhaps most significantly, in Cote d’Ivoire, a rule that required presidential candidates to have both their parents born in the country proved extremely destabilizing, because most citizens in the north, who are also predominantly Muslims, often have cross-border links with people of Burkina Faso and other countries. The rule was ultimately abolished in constitutional reforms in 2016.⁸

Another important aspect of presidential elections that continues to be a recurrent source of controversy relates to term limits.⁹ Changes to applicable rules to extend incumbent terms represent the starkest manifestation of a decline in democratic space and autocratisation.¹⁰ Because term limits represent clear rules that are hard to bypass otherwise, incumbents often target their amendment – either directly or indirectly through the adoption of ostensibly new constitutions and, in certain cases, through questionable judicial interpretations.¹¹

Term limits are particularly often altered or removed in countries where power is personalised with weak parties (such as Congo Republic, Djibouti, Cameroon, Uganda) and less likely to occur in countries with equally dominant but strong parties or in competitive contexts. In countries with dominant but strong parties (such as Tanzania, Botswana and South Africa), term limits have been consistently respected, despite the parties having the numbers to change the rules, implying the importance of term limits in regulating intra-party power dynamics and alternation of power. Indeed, Tanzania became the first African country to witness an alternation of power as a result of term limits in 1995, while the country was still practically a one-party state. Similarly, in countries with relatively competitive contexts, term limits have largely remained stable, with efforts at altering them defeated, such as in Nigeria, Malawi and Zambia.

Other constitutional aspects of presidential electoral rules are not frequent subjects of reform, mainly because incumbents find ways to rig the applicability of the rules, without the need to change them. Even reforms related to the electoral system, with some countries shifting from runoff to plurality, may partly be explained by cost and other inconveniences, rather than simply as the last nail on the democratic coffin, as are changes to term limit provisions.

Additional aspects that have often generated controversy relate to registration fees for presidential candidates, which are not always regulated at the constitutional level but can prove a major hinderance to some candidates. Most recently, Nigeria’s main opposition parties introduced a nomination fee for presidential aspirants of up to 100 million Nira (about USD 250,000) and lower amounts for other offices.¹² Interestingly, these fees are set by the parties, rather than by law.

Perhaps most notably, the introduction of a ‘sponsorship’ systems has become controversial in recent years in Senegal, Benin and Burkina Faso. A 2018 Senegalese law established a ‘citizenship sponsorship’ requiring presidential candidates to obtain the support of about 1 per cent of the voters in at least seven regions of the country.¹³ While the law is intended to streamline electoral competition and ensure nationwide support for candidates (therefore reduce regionalised politics), it tends to benefit incumbent presidents and parties.

More seriously, in Burkina Faso, 2020 reforms to the electoral code require presidential candidates to be sponsored by at least fifty elected officials. When sponsors are municipal councillors, they must be located in at least seven of the thirteen regions of the country.¹⁴

Similarly, Benin introduced a party sponsorship system in 2018 requiring candidates to obtain endorsement of at least sixteen members of the National Assembly (of which there are eighty-three members) or mayors (of which there are seventy-seven).¹⁵ While the number may seem relatively low, the requirement came after opposition parties boycotted the 2018 legislative elections because of alleged repressive tactics and changes to the electoral law establishing, among other rules, a requirement on political parties to receive a ‘certificate of conformity’ from the Ministry of Interior, to secure at least fifteen members from each municipality of Benin, as well as a ban on party alliances from presenting list of candidates for elections.

The rules in Benin were adopted ostensibly to stem the fragmentation and proliferation of candidates and parties (with the country of around 12 million people reportedly home to more than 200 parties before the changes) and encourage nationwide and ideological political formations, rather than regionalised identity-based parties. Nevertheless, in effect this meant that opposition candidates needed the endorsement of members of parliament and mayors aligned with the ruling coalition. These rules have proved extremely controversial partly because they are not regulated at the constitutional level and are therefore seen as unilateral and even opportunistic and capricious. Indeed, the Benin government has arrested and convicted key opposition leaders following instability after the latest elections where President Patrice Talon won re-election (despite promising during his presidential campaign not to run for re-election and even to amend the constitution to impose single terms on presidents).¹⁶ Benin, once a paragon of democracy in Francophone Africa, is now classified as only partly democratic in all major democracy rankings.

Overall, provisions such as term limits and the two-round system were largely adopted in the 1990s in African constitutions as major concessions at moments of incumbent vulnerability and democratic euphoria during the transition towards democracy. Old and new dominant leaders and parties have since then sought to alter the rules and, on key occasions, succeeded, often marking a significant point in democratic backsliding.

6.2.2 Electing Parliaments (for First Chamber)

Unlike electoral rules for presidential elections, many African constitutions leave key aspects of parliamentary elections for legislative regulation. Notably, constitutions in Francophone African countries rarely even determine the applicable electoral system for the legislature, instead leaving it to determination through organic laws (which often required an absolute majority to enact). For instance, the 2016 Constitution of Cote d’Ivoire, one of the latest in Francophone Africa, provides that an organic law determines ‘the number of members of each house, the conditions of eligibility and appointment, the system of ineligibilities and incompatibilities, the methods of voting and the conditions under which new elections should be organised’ (article 90). Compare this with the 2010 Constitution of Kenya where the Constitution specifically determines the number of members of the National Assembly as well as manner of their election, including representation of women, as well as qualifications and disqualifications (articles 97–99).

Most countries in Africa have adopted a list proportional system (particularly common in civil law countries [Francophone and Lusophone], first-past-the-post system [mainly common in Anglophone countries] and mixed electoral system [parallel].¹⁷ Constitutional rules regarding legislative elections have generally stayed stable, although it is not clear whether the rules are results of accident and historical colonial diffusion, or deliberation, reflection and consensus. While the legislative framework for the electoral system has changed on occasions, particularly with the transition to proportional systems in Francophone countries, it is unclear whether changes to legislative electoral rules have been tampered with to distort outcomes. This may partly be because incumbency advantages and outright election rigging often deliver tolerable results, while electoral reforms may be seen as unnecessary and even undesirable and visibly capricious.

Nevertheless, there are cases where electoral reforms have proved extremely contested. For instance, in 2018, Benin changed the electoral code imposing among other things a 10 per cent threshold for winning seats in the National Assembly.¹⁸ While this was ostensibly intended to crack down on party fragmentation, it favoured the incumbent party and led to opposition boycott of the 2019 legislative elections. Indeed, following its observation of the elections, the African Union Election Observation Mission noted that the 10 per cent threshold ‘appears to contradict the fundamental principles of fairness and equal suffrage’.¹⁹ Regardless of the propriety of the applicable rules, the Benin experience shows that the lack of constitutionalisation of fundamental aspects of the electoral system could lead to unilateral alterations by incumbent parties without securing broad consensus. Indeed, in Benin, the incumbent party won all the legislative seats in 2019, which it then used to adopt constitutional amendments, without the need to submit them to a referendum – a referendum is only required if the proposed amendment does not secure a four-fifths majority in the National Assembly (Constitution of Benin, 1990, article 155).

6.3 Political Parties

Political parties are the principal engines of competitive democracy, straddling both the private and public spheres. The post-independence African constitutions largely left the issue of political parties unregulated. In any case, the post-independence period of multi-partyism was remarkably short and many of the countries, with the notable exception of Botswana, Mauritius and The Gambia, established de facto or de jure one-party systems.²⁰

Multi-party politics returned in the 1990s as central to the waves of democratisation across Africa. Accordingly, the new constitutional frameworks recognised the right to political competition and opposition as critical to democratic dispensation and to peace and stability. Similarly, the ACDEG requires states to strengthen political pluralism and recognise ‘the role, rights and responsibilities of legally constituted political parties, including opposition political parties, which should be given a status under national law’ (article 3) . Fombad has argued that the constitutional recognition and regulation of political parties ‘is not only imperative, but its broad scope to avert or at least limit the risks of one-party dictatorship is crucial’.²¹ In this regard, Chilton and Versteeg have found that constitutional rules with natural constituencies, such as those applicable to political parties, are systematically associated with better practices and compliance.²²

The level of detail with which African constitutions recognise and regulate political parties varies. In general, more recent constitutions across the continent and constitutions in Anglophone African countries tend to have more elaborate provisions, while Francophone countries tend to leave the regulation of parties to organic laws. Fombad has provided a useful analysis of twenty-four African constitutions partly drawing on a framework developed by Khaitan.²³ Notably, constitutional regulations of parties in Anglophone countries seem to aim at ensuring the democratic character of political parties in view of their public function, while in civil law countries, where it exists, the aim is mainly to protect parties from the state, in view of their private character.²⁴

Many constitutions provide for equality of treatment of parties, principally through equitable access to publicly funded media, often during electoral periods, but in some cases all the time (Constitution of Gabon [article 95], Ghana [article 55(11)] and South Africa [section 197]). For instance, the Ghanaian Supreme Court has ruled that opposition parties have the right to be granted access to public media to respond to an annual budget presentation by the government.²⁵ At the continental level, the African Commission on Human and Peoples’ Rights has called on member states to ensure that political parties, especially opposition parties, ‘are given equitable access to state controlled media and resources’.²⁶

6.4 Electoral and Party Rules during Backsliding: Example of Benin

In general, changing the rules is too visible a strategy that incumbents with other options are likely to avoid. All things remaining equal, incumbents would arguably tend to prefer irregular means to undermine opposition and other critical groups. Nevertheless, constitutional and legal changes have been used to promote ostensibly legitimate political goals, while in practice unduly undermining opposition parties. Such changes are perhaps more likely in countries where irregular means are less readily acceptable, in view of the relatively open and politically competitive context.

Where constitutional rules exist, they can be effective even in largely authoritarian settings. For instance, in cases of presidential term limits, authoritarian incumbents often have no choice but to change the constitutional rules. Although term limit provisions have been bypassed in many countries, there are also cases where constitutional rules have worked in stymieing efforts at removing term limits – for example, Nigeria, Malawi, Zambia. Indeed, some incumbents opted to resort to courts (e.g., Senegal 2012) or simply delayed elections to practically extend their terms (DRC 2016), because of the political impracticality of constitutional change. In Burundi, the incumbent resorted to the constitutional court after failing to secure the numbers to amend the constitution to allow a third term.²⁷

The limited scope or absence of constitutional regulation of key aspects of political parties, as well as broadly legislative electoral rules, means that they are rarely directly targeted through constitutional change. Instead, incumbents would either resort to irregular means or changes to the electoral laws to pursue their objectives. As a consequence, the political party and electoral rules tend to be relatively stable. But not everywhere.

In Benin, since 2018, a number of legislative reforms, noted above, were enacted, which the opposition objected to – including notably the requirement that political parties needed certificates of conformity from the Ministry of Interior, a duty to secure at least fifteen members from each municipality of Benin and a ban on party alliances from presenting list of candidates for elections. The law also exponentially increased the fee to field candidates.²⁸ This was in addition to reports of repressive government tactics.²⁹ Because of a combination of these tactics and legal changes, opposition parties boycotted the 2019 legislative elections, which contributed to a very low turnout of 27 per cent, but a total victory to the incumbent president’s ruling coalition. The ruling coalition then used its dominance to adopt constitutional amendments through an accelerated/summary procedure, the first ever successful amendment since the adoption of the 1990 Constitution through a National Conference. Parliament adopted the amendments unanimously, which precluded the need for a national referendum, which is not necessary if the amendment receives a four-fifths majority in the unicameral National Assembly. While most of the amendments weren’t too controversial – such as the establishment of position of vice president, increased quotas for women, banning of death penalty, grouping of elections (holding them at the same time) – some were. Notably, the amendments effectively banned independent candidates from presidential elections, constitutionalising the party sponsorship system introduced in the earlier legislative reforms.

The amendments to the law and the constitution were challenged in the Constitutional Court but were approved. However, the Court was seen as captured by the incumbent president after he appointed his former personal lawyer as the chair of the Court. Two cases were then filed before the African Court on Human and Peoples’ Rights challenging the reforms on both procedural and substantive grounds.³⁰ The African Court found that the constitutional amendments were not adopted through ‘national consensus’ as provided for in the ACDEG (article 10(2)). Interestingly, the principle was first formally propounded by the Constitutional Court of Benin³¹ before its subsequent adoption as a continental standard in the ACDEG. Having found the entire amendment invalid on procedural grounds, the African Court did not deem it necessary to discuss the substantive compatibility of the amendments with continental standards. Nevertheless, the Court has ruled in another case involving Tanzania that a ban on independent candidates, even if done through the constitution, is incompatible with the African Charter.³² While decisions of the African Court are technically only binding on the specific respondent state, they have crucial ‘horizontal’ importance in other contexts, especially when the decisions relate to the validity of legal and constitutional provisions, rather than merely their interpretation or application.³³

Displeased with the cases before the African Court, Benin chose to withdraw the declaration allowing individuals and non-governmental organisations to submit cases before the Court – therefore preventing future cases against Benin – and has not implemented the decisions.³⁴ Interestingly, despite the fact that the African Union is required to enforce decisions of the African Court, the AU chose to send election observers to the 2021 elections held under rules that were found to be incompatible with the African Charter.

6.5 Addressing Winner-Takes-All Politics through Inclusive Majoritarianism

Electoral and political party rules are foundational to democratic contestation. In a context where electoral contests are often perceived as high stakes winner-takes-all politics – as former Nigerian President Olusegun Obasanjo quipped ahead of the 2007 elections, ‘a matter of life and death’– incumbents often resort to subtle and non-subtle oppressive tactics and in certain cases use or change the rules to increase their chances of victory and dominance.³⁵ Winner-takes-all politics involves not only the overbearing dominance of the incumbent leader and party over status, power, governance and distribution of economic rents but also the active ‘marginalization of perceived political opponents and the feeling of exclusion from the governance process [and access to resources] by those who do not belong to the government/ruling party’,³⁶ rendering them spectators, rather than critical partners in governance and development. Accordingly, winner-takes-all thinking could lead to a situation where ‘opposition parties may end up becoming desperate to win power by all means and at whatever cost; whilst the incumbents, mindful of the cost of losing elections, may also prepare to maintain power by all means and at anybody’s expense’.³⁷

It would not be an exaggeration to suggest that the most central challenge to democratisation and peace, intra- and inter-group tensions, stability and development in Africa lies in winner-takes-all, zero-sum politics. Accordingly, to enable progress in constitutional democracy in Africa, ensure the stability of the rules of the game and preclude their abusive changes, constitution makers need to recognise the manifestations of and tackle head-on the scourge of winner-takes-all politics. This political culture coexists with the widespread public support in Africa for ‘consensual democracy’, one that ‘places limits on the extent of political competition, but without compromising the principle of political accountability’.³⁸

This support for ‘consensual democracy’ arguably requires the recognition and empowerment of opposition groups, including in constitutional and legal reform processes, as well as in the exercise of constitutional powers. The issue of opposition empowerment has received some scholarly attention,³⁹ broadly as necessary to encourage political moderation,⁴⁰ and also in the context of constitutional amendments.⁴¹ Opposition empowerment not only enhances checks on the incumbent, it also increases the stakes for opposition groups in contributing to peaceful electoral and broadly democratic processes and accepting electoral outcomes – ‘losers’ consent’.⁴²

This section briefly discusses existing mechanisms in African constitutions empowering the opposition, mainly using the implications of the principle of ‘national consensus’ as a precondition for legitimate constitutional amendments and the experience from Seychelles, which gives the opposition almost equal role as the incumbent in constituting courts and other democracy, rule of law and accountability promotion and protection institutions. In combination, the principle of national consensus and equitable role of the opposition in making key appointments reduce chances of abusive constitutional changes and incumbent capture of critical institutions and therefore can serve as a robust antibody against democratic backsliding. More positively, opposition empowerment rules necessitate a culture of regular engagement and cooperation among political rivals, which could enable trust, tame polarisation and winner-takes-all thinking and ultimately contribute to nurturing the legitimacy, stability and resilience of democratic systems. Normatively, as well, rules empowering the opposition ‘ensure a more robust version of representation in politics, and hence a more robust version of legitimacy for democratic institutions’.⁴³

The principle of national consensus in the adoption of constitutional amendments was formally adopted in the ACDEG⁴⁴ with a view to preclude opportunistic and abusive constitutional changes, mainly in view of a pattern of reversal of key concessions in the post-1990 democratisation and constitutional reform processes, notably presidential term limits. Nevertheless, the Constitutional Court of Benin had earlier adopted it as a fundamental principle of constitutional change, which likely inspired the drafters of the ACDEG.

The ACDEG requires state parties to ‘ensure that the process of amendment or revision of their constitution reposes on national consensus, obtained if need be, through referendum’ (article 10(2)). The principle of national consensus superimposes on the constitutional amendment procedures established in national constitutions.⁴⁵ Accordingly, countries cannot simply argue that they have followed the specific amendment procedures outlined in the respective constitutions to justify changes.

The principle of national consensus effectively empowers opposition groups beyond supermajority and other rules as it necessitates deliberation, dialogue and negotiation – and potentially consensus – beyond the four walls of parliament. The principle is particularly critical regarding electoral and party rules that should be protected from unilateral alterations, regardless of the dominance of any political group – considering that incumbent parties or coalitions have in many African countries secured absolute dominance in formal political institutions.

Indeed, in Benin after the 2019 legislative elections, the ruling alliance secured full control of the parliament. Nevertheless, arguably because of the principle of national consensus established by the Constitutional Court, and also because of escalating political tension, the president convened an ostensible national dialogue. The dialogue ultimately generated certain fundamental principles to provide the basis for reforms. Nevertheless, some opposition groups boycotted the dialogue and even some who attended complained that the unveiled outcomes did not reflect the deliberations in the dialogue. As noted above, although the Constitutional Court approved the amendments, the African Court found that the dialogue was insufficient to comply with the principle of national consensus.

The African Court – and broadly the African Union and the bodies monitoring compliance with the ACDEG – has not expounded on what national consensus exactly entails, particularly whether a critical mass of opposition groups should be on board with constitutional amendments. The 2000 Lomé Declaration on the Framework for an OAU Response to Unconstitutional Changes of Government recognises as fundamental principles the promotion of political pluralism, the principle of democratic change of government and ‘recognition of a role for the opposition’ (emphasis added).⁴⁶ The Declaration also specifically calls for ‘guaranteeing access to the media for all political stake-holders’.

In view of these continental standards, the value of the principle of national consensus would be maximised in protecting constitutional provisions, particularly those relating to the democratic electoral game, where no player is allowed to change on their own the rules by which they must play, regardless of their dominance. In a different article, this author has argued that moving beyond supermajority and/or referendum rules and requiring direct cross-party approval of the most abused constitutional rules may be necessary to lead to a form of democracy founded on inclusive majoritarianism.⁴⁷ The legitimacy and validity of amendments to the ‘democratic core’ should depend not simply on the size/quantity of the supermajority but its quality/inclusiveness (of more than one group’s voice).

Beyond constitutional change, the Seychelles Constitution perhaps has the most progressive and comprehensive guarantees for the involvement of opposition groups in governance, effectively precluding any incumbent from capturing rule of law, accountability and democracy protecting institutions. The Constitution provides for a Constitutional Appointments Authority with the mandate to propose names for key appointments and removals, including the Attorney General, members of the Electoral Commission, highest judges and the Ombudsperson (articles 139–142). The Authority has five members: two each selected by the President and the Leader of Opposition and a chairperson selected by the four members. The members serve guaranteed seven-year terms (which does not overlap with presidential terms) and may only be removed under strictly defined processes. The Leader of Opposition is elected by National Assembly members who are not from the party of the incumbent president, hence could be from the legislative majority or minority (article 84(5)). The salary, allowances, gratuity or pension payable to the Leader of the Opposition are determined through legislation but ‘shall be not less than those payable to a Minister and shall be a charge on the Consolidated Fund’. While other African countries have recognised the role of the opposition in the appointment of a handful of members of election management bodies, Seychelles stands out in building a representative and impartial appointment process to courts and other rule of law, accountability and democracy promoting institutions.

In addition to the principle of national consensus and role of the opposition in the Seychelles, African constitutions are increasingly recognising the status and roles of the opposition. In Burundi, the 2018 Constitution requires that the President and Vice President must belong to different political parties/coalitions, as well as different ethnic groups, or are independents of different ethnicities (article 124). A 2016 amendment to the Constitution of Senegal introduced the position of Head of Opposition and requires the adoption of an organic law outlining the rights and duties of the opposition (article 58). The 2016 Constitution of Cote d’Ivoire guarantees the parliamentary opposition the right for adequate and effective representation in all the bodies of Parliament (article 100). The now repealed 2014 Constitution of Tunisia assigned the chair of the Finance Committee and rapporteur of the External Relations Committee to the opposition (article 60). The Tunisian opposition is also entitled to establish and head a parliamentary committee of enquiry annually. The Constitution of Rwanda establishes a National Consultative of Forum of Political Organizations as a permanent forum to bring together political organisations for the purposes of political dialogue and building consensus and national cohesion (article 59). While the focus of this chapter is on constitutional rules, which provide the most effective protection, many countries in Africa also empower the opposition, particularly in leading and representation in parliamentary committees, in various ways through the standing orders of parliament.

Overall, there is increasing recognition in Africa that winner-takes-all politics poses a fundamental challenge, and innovative approaches both at the national and continental level have emerged to institutionalise antidotes to the winner-takes-all thinking, principally through the empowerment of the opposition. In a way, constitution makers are starting to imagine and craft constitutions from the perspective of electoral losers, rather than merely the winners. In view of these developments, a more systematic deliberation on the issue is critical at national, sub-regional and continental levels to identify the various modalities of opposition empowerment at the constitutional, statute and legislative standing order levels to assess their performance in enabling stable, legitimate and resilient democratic and constitutional systems.

6.6 Conclusion

In summary, constitutions in Africa, as elsewhere, regulate to a different degree key aspects of the rules of the democratic game – electoral systems, political parties, courts and democracy and accountability promotion institutions. Constitutions in Anglophone African countries tend to have elaborate rules on the rules regulating the terms of democratic political contestation. In contrast, Francophone African countries tend to be constitutionally minimalist, leaving large aspects of these key issues, even with presidential elections, to organic law. Despite the differences, however, the constitutional rules have largely remained relatively stable. In most cases, democratic backsliding often happens through irregular processes, in some cases through legal reform and rarely through constitutional change, with the notable exception of presidential term limits.

Where there has been constitutional change, the pattern is decidedly towards better guarantees and protection of electoral rules. This is not to underestimate the importance of constitutionalisation of the fundamental rules to reduce the chances of formal backsliding. Indeed, the trend is that newer constitutions tend to regulate more of these rules of the democratic game than their predecessors.

Beyond the rules of contestation, this contribution argues that central to the battle against democratic backsliding may be the imagination of constitution design from the perspective of electoral losers, rather than merely electoral winners, more specifically, through ways of equitable participation of the opposition in governance, rather than merely opposing. The requirement of national consensus for the adoption and change of fundamental aspects of democratic competition ensures relative stability of the rules and prevents a dominant incumbent from self-servingly altering the rules of the game. The empowerment of the opposition in constituting the membership of the institutions central to rule of law, democracy and accountability – as in Seychelles – can reduce the chances of political capture, therefore enhancing the overall legitimacy of the political system, enhancing the stakes for all to accept electoral outcomes, enabling consistent engagement and dialogue and trust among rival political actors and, overall, building a stable foundation for constitutional democracy. This is not to say that the deliberate empowerment of the opposition is an unmitigated good. Such empowerment could potentially lead to political paralysis and deadlock. Accordingly, it is important to design systems that encourage deliberation, not blockade, which necessitates deadlock breaking mechanisms (European Commission for Democracy through Law [Venice Commission] 2018).⁴⁸ A balance must be struck to ensure that the desire to check the tyranny of the political majority through including the opposition in governance does not degenerate into the tyranny of the political minority.

Accordingly, constitutional designers in Africa (and beyond) should reimagine democracy beyond majoritarian conceptions and seek ways to empower the opposition through inclusive governance. Scholars should also explore ways through which countries in Africa and across the world have empowered the opposition, how particular approaches have operated in practice, and propose ways to refine the approaches based on comparative practice.

7 Parties versus Democracy Addressing Today’s Political Party Threats to Democratic Rule

Tom Gerald Daly and Brian Christopher Jones

7.1 Introduction: The Political Party Threat to Liberal Democracy Worldwide

The growing threat to liberal democracy worldwide is, in many ways, a political party threat. Recent years have seen the rise of a range of populist, illiberal, nativist, xenophobic, far-right, and neofascist parties.¹ In many places, parties with a questionable commitment to liberal democracy have entered government, while others remain outside government but are making gains.² We also see established democratic parties in government that have threatened or incrementally dismantled democratic structures through subversion by an outsider or the ascendance of an extremist wing. These threats are studied under rubrics including “constitutional retrogression,”³ “constitutional capture,”⁴ and “democratic decay.”⁵ While this phenomenon is often framed as an executive-led problem, it also needs to be understood as a political party problem.

Parties and party leaders occupying an ill-defined space on the political spectrum between the center and extreme – the “far-right lite” – now present a much greater threat to democratic governance than overtly antidemocratic fringe outfits, such as Germany’s neo-Nazi National Democratic Party (NPD). Such parties also frustrate, in new ways, the application of existing public law and policy mechanisms to address democracy-threatening parties, including refusal of registration, thresholds for entering parliament, application of the criminal law, outright banning, the erection of “cordons sanitaires” to freeze them out of governance, or a practice of considered engagement. Crucial features of contemporary political party threats include their ambiguous nature, their growing size, and the subversion of democratic parties by errant leaders or extremist factions.

This chapter makes the following central claims: that contemporary political party threats require us to more systematically map the key threats posed, to pay greater attention to crafting novel public law and policy solutions to address these threats, and to reflect anew on our fundamental assumptions about the relationship between political parties and the functioning of liberal constitutional democracy itself. At a time when political party systems are transforming worldwide, and certain parties’ core function is shifting from a broadly rational vehicle for channeling citizen policy preferences to a more emotive representation of identity, this chapter aims to ignite discussion and debate on these developments.

The chapter contains four sections. Section 7.2 addresses the enduringly awkward relationship between democratic governance and political parties, as both essential mediators between the public and State and forces that can frustrate the design and functioning of the democratic system. Section 7.3 discusses conventional approaches to political parties perceived as threats to democratic governance. Section 7.4 highlights the inadequacies of existing approaches to address the threats posed by contemporary political parties. Section 7.5 canvasses a number of potential innovations in responding to such threats, with the aim of spurring deliberation on this crucial issue.

7.2 Political Parties: Central to Democracy but Orphans of Constitutional Thought

Despite being central to contemporary understandings and conceptualizations of functioning liberal democracy, political parties occupy an enduringly awkward position in democratic governance and constitutional law, representing both a potential threat to democracy and a virtually unavoidable medium between the state and the people in facilitating democratic governance in complex modern polities. This tension has deep historical roots. In crafting the US Constitution, James Madison warned of the “factional threat” represented by a group “who are united and actuated by some common impulse of passion, or of interest, adverse to the rights of other citizens, or to the permanent and aggregate interests of the community.”⁶ While the US Constitution itself says nothing of parties, its entry into force shortly predated, even spurred, the coalescence of the US political system into two clear groupings, centered on the issue of the extent of federal power, and prefiguring the enduring two-party system central to contemporary US democracy.⁷

Some 4,000 miles distant, political clubs in revolutionary France arose in the heady years of newly won political freedom following 1789, which saw a flowering of open political activity and exchange of ideas. However, Jacobin clubs, in particular, having played a key role in the height of The Terror from 1793–1794, during which the revolutionary government pursued its aim of countering its internal and external enemies through extreme violence, were closed down after the end of The Terror in 1794.⁸ The terms “terrorism” and “terrorist” are said to have been invented retrospectively to describe the Jacobins and the methods they employed.⁹ In France, then, the first (proto-)parties rapidly came to be viewed as antithetical to good governance. Yet, despite their increasing systemic importance, successive constitutions remained silent on the role of parties as the French Republic repeatedly foundered and renewed itself.

Despite the concurrent rise of constitutional government and political parties across the long nineteenth century (i.e., 1789–1914), constitutions worldwide largely overlooked parties as an essential element of the modern democratic state. As Aradhya Sethiya offers: “If political theory saw parties as anti-democratic, the eighteenth- century constitutions considered them constitutional externalities” or even “orphans of constitutional law.”¹⁰ In the post-1945 era, the most common early references to political parties in constitutional texts concern their registration and the constitutional power to ban parties opposed to democratic rule: originally found in the 1949 Basic Law of West Germany and spreading in subsequent decades to states worldwide, including Spain, South Korea, Israel, and various states in Central and Eastern Europe after post-1989 transitions to democratic rule (e.g., Czech Republic and Poland).¹¹ In recent decades in continental Europe, thicker constitutional recognition has transformed political parties “from socio-political organizations into integral units of the democratic state,” viewed as an attempt to shore up their legitimacy as their claim to democratic representation has weakened, not least due to declining membership.¹²

However, in constitutional law scholarship, parties have all too often been ignored or treated as an unwelcome guest, running amok around the three pristine pillars of ordered government sketched in the constitutional text. Not so in political science, where scholars, more interested in whomever exercises power and less hidebound by the niceties of constitutional texts and ideals, have expended much more energy on understanding precisely how political parties operate within the democratic system.¹³ A rich literature analyzes everything from interparty relations to intraparty dynamics, to sweeping shifts in political party systems.¹⁴ However, advances in legal actors’ understanding of political parties as constitutional actors have been made in the past two decades.

Kommers has framed the Federal Constitutional Court of Germany’s case law on political parties as a “jurisprudence of democracy,”¹⁵ shaping the electoral system with the aim of ensuring a genuinely representative political system and bringing their roles within the constitutional realm. As well as insistently affirming the core democratic role of political opposition in its early decades and beyond, the court in key decisions granted political parties the power to defend their institutional rights before the court in a similar manner to other state organs, struck down restrictive candidacy laws, and upheld a law setting a 5 percent threshold of votes cast for parties to enter parliament, to ensure “orderly” governance in an electoral system characterized by diffuse voting patterns. The latter outcome reflected memories of the instability inflicted on Weimar’s parliamentary system by a “chaotic carousel of shifting coalitions and collapsing governments, of immobile parliaments repeatedly dissolved.”¹⁶

US scholars have crafted a “law of democracy” literature focused on an institutionalist analysis of the true workings of the democratic system, which underscores the serious tensions between real-world practice and the scheme set out in the venerable constitutional text. In a 2005 article, Pildes and Levinson argued that the original Madisonian design of the US Constitution, predicated on healthy interbranch competition as a means of preventing excessive concentration of political power and the concomitant risk of a tyrannical government, had been almost immediately superseded by the simultaneous emergence of the political party system.¹⁷ For Levinson and Pildes, the continuing focus on this outmoded model of interbranch competition elides the ways in which disciplined political parties can functionally fuse executive-legislative branch operation, which has been exacerbated by the sharpening of ideological interparty divisions through factors including the rise of gerrymandering by both parties and the strengthening of internal party discipline, which renders branch interests “contingent upon shifting patterns of party control.”¹⁸

This analysis is couched in a broader strain of recent US scholarship highlighting the way in which other long-term phenomena, including the growth of the administrative state and of (private) economic power, frustrate the ideals, understandings, and deep assumptions underlying the constitutional scheme and constitutional thought.¹⁹ Yet, despite attempts to understand and reconceive political parties in constitutional terms due to their unavoidable centrality to the exercise of public state power, in the US system (and other states such as Australia and South Africa)²⁰ they are generally viewed in constitutional terms as private entities, under-regulated, or at best cuckoos in the constitutional nest.²¹

While the analysis above remains largely framed as analyzing the shortcomings of “ordinary” politics in systems populated by parties broadly committed to democratic governance,²² public law scholars’ focus on the centrality of parties to functional democratic governance has started to intensify as parties hostile to liberal democracy have gained ground and various governing parties worldwide have actively diminished accountability and rights-protecting organs (independent courts, media, and civil society organizations), while maintaining a veneer of legality and democratic rule through sophisticated manipulation of law and continued elections.²³ This presents a challenge of a different order and magnitude compared to the imperfect systemic functioning analyzed by Pildes and others, raising more intensely the risk of tyrannical government through excessive concentration of power and subversion of the constitutional framework.

In many states worldwide, the political party system is now unable to perform the essential mediating and representative role essential to adequately functioning representative democracy. This is due not only to long-established trends such as declining membership but also to the intensification of extreme polarization and “invidious partisanship,”²⁴ the prioritization of partisan advantage over fidelity to constitutional and democratic governance,²⁵ the fuller “capture” of parties by elite or sectoral interests, and party “capture” of the state itself through domination of all previously independent democratic institutions, often facilitated by the fragmentation or weakness of the opposition.

Further complicating the picture, perhaps the defining feature of the political party landscape in many states suffering democratic decay today is flux: marginal parties are growing, new parties are forming, long-dominant centrist parties are losing support, more extremist wings of large parties are in the ascendant, and – the greatest challenge of all – recent years have witnessed the rise of parties that are ambiguous in terms of their commitment to liberal democratic rule, rather than avowedly antidemocratic. The party system, quite settled for decades in many states, has become a churn of change. Given this churn, it is useful to briefly map existing approaches to problematic parties.

7.3 Conventional Approaches to Antidemocratic Parties

This section summarizes the three principal ways – legal, constitutional, and policy approaches – states have attempted to address parties perceived as threats to, or inimical to, the democratic system, and highlights their inadequacy in remedying the novel democratic threats posed by contemporary parties.

7.3.1 Legal Approaches

7.3.1.1 Registration Conditions

Registration requirements (and refusal to register) have been used to curtail threats by making it more difficult for fringe and extremist parties to gain ballot access. While in some states – especially in long-established common law democracies – these may only consist of “bureaucratic niceties,” such as form- filling and fees,²⁶ in other jurisdictions requirements are “complex and lengthy.”²⁷ Even if parties meet all the formal bureaucratic requirements, state authorities are often empowered to refuse registration based on the wider aims of the party or because of incongruity with party laws or constitutional standards – although such refusals can usually be appealed.²⁸ While research suggests that the types and forms of documentation required for political parties are becoming lengthier and more complex, this has not kept democracy-threatening parties off the ballot. Savvy parties are aware of these restrictions and are unlikely to divulge information that may lead to registration refusal.

7.3.1.2 Thresholds for Entering Parliament

Thresholds, defined as “the legally prescribed minimum number of votes needed for a party to take part in distribution of parliamentary seats,”²⁹ are designed to protect parliaments against extremist or fringe parties that may gain a small but not insignificant number of votes. Usually set between 3–7 percent, they can be higher or lower,³⁰ and can also relate to regional versus national vote percentage, or even for party coalitions.³¹ Beyond the legal threshold, there is also a natural threshold that parties must surpass in order to gain seats, namely, the percentage needed to obtain one seat at the district level, which tend to be very significantly higher.³² For example, in the United Kingdom’s (UK) majoritarian system the natural threshold to secure a seat is 35 percent (preventing the UK Independence Party from gaining more than a single seat in the 2015 elections despite obtaining 12.6 percent of the national vote).³³ As with registration requirements, elections in recent years have demonstrated that many state thresholds are not keeping threatening political parties out of power.

7.3.1.3 Applications of New and Existing Law

Jurisdictions are often hesitant to restrict specific parties because of the implications this could have for rights and liberties, such as freedom of association and expression, and foundational values such as democratic pluralism. Yet, states commonly punish extremist parties or party leaders through terrorism, hate speech or incitement laws,³⁴ criminal law,³⁵ tax fraud laws,³⁶ and campaign funding regulations,³⁷ which can lead to parties breaking down or voluntarily dissolving. However, this may not be the best strategy to defuse the long-term problem and may even prove advantageous – rather than debilitating – for the targeted party, by bolstering its status and electoral success, such as Jean-Marie Le Pen’s success after his conviction for assault during a 1988 election campaign.³⁸ For states ordinarily less willing to tackle parties through the law, one-off restrictions include attempts in the 1950s to suppress or ban the main communist party in both the United States and Australia, which failed due to weak enforcement of a suppressive law (US)³⁹ or its being struck down by the apex court (Australia).⁴⁰ Alternative strategies, such as cutting off media access or government funding – which Germany’s Bundestag opted for in 2017 after the Constitutional Court refused to dissolve the neo-Nazi National Democratic Party (NDP)⁴¹ – have limited effect in the social media age and could even incentivize foreign or illegal funding strategies or the party dissolving and reregistering under a new name.

7.3.2 Constitutional Approaches

7.3.2.1 Election System Tinkering

Can particular election systems facilitate or diminish political party threats? While proportional systems have proliferated on the basis that they are more democratic by according voters a broader electoral choice and by constructing a more representative parliament after elections,⁴² Rosenbluth and Shapiro argue that this can not only lead to haphazard, weak, or deceptively representative coalition governments but also permits fringe and extremist political parties into the system.⁴³ They argue that having two strong parties in a majoritarian single-member district (SMD) system produces the best democratic outcomes.⁴⁴ Counter-arguments here include: in many jurisdictions any wholesale electoral system change would be difficult and unlikely to be achieved within a short time-frame, fragmentation may be rooted in longstanding political traditions, major traditional political parties in majoritarian systems can still be captured by authoritarian-leaning populist candidates hostile to liberal democracy,; and the latter problem has been aggravated by changes to “democratize” party leadership election methods, which, compared to more traditional selection of leaders through party or parliamentary leadership, removes barriers for questionable candidates.⁴⁵ Thus, constitutional tinkering of the electoral system is at best a medium-term option and, even if successful, is no panacea.

7.3.2.2 Banning or Dissolving Political Parties

The power to dissolve political parties based on their purportedly antidemocratic platform or operation is a feature of many democratic constitutions globally, representing one of the most controversial weapons in the arsenal of a post–World War II “militant democracy” capable of protecting itself from threat or collapse by employing illiberal means. The most influential model has been the 1949 Basic Law of West Germany: Article 212(2) empowers the Federal Constitutional Court to ban political parties that “seek to undermine or abolish the free democratic basic order or to endanger the existence of the Federal Republic of Germany.” While frequency of use varies considerably (from highly frequent in Turkey to not a single instance in Poland) and some sizeable parties have been dissolved (e.g., former ruling Communist parties in Latvia and Lithuania), they largely target fringe parties or specific state issues (e.g., secessionist parties).⁴⁶ One study indicates that twenty of thirty-seven European states analyzed had banned at least one party since 1945, totaling fifty-two bans in all, including both post-authoritarian states and states without experience of authoritarian rule.⁴⁷ The core expressive and associative political freedoms such bans affect are not only recognized in national constitutions but also in international human rights instruments, including the International Covenant on Civil and Political Rights (ICCPR) and regional conventions such as the European Convention on Human Rights (ECHR).⁴⁸ In 2000, the Council of Europe’s Venice Commission set out seven guidelines for political party dissolution,⁴⁹ drawing heavily from the case law of the European Court of Human Rights (ECtHR). However, these have not necessarily ensured clarity⁵⁰ and some of the ECtHR’s judgments have come under heavy criticism.⁵¹

Bligh and Müller have argued for a reconsideration and new understanding of party bans on the basis that novel challenges and different types of authoritarianism have arisen.⁵² As Bligh observes, “the dominant approach continues to be preoccupied with the Weimar scenario” of democratic breakdown in 1920s Germany, spurred by overtly antidemocratic actors.⁵³ Both public law and political science literature emphasizes the deficiencies of bans:⁵⁴ normatively, as being undemocratic and open to abuse, resting in intractable tension with adherence to democratic pluralism; and practically, as being “pointless”, “counterproductive,”⁵⁵ and diverting attention from more effective methods, such as policy approaches.⁵⁶

7.3.3 Policy Approaches

7.3.3.1 Cordons sanitaires

“Cordons sanitaires” entail parliamentary parties adopting a policy of refusal to engage with extremist or threatening political parties that have entered parliament. There is little consensus that this is effective: some question the efficacy of “quarantining” in that targeted parties may not always become pariah parties and could exert influence through other means. Although repressive measures can have the effect of pushing out a “small minority” of members from extremist parties, such actions may also attract potential newcomers because of the party’s “persecuted” status and can also lead to the establishment of clandestine networks and a hardening of extremist positions.⁵⁷ As Downs stresses, “denial, rejection, and repression have largely failed to mitigate extremism in the cases where they have been adopted as dominant strategies.”⁵⁸

7.4 Democratic Threats Posed by Contemporary Parties

7.4.1 The Rise of the “Far-right Lite” Party

In recent years, the clearest global trend in political party systems is the rise of parties with a more ambiguous relationship to liberal democracy and more significant electoral support (e.g., France’s Front National, Poland’s Law and Justice Party (PiS), Brazil’s Social Liberal Party (PSL), or India’s Bharatiya Janata Party (BJP).⁵⁹ The nature of these larger parties is a key obstacle to addressing the threat they present. They often present a form of “far-right lite,”⁶⁰ with partially detoxified platforms that steer away from any overt challenge to democratic governance and tend to frame their racist, xenophobic, and illiberal views in a more sophisticated manner than previous problematic parties (although the Brazilian context has featured more overt authoritarianism at times).

Moreover, a party like the Alternativ für Deutschland (AfD) – like most parties – is not a monolithic bloc of one mind on all issues. Its success appears partly based on its ability to offer the electorate two political “flavors”: a relatively moderate face, which frames anti-immigrant and other views as eminently sensible, and a much more strident and virulent face, which speaks against “an invasion of foreigners” and is capable of shocking statements.⁶¹ The AfD has made an art of walking back extreme statements with contrary statements from its more moderate wing. Thus, it is hard to fit the party into the established framework for addressing antidemocratic parties under the German Constitutional Court’s case law or accepted understandings of such bans reflected in the Venice Commission’s guidelines. Moreover, especially when the AfD has been the official opposition, a “cordon sanitaire” policy has been neither practically feasible nor democratically defensible.

Indeed, in the context of the Constitutional Court’s refusal to ban the NPD in 2017 – strongly criticized by political actors, as discussed above – Stefan Thiel approved of the Constitutional Court’s judgment as reflecting the view that German society “must adapt to fight extremist ideologies chiefly in the political, rather than the legal arena. First and foremost, this requires engaging with at times uncomfortable viewpoints, an active engagement of civil society in political debate and tolerance of dissent.”⁶²

7.4.2 The Size of “Far-right Lite” Parties

Second, and further undermining the potential application of existing mechanisms, is the size of contemporary democracy-threatening parties. While such “hybrid” parties long occupied the fringes of democratic political party systems, especially in Europe, it has been argued that they have now displaced liberal parties as the “third ideological authority” beyond Conservative and Christian Democrat parties, and Social Democrat Parties.⁶³ This means that Thiel’s point above gains added force: the larger a potential antidemocratic party, the more foolhardy (and less justifiable) it may be to attempt to suppress it by legal means, or to attempt a policy of exclusion or containment. It is tempting to argue that such parties should be targeted before they have the chance to grow, through refusal to register, application of the criminal law, or outright bans. However, this would assume that a party’s platform and views are explicitly antidemocratic, whereas contemporary parties present much more ambiguous fronts. It may only be when a party is in power that its true threat to democratic rule becomes apparent. Here, the “Weimar” scenario of overt aversion to democratic rule does not apply, which precludes the application of any banning mechanism.

7.4.3 The Entry of Antidemocratic Parties into Government

Apart from research on the banning of former ruling parties,⁶⁴ the majority of the literature on antidemocratic political parties focuses on contexts where the main political territory is occupied by “mainstream” parties within the acceptable ranges of the democratic political spectrum. However, in multiple states, parties hostile to liberal democracy have entered government, sometimes with significant majorities or in coalition. Some, like Hungary’s Fidesz Party, Poland’s Law and Justice Party or India’s BJP have secured multiple terms in government while others have been ousted after one term (e.g., Bolsonaro’s PSL in Brazil). The problems canvassed above regarding the futility of applying existing remedial measures to parties like the Front National and AfD are exacerbated in the case of a variety of parties that have, once in government, tended to incrementally hollow out democratic structures, crafting a hybrid governance system with few constraints on executive power but retaining elections. The examples of the Law and Justice (PiS) party in Poland and Fidesz party in Hungary demonstrate how difficult it is to deal with this issue before parties come to power. Indeed, both parties started as what seemed to be liberal-democratic parties; Fidesz (whose name comes from Fiatal Demokraták Szövetsége; Alliance of Young Democrats) emerged from a liberal student activist movement; PiS emerged from the Solidarity movement that spurred the Polish democratic transition ending Communist rule.

7.4.4 The “Subversion” of a Democratic Party by an Outsider/Extreme Wing

A different form of threat is posed by the takeover (or “populist capture”)⁶⁵ of a long-established “good” party by a “bad” leadership, whether by an individual outsider or an extremist wing. This tends to be the only choice available to authoritarian-leaning political forces where the nature of the established party system precludes the formation of a new party.⁶⁶ Examples include Rodrigo Duterte in the Philippines and Donald Trump in the United States.

In neither scenario could existing mechanisms tackle the problematic rise of these individuals. In a two-party system such as the US, using criminal law, “cordon sanitaire” techniques, or other existing mechanisms against the subverted party simply could not work without distorting the entire political party system, and would inescapably be viewed as partisan in nature. It is important to emphasize the distinction between party leadership and the party itself. For instance, the 2011 Venice Commission guidelines on banning political parties emphasizes that the activities of party members as individuals (including leaders) cannot provide the basis for dissolution, especially if such action runs counter to the political party’s constitution or activities, unless it can be demonstrated that the activity was taken by the party’s statutory body.⁶⁷

In cases of “subversion” of a democratic party by an outsider, rather than the cost of measures that target the entire party, it may be more effective to wield the scalpel of targeted measures to remove a corruptive leader, such as impeachment.⁶⁸

7.4.5 The Dominance of a Party by Unaccountable or Shadow Insiders

A “subverter” is not always an outsider, nor in a formal position of apex power in the State. In this connection, internal party dynamics appear increasingly important as a factor. This raises two difficult issues. First, to what extent can the activity of a dominant figure such as Poland’s Jarosław Kaczyński be separated from the party itself? After all Kaczyński is leader of the PiS party but only joined cabinet in 2020 until PiS was ousted in 2023. Second, what democratic concerns are raised by the level of dominance exercised by an unaccountable individual or group of unaccountable individuals? Where government policy and activity is excessively influenced by one figure, this appears to cut against the most foundational safeguards of a democratic system, such as the separation of state powers – acutely heightening the concerns highlighted by Levinson and Pildes in the US context regarding the impact of party dynamics on excessive concentration of power. Effectively, the separate branches of government become simply different arms of the party, rather than separate “sovereign” entities that check and balance one another’s power in concordance with the Constitution as well as acceptable constitutional practice in a democratic society. Such concentration of power in one individual also renders the link between the electorate and party more tenuous. It is an issue that requires much more attention in the literature.

7.5 Contemplating New Public Law and Policy Approaches

Effectively addressing the novel challenges to democratic governance posed by contemporary political parties requires new mechanisms, based on key lessons from the debate concerning existing and historical approaches to antidemocratic parties, including: falling into the trap of mechanisms that can be characterized as elite or partisan frustration of the will of the people, assuming that antidemocratic parties will be easy to identify, and distinguishing between party leaderships and the parties themselves. This section contemplates an indicative list of possible public law and policy options for addressing the difficult threats raised by contemporary political parties.

7.5.1 Can We Just Trust Courts to Make the Right Call?

A clear point of consensus across jurisprudence, scholarship, and practice is that the most serious forms of controlling political parties, such as bans, should be the responsibility of the constitutional court (or equivalent).⁶⁹ It therefore may be tempting to suggest that courts could be accorded much broader regulatory powers; for instance, to perform periodic party assessments for commitment to the constitution and rule-of-law principles or to assess parties’ internal democratic procedures to prevent excessive dominance by one figure or faction. Grounds for regulation could be reframed in wider terms, allowing more discretion to constitutional courts to take a tailored approach to each party, with more flexible standards of scrutiny.

However, such an argument dissolves in the face of four issues. First, existing jurisprudence on party bans and regulation at both the national and international levels has attracted significant criticism, not least the inconsistencies in the ECtHR’s case law, discussed above.⁷⁰ Second, courts may, for good reason, be unwilling to employ such an expanded regulatory power on the basis that it would mire them in partisan politics – especially regarding regulation of parties with more than marginal support. Third, packed courts in backsliding or fragile democracies might wield broad regulatory powers aggressively. As Cavanaugh and Hughes observe: “The use of [militant democracy] measures may well erode and devalue the very principles that they seek to protect.”⁷¹ Finally, even where independent courts remain in place, their characterization by authoritarian-leaning populist forces as elite liberal institutions could mean that intervention may strengthen support for such parties by allowing them to present themselves as victims of entrenched elites.

7.5.2 Nonjudicial Options

The party regulation model in states such as the United Kingdom may point to a less court-centered, approach. For instance, the UK’s party proscription process under the Terrorism Act 2000 is wholly executive-based via the Home Secretary, but this is tempered by the Act’s framework for deproscription. A proscribed party may apply to the Home Secretary for deproscription and, if declined, may appeal to a Proscribed Organisations Appeal Commission (POAC) consisting of one senior judge and two other members of the Commission (usually accomplished lawyers),⁷² with a right of further appeal to higher UK courts.⁷³ Thus, rather than court-centered from the beginning, the process of deproscription becomes increasingly court-focused only after decisions have been made again by the Secretary of State and then by an independent Commission. This may insulate the courts from accusations of political decision-making, as they are not the initial adjudicators on party dissolution.

Ideally, party regulation should involve multiple branches of government, incorporate quasi-judicial entities (e.g., independent commissions), and not place dissolution into the hands of one group or institution.

7.5.3 Emerging International Mechanisms

In the European Union (EU), approaches to contemporary illiberal parties, perhaps inescapably, have an international dimension. Alongside a long-standing but unsuccessful campaign to have the Hungarian and Polish governments sanctioned under Article 7 of the Treaty on European Union (TEU) for breach of fundamental values of the EU (e.g., democracy, rule of law) contained in Article 2 TEU,⁷⁴ and the pushback by both the Court of Justice of the European Union (CJEU) and national courts,⁷⁵ yet another gambit has emerged, focused on the parties themselves qua parties, rather than executive actors.

Pech and Alemanno in 2018 called on the European Parliament to request the EU party regulation body⁷⁶ to verify whether the European’s People Party (EPP) (which groups together a range of national parties, including Fidesz, the ruling party in Hungary) is in compliance with the EU’s fundamental values as set out in Article 2 TEU.⁷⁷ This legal mechanism (in a little-known EU Regulation)⁷⁸ had never been invoked and was perceived as potentially providing an avenue for Fidesz’s deregistration as a European political party, thereby, at least by implication, diminishing its power and damaging its domestic standing. Interestingly, the request for review of a party by the EU regulator can be made not only by other EU organs (the Council and Commission) but also by “a group of citizens,” although the latter is possible solely in the event of a “manifest and serious breach” of EU values. This option needs to be examined in light of the discussion above concerning the value and utility of repressive measures. Although it is not a party ban, with the relevant Regulation placing emphasis on political pluralism, it has not been pursued and appears as a measure of last resort from the EU law perspective.⁷⁹

7.5.4 Incentivizing Opposition Rights

It is abundantly clear from the literature that, in countering governmental degradation of the democratic system, opposition parties matter. Levinson and Pildes have suggested that a key measure to address the democratic deficiencies of the US political party system would be to adopt the European notion of opposition rights, that is, “measures to empower the minority party to oversee government action, such as the power to initiate investigations, to obtain information through the subpoena power or other means, or to control audit or similar oversight committees.”⁸⁰ More recently, this is a central plank of Huq and Ginsburg’s argument for rendering the US political system more resilient against backsliding⁸¹ and Khaitan’s argument for pushing back against the illiberal agenda of the ruling BJP party under Prime Minister Modi in India (as well as multiparty appointments for, and greater independence of, fourth branch institutions).⁸²

However, for some states, a focus on opposition rights is of little benefit where there is a seriously diminished or fragmented opposition. In the long term, these could be written into law, but for the short term – and again, as a measure of last resort due to democratic legitimacy concerns – the most effective approach may be to offer enhanced international funding for opposition coalitions that form a unified front against a ruling party that has demonstrably sought to entrench itself in power through the capture of independent accountability institutions and changes to electoral laws, although in many states this may be frustrated by laws against foreign funding.

7.5.5 Stronger Controls on Electoral Manipulation

One of the greatest threats to the very core of democratic functioning is the use of law to degrade the fairness and transparency of the electoral process, such as state laws establishing extreme gerrymandering and voter suppression in the United States.⁸³ These measures, again, break the link of true representation that renders the party a legitimate channel of the electoral majority.

How can this be addressed? In the immediate term, the clearest backstop is international condemnation. However, this requires an in-depth understanding of often sophisticated manipulation of electoral laws, which can be a hard sell to foreign political leaders and organizations. In the longer term, new constitutional design options might be considered, drawing on Dixon and Landau’s notion of “tiered constitutional amendment” – namely, creating different constitutional amendment requirements for different parts of the constitution,⁸⁴ which in the electoral arena, could require special and more onerous procedures or supermajorities to amend electoral law and transform electoral agencies.

7.6 Conclusion

This chapter sought to highlight key threats posed by political parties to the endurance of representative liberal democratic governance worldwide, to generate debate by putting a range of potential remedial options on the table, and to spur reflection on the need for a fundamental reorientation of deep constitutional assumptions concerning the role and democratic purpose of parties today. While it is impossible to be comprehensive or definitive regarding solutions, we have aimed to emphasize the urgent need for greater attention to the often ambiguous ways in which parties now threaten democratic governance. Despite prevalent analysis of the global authoritarian populist turn as based on a revolt of the electorate wrenching democracy from entrenched and out-of-touch elites – and there is considerable truth to that perspective – it is also a story of new elites delivering us charlatans, fake democrats, and fake democracy. Perhaps the most immediate lesson from this discussion is that to frame the challenges facing democratic rule worldwide as an executive, or even leadership, problem, is to miss the deep structural role that parties play in processes of democratic deterioration and decay. Worldwide, political parties are also learning from one another, as seen at the time of writing in how rapidly Slovakia’s new government – a coalition of Direction – Social Democracy (Smer-SSD), Voice – Social Democracy (Hlas-SD), and the nationalist Slovak National Party (SNS) – is implementing the “authoritarian playbook” developed in large part by far-right lite party governments in Hungary and Poland.⁸⁵ These negative dynamics may remain even when specific democracy-threatening incumbents are ousted in elections across the world, as we have seen in states such as the USA, Brazil and, more recently, Poland. There is no doubt that contemporary democracy requires wider rethinking and renewal, and solutions must go far beyond trying to turn the clock back to the status quo ante. But we must start somewhere: democracy-threatening parties are going nowhere.

8 What Is the Value of a Constitutionalized Right to Vote?

Yasmin Dawood

8.1 Introduction

In an era marked by democratic backsliding on the global stage, questions have been raised about the ability of constitutional safeguards to forestall authoritarian retrenchment.¹ Does the constitutional design of elections matter for the sustainability and functioning of democratic governance? This chapter seeks to address one aspect of this larger question by focusing on the right to vote. Given the norm of universal suffrage, the importance of the right to vote is virtually undisputed. What is the value, however, of a constitutionally enshrined right to vote? Does it matter whether or not the right to vote is constitutionalized?

It has long been accepted that the protection of constitutional rights is dependent, at least in part, on government enforcement and societal mobilization.² As James Madison famously noted in The Federalist Papers, liberty-protecting constitutional provisions amount to “parchment barriers” that are often unequal to the “encroaching spirit of power” embodied by the legislature.³ Constitutional rights provisions do not necessarily translate into actual protections for citizens simply by virtue of being included in the constitutional document.⁴ Rights-protection is contingent upon the work of public institutions, such as the courts, and more generally upon effective government.⁵

A recent empirical study, while confirming the core insight that constitutional rights do not on their own protect citizens from a government’s repressive actions, raises questions, however, about whether judicial independence and democratic accountability mechanisms are the prime causal factors leading to rights-protection.⁶ Adam Chilton and Mila Versteeg argue that those rights that are organizational in character (in the sense that there are organizations that are invested in protecting the right) are harder for governments to violate than rights that are individual in character (in the sense that these rights are largely relied upon and defended by individuals).⁷ That being said, they note that governments are usually successful when they are “determined to erode the protections provided by certain rights.”⁸

In this chapter, I claim that the contingent nature of rights protection is particularly pronounced with respect to the right to vote – and this is the case for two reasons, both of which are tied to the distinctive features of the right to vote qua right. First, I suggest, the right to vote is multidimensional: it is composed of constitutional, statutory, regulatory, and jurisprudential elements that interact with one another to collectively produce “the right to vote.” Second, the right to vote is a structural right in the sense that it is dependent upon an entire infrastructure of institutions to exist and perform its function. I suggest that these features of the right to vote – its multidimensional and institutional nature – make it uniquely susceptible to being undermined by political forces.

Although the constitutional enshrinement of the right to vote does not on its own protect voting rights, I claim that, as a normative matter, constitutions undoubtedly should recognize the right to vote. In addition to its potential impact on voting rights protection, I suggest that a constitutionally enshrined right to vote is crucially important for its expressive functions. A constitutionalized right to vote expresses a commitment to various democratic values and, in addition, establishes normative baselines regarding universal suffrage, political equality, and democratic representation. In democracies, these expressive functions can indirectly serve to protect voting rights.

However, the expressive functions of the right to vote can also, paradoxically, undermine democracy by furnishing autocrats in competitive authoritarian regimes with “democratic cover” while they are undermining the key determinants – competitive elections, rival political parties, freedoms of speech and association – that render voting meaningful. At a more general level, the expressive function of constitutional rights and structures raises questions about the ways in which authoritarian regimes use the mechanisms of democracy, such as voting rights and elections, to create the impression of democratic legitimacy while simultaneously eroding it in practice. However, the existence of democratic structures such as elections and voting, even when heavily manipulated, may nonetheless exert some constraints on elected autocrats, at least in comparison to the absence of such constraints in fully authoritarian regimes. For this reason, the claim that the right to vote ought to be constitutionalized remains normatively appealing, although the overall force of the claim is qualified in light of this paradox.

This chapter is organized in four sections. Section 8.2 introduces the multidimensional nature of the right to vote, while Section 8.3 focuses on the institutional dimension. Section 8.4 explores the claim that a constitutionalized right to vote can directly protect voting rights. Section 8.5 argues that a constitutionally enshrined right to vote is normatively valuable for its expressive function notwithstanding the potential that it will be used as a tool of autocratic entrenchment. The conclusion summarizes the main themes.

8.2 The Multidimensional Right to Vote

The right to vote is multidimensional: it is composed of constitutional, statutory, regulatory, and jurisprudential elements and the interactions among them. In many democracies, the right to vote is constitutionally entrenched. Even when it is constitutionally recognized, the right to vote is also comprised by myriad statutory and regulatory provisions that determine the eligibility and opportunity to vote. Rules that impose citizenship or minimum age requirements on voters, or that require certain forms of voter identification, or that deny the franchise to those convicted of a crime exert an important influence on the contours of the right to vote. These statutory elements of the right to vote can also be described as a “legislative” or “political” component.⁹ Statutes are enacted by political bodies to achieve certain political ends; hence a statute is both a legal and a political phenomenon. At times, the formal constitutional strand of the right to vote can be undermined by the statutory elements of the right to vote. This is particularly the case when the rules governing the eligibility and opportunity to vote are crafted to augment or even entrench partisan advantage.¹⁰ Judicial decisions that interpret these constitutional, statutory, and regulatory provisions likewise play a crucial role in delineating voting rights. As such, these various strands – constitutional, statutory, regulatory, and jurisprudential – taken together produce “the right to vote.”

A snapshot view of the right to vote in Canada illustrates the multidimensional nature of the right to vote. In Canada, Section 3 of the Charter of Rights and Freedoms provides that “every citizen of Canada has the right to vote in an election of the members of the House of Commons or of a legislative assembly and to be qualified for membership therein.”¹¹ In addition to the constitutional text, the right to vote is further specified in the Canada Elections Act,¹² which sets forth a vast array of rules governing federal elections. Most directly, these statutory rules determine the eligibility of voters and the mechanics of voting and vote counting. Provinces and municipalities have their own electoral statutes, which influence the right to vote at the local level. The right to vote is also comprises regulatory rules promulgated by Elections Canada, the electoral management body in charge of federal elections. Finally, the contours and content of the right to vote have also been determined by judicial decisions. The Supreme Court of Canada and lower courts have addressed various topics that bear directly on the right to vote, including voter qualification rules, voter identification requirements, the disenfranchisement of prisoners, residency rules, and the entitlement to vote.¹³

A cursory examination of global trends likewise suggests that many nations have enacted a complex array of rules and regulations that comprise the right to vote. For instance, citizenship is a qualification for registering to vote and for voting in national elections in over 95 percent of countries.¹⁴ The vast majority (87 percent) of countries around the world set their minimum voting age at eighteen years of age.¹⁵ An overwhelming majority of voters around the world are required to present some form of identification in order to vote.¹⁶ As of 2004, approximately 72 percent of democracies impose restrictions on voting for those who have been convicted of a crime.¹⁷ These statistics provide some indication of the complexity of rules that govern the right to vote in numerous countries.

The complexity of the rules comprising the right to vote is also evident in the United States. While the US Constitution includes many provisions concerning voting and political participation, it does not contain an affirmative right to vote. The “most explicit protections of the franchise … are phrased almost entirely in the negative – that is, they simply prohibit particular forms of disenfranchisement.”¹⁸ In addition to these constitutional provisions, there are countless statutory and regulatory rules governing the right to vote – a complexity that is amplified by the fact that individual states determine the qualifications for voting. Electoral regulations are likewise developed and applied at the state level. Every aspect of the right to vote is heavily specified by rules and regulations, including, for instance, the layout of a ballot, the methods for counting votes, the eligibility of voters, and the location of polling places. Court decisions, at both the federal and state levels, further specify the content and contours of the right to vote.¹⁹

The example of voter suppression in the United States illustrates not only how the right to vote is composed of various components but also how the statutory and jurisprudential strands of the right to vote can undermine the constitutional guarantee. The right to vote has had a long and turbulent history marked by racial discrimination and exclusion.²⁰ After the passage of the Fifteenth Amendment,²¹ states turned to facially nondiscriminatory disenfranchising tactics – including poll taxes, literacy tests, character tests, property and residency requirements, secret ballots, all-white primaries, and voter registration rules – to prevent African Americans from voting.²² Grandfather clauses waived such requirements for low-income and illiterate white voters. The Voting Rights Act of 1965, which prohibits racial discrimination in voting, was enacted to enforce the voting rights guaranteed by the Fifteenth Amendment of the US Constitution.²³

Although the US Supreme Court was generally protective of voting rights after the advent of the civil rights era,²⁴ in recent years, the conservative majority of the Court has undermined voting rights. In the wake of the Court’s decision in Shelby County v. Holder,²⁵ which effectively dismantled the preclearance process under the Voting Rights Act, states have passed a number of statutes that imposed new restrictions on the eligibility and opportunity to vote. These restrictions, which include stringent voter identification laws and complex registration requirements, amount to a new form of vote denial.²⁶ Given the confluence of racial identity and partisan affiliation in the United States, laws governing the eligibility and opportunity to vote are often designed to depress minority voting in order to achieve a certain partisan outcome.²⁷ In a recent case, Brnovich v Democratic National Committee,²⁸ the conservative majority of the Supreme Court considerably weakened another provision (section 2) of the Voting Rights Act, which prohibits voting regulations that have a disproportionate impact on minority voters. The Court announced a new approach to section 2 vote denial claims, which makes it arguably more difficult for plaintiff voters to prevail against voting restrictions.²⁹

8.3 Institutions and the Right to Vote

A related point is that the right to vote has an institutional dimension. The right to vote is a “structural right” because its existence depends upon on an entire infrastructure of political institutions.³⁰ Structural rights are individual rights that take into account the broader institutional framework within which rights are defined, held, and exercised. Structural rights theory holds that the participation of individuals is key (hence the emphasis on rights) but that individuals participate within an institutional framework that is constituted by relations of power (hence the emphasis on structure). Rights do not exist in a vacuum but are instead exercised within a particular political, institutional, and societal context. While structural rights theory can be applied to many kinds of rights, it has particular salience for democratic rights.

For instance, the right to vote, while held by individuals, presupposes the existence of a wide array of institutions and actors, including elections, candidates, political parties, constituencies, candidates, legislatures, and so forth. By contrast, other individual rights, such as the right to liberty, necessarily inhere in individuals and are not dependent for their exercise upon the existence of a prior institutional framework. To be sure, and as described in Section 8.2, every right requires institutions in order to be enforced. However, the right to vote cannot even be conceived of as a right in the absence of an entire system of institutions. In addition, the right to vote is dependent upon effective democratic governance to provide not only voting but also the associated institutions and processes, in particular free and fair elections, that render such a right meaningful.³¹ In this way, the right to vote confounds the usual distinctions between negative rights and positive rights; indeed, it comprises and combines elements of both.³² Along with a properly functioning set of institutions, the right to vote is also dependent to a large degree on effective electoral administration.³³ As a caveat, the notion of a structural right is conceptual; it captures something relevant about the institutionalized nature of the right to vote but it does not imply that the holder of the right to vote is necessarily entitled to all the relevant institutions.

The political and institutional mechanisms by which votes are translated into power also have an impact on whether the right to vote is meaningful. The aggregation of votes,³⁴ the formation of electoral districts, and the influence of partisanship on electoral rules,³⁵ for example, can have an effect on how and whether votes count. Broader electoral rules – such as those affecting campaign finance – can undermine the relative power of citizens’ votes. In the United States, for example, campaign finance rules have accentuated the disproportionate political influence of the wealthy. Political campaigns are largely funded by the so-called donor class, a wealthy and powerful minority.³⁶ Empirical research has shown that elected representatives are more responsive to the preferences of the affluent rather than those of most citizens,³⁷ which has distorted policymaking in Congress.³⁸

A meaningful right to vote is also arguably dependent on more diffuse sociological factors, such as the availability of informed voting. While social media has provided citizens with new venues for expression and information-gathering, it has also flooded public discourse with disinformation and fake news. Social media produces echo chambers, exacerbates polarization, and creates bias.³⁹ Individual actors and foreign governments have used social media to influence elections by targeting citizens with fake news.⁴⁰

8.4 Constitutionalizing the Right to Vote

To summarize thus far: the right to vote is comprised of multiple strands – constitutional, statutory, regulatory, jurisprudential – and is inextricably embedded in an institutional framework. Two implications emerge from these observations. First, the “right to vote” is heavily specified by rules and regulations that determine not only its outer boundaries but also its internal content. It is also unintelligible as a right in the absence of an array of institutions and processes. These two features of the right to vote distinguish it, I suggest, from other kinds of rights, such as the freedom of speech. To be sure, the difference is one of degree, not kind: while other constitutional rights, such as the freedom of speech, are determined to some extent by laws and judgments and are exercised within an institutional context, they are, in comparison to the right to vote, relatively less dependent on legislative and judicial specification and relatively less dependent on institutional mechanisms for their realization.⁴¹ The second implication stems from the first, namely, that the multidimensional and institutional features of the right to vote impose certain constraints on how effectively a bare constitutional right to vote can protect voting rights.

That being said, does the constitutional backstop of a right to vote make a difference? There is no question that constitutional structures matter, but it has proven to be difficult to reach a consensus about which structures matter and why. For instance, there are long-standing debates about whether presidential or parliamentary systems are more stable,⁴² whether the variation is a function of institutional factors other than regime type and the separation of powers,⁴³ or whether economic or cultural factors matter more than institutional ones.⁴⁴ Leaving aside questions of regime stability, institutional features, such as presidential or prime ministerial selection devices, may matter a great deal for other sorts of issues, such as the kind of democratic politics that results.⁴⁵ Or it may be the case that in other circumstances it is not possible to draw valid inferences from institutional rules alone.⁴⁶ A constitution may not even mention key institutional features, such as political parties, which have proven to be indispensable to politics.⁴⁷ And finally, constitutions may provide the legal framework for democracy,⁴⁸ at varying degrees of success,⁴⁹ but they may be fully compatible with competitive authoritarian regimes and fully authoritarian regimes.⁵⁰

Another way to approach this issue is to ask whether the absence of an affirmative right to vote, for instance in the US Constitution, makes a difference to the protection of voting rights. It is possible that a generalized and affirmative right to vote in the US Constitution would have made it easier for legislators to enact pro-voting rights legislation, or alternatively, would have furnished courts with greater tools to strike down legislation that undermined voting rights. It may be the case, however, that the protection of voting rights is more dependent on non-constitutional factors, such as statutory rules, political forces, and legal norms, which buttress the right to vote. For instance, the existence of compulsory voting in Australia⁵¹ suggests that non-constitutional rules can play an important role in safeguarding the right to vote. Indeed, in democracies without written constitutions, the franchise is protected by statute, as in the United Kingdom. International legal norms may also serve as a safeguard for the right to vote. For example, Article 25 of the International Convention on Civil and Political Rights provides every citizen with the right and the opportunity to “vote and be elected at genuine periodic elections which shall be by universal and equal suffrage and shall be held by secret ballot, guaranteeing the free expression of the will of the electors.”⁵² Broader cultural and social norms around the right to vote may provide important safeguards for voting even if formal franchise rights are weakly protected. In sum, contextual factors and the non-constitutional elements of the right to vote must be considered when evaluating the impact of the formal constitutional guarantee of the franchise in any given jurisdiction.

Given this complexity, it may be more useful to conceive of a constitutionalized right to vote as part of the “minimum core,” which is defined by Rosalind Dixon and David Landau as those institutions and rights that are required to preserve competitive democracy.⁵³ This minimum core can be used to assess and compare the performance of different constitutions. On this view, it may be less important to disaggregate the individual effect of each component of competitive elections and more important to focus instead on the extent to which constitutional structures, taken as a whole, protect democracy.⁵⁴ While I am in favor of minimalist approaches to democracy,⁵⁵ I think that even the most minimalist version of competitive democracy is actually fairly “thick” with respect to the complex interplay of rules, processes, and institutions that are inevitably at stake. In this chapter, I have defended, as a conceptual matter, a thicker multidimensional conception of the right to vote, which emphasizes the myriad details that underlie the seemingly simple act of casting a ballot. This thicker conception, I suggest, would still fall under minimalist conceptions of democracy.

8.5 The Expressive Functions of a Constitutionalized Right to Vote

The multidimensional and institutional features of the right to vote have the potential to impose significant constraints on voting rights. While the constitutional enshrinement of the right to vote can have a protective impact on voting rights, it provides no guarantee that it will do so. That being said, a constitution should protect the right to vote. I claim that a constitutionally enshrined right to vote is normatively valuable inasmuch for its expressive functions as it is for its direct impact on voting rights.

Legal rules have an expressive dimension.⁵⁶ The constitutional right to vote expresses and symbolizes the values of political equality, respect, and belonging. As Judith Shklar observed, the right to vote is a “certificate of full membership in society” that “confers, and in some ways, defines, full citizenship.”⁵⁷ Those who are denied the vote “feel dishonored, not just powerless.”⁵⁸ Individuals who are excluded from the franchise are, according to Charles Beitz, “socially dead” as they are “not publicly recognized as persons at all.”⁵⁹ Not only does the right to vote confer belonging and dignity, it is also closely connected to the principle of equality.⁶⁰ The right to vote is a “minimal condition of political equality”;⁶¹ it is also, crucially, a public expression of that civic equality.⁶² In the United States, the right to vote has long been connected to the principle of racial equality but in recent years it has also been understood through the lens of universalist principles of equal voting, which apply regardless of race.⁶³ To be sure, the expressive value of a constitutionally enshrined right to vote may change depending on the wording of the guarantee and the culture and history of the relevant jurisdiction.

In addition to expressing certain values, the right to vote also expresses and thereby establishes normative baselines concerning universal suffrage, political equality, and democratic representation. It is a fundamental normative commitment of a democracy that all citizens should have the right to vote.⁶⁴ The right to vote is viewed as a preservative right because it enables the protection of all other rights through the mechanisms of political participation, democratic representation, and accountable government.⁶⁵ These normative baselines of universal suffrage and political equality can set an outer limit for subconstitutional regulation. That is, the expressive value of the right to vote lies in serving as an outer boundary – at least in theory – for statutory and judicial restrictions on the right to vote. While these restrictions may undermine the right to vote, the existence of the normative baseline of universal suffrage would make it costly as a reputational matter to eradicate the constitutional protection altogether. A more general observation is that the aspirational dimension of constitutions and constitutional rights⁶⁶ can have a real-world impact on politics. The effect of aspirational constitutions may be subtle in that they help to thwart decay.⁶⁷ Or they may, as in the case of the right to vote, serve as a rallying point to mobilize citizens and bring about progressive political change.

To be sure, there are empirical difficulties associated with the claim that a constitutionally enshrined right to vote establishes a normative baseline of universal suffrage that can serve as an outer boundary for subconstitutional regulation. One challenge with assessing this claim is that it would be difficult to disentangle the effect of the constitutional dimension of the right to vote from the effects of its other aspects in a manner that would produce a definitive conclusion. Another complication is that legislatures and courts can undermine voting rights while still protecting universal suffrage. Even though all citizens may nominally have the right to vote, a state can enact voting rules that place considerable restrictions on the eligibility and opportunity to vote. It may be difficult to ascertain the extent to which the normative signal of constitutionally endorsed universal suffrage is constraining legislatures and courts.

A more significant challenge is that a constitutionally enshrined right to vote can also, paradoxically, undermine democracy. The electoral route to competitive authoritarianism is a well-documented avenue to democratic backsliding.⁶⁸ Once the autocrat is elected to office, the right to vote and associated constitutional rights and structures provide “democratic cover” to autocrats, furnishing them with democratic legitimacy as they strip out democratic safeguards. Competitive authoritarian regimes retain the trappings of an electoral system, including the right to vote, while going to considerable lengths to manipulate the election to entrench themselves in power.⁶⁹ Thus, they “practice authoritarianism behind the institutional facades of representative democracy.”⁷⁰ Although they hold regular multiparty elections with universal suffrage, they deploy a range of manipulative strategies to win elections, including prosecuting candidates, banning parties, intimidating voters, harassing journalists, and forging election results.⁷¹ Political parties and candidates in opposition face significant hurdles.⁷² Constitutional rules and entities, including the courts, are subverted to entrench the autocrat in power.⁷³ Private entities, such as the media, are likewise co-opted to support the regime.⁷⁴ Rather than providing for accountability and the possibility of new leadership, elections are designed to entrench the incumbent autocrats.

Autocrats hope that these periodic elections provide a “semblance of democratic legitimacy” for both domestic and foreign actors.⁷⁵ The expressive function of the right to vote and elections confers democratic legitimacy on autocrats even as they ensure their continued grip on power by undermining democracy. At the same time, the existence of even nominally competitive elections may incentivize autocrats to be somewhat more attentive to the needs of the people, at least in comparison to the leaders of fully authoritarian regimes. A formal recognition of a universal right of suffrage in the constitution could increase the costs for autocrats who might otherwise be tempted to eradicate elections. For this reason, a constitutionalized right to vote is normatively appealing notwithstanding its paradoxical nature.

8.6 Conclusion

In this chapter, I have claimed that while a constitutionally enshrined right to vote can help to strengthen voting rights in practice, it does not on its own provide a guarantee that the right to vote will be meaningful. The right to vote is both multidimensional and institutional, which suggests that the sub-constitutional components of the right make a far greater difference to the reality of voting than the bare fact that the right is constitutionalized. That being said, I have argued that, as a normative matter, constitutions should recognize the right to vote. A constitutionalized right to vote plays a crucial expressive function by promoting democratic values and establishing normative baselines concerning representation, universal suffrage, and political equality. These values and baselines can indirectly protect voting rights in a democracy.

However, the expressive function of constitutional rights and structures can, paradoxically, undermine democracy by providing autocrats in competitive authoritarian regimes with democratic cover while they are undermining constitutional safeguards. That being said, the reality of elections and voting, even when heavily manipulated, may exert some beneficial constraints on elected autocrats. For this reason, the claim that the right to vote ought to be constitutionalized is normatively justifiable.