A central function of a written constitution is to enhance the stability of the political system that underpins it. Constitutions often do this by adopting formal norms of entrenchment, or imposing formal barriers to constitutional change that depend on more than mere legislative majority support. In some systems, political conventions and practices may serve the same function and create effective informal barriers to constitutional change. These formal and informal barriers to change help to protect institutions and values by entrenching them. Thus, recent literature has suggested that one key metric for the success of a constitutional order is its overall “endurance”, or capacity to withstand pressures for constitutional change (Elkins, Ginsburg, and Melton 2009).

We argue in this chapter that for those constitutions with a democratic starting point, one important criterion for constitutional success is the degree to which the constitution preserves a minimal or thin conception of competitive democracy. We do not argue that this is the sole criterion for success, but it does allow for relatively structured comparisons across countries and it would be difficult to argue that a constitution failing this criterion has been successful. Thus, constitutional designers and scholars working from a democratic starting point should be most concerned with the endurance of those parts of a democratic constitution protecting this thin version of democracy: what one might call the minimum core of a democratic constitution. The endurance of the minimum core is normatively important in determining constitutional success because the failure to protect the minimum core will lead to the erosion of democracy. However, the desirability of change to the rest of a democratic constitution is more ambiguous from a normative perspective; such change may be harmful, neutral, or even affirmatively beneficial from the standpoint of constitutional theory.

This point has consequences for both the conceptualization and measurement of constitutional change. From a conceptual perspective, it suggests that constitutions are heterogeneous documents with different kinds of levels of provisions, and changes at these different levels have different normative consequences. We distinguish at least two different levels at which constitutions operate: the minimum core and more ordinary law.

INTRODUCTION

Constitutional ambiguity, vagueness, and indecision have been recently recognized by growing number of scholars and practitioners as a useful tool for mitigating conflicts over ideational issues during constitution-drafting processes (Bali and Lerner 2016; Brown 2008; Dixon and Ginsburg 2011; Lerner 2011; Shankar Forthcoming; Sunstein 2001). Under conditions of deep disagreements over the religious/national identity of the state or when the drafters are polarized around other ideological questions, it had been argued, drafters managed to craft democratic constitutional arrangements by using a variety of incrementalist or permissive constitutional strategies such as the use of vague and ambiguous language, the deferral of decisions to future parliaments, or even defining certain provisions as non-justiciable (Jacobsohn 2010; Lerner 2011). Yet what is the long-term impact of such incrementalist/permissive strategies? To what extent do they allow, for example, for the emergence or consolidation of a stable democratic order? Do they manage to promote or rather impede the protection of human rights, gender equality, or minority rights?

The long-term consequences of formal constitutions have become an emerging topic of interest in the field of comparative politics and comparative constitutionalism. Most works tend to explore the influence of particular provisions within formal constitutions, such as electoral rules or bill of rights (e.g. Horowitz 1985; Law and Versteeg 2013). By contrast to these studies, which look at the impact of formal written constitutional provisions, this chapter aims at exploring the consequences of decisions made by constitutional drafters which in many cases did not have a clear formal manifestation within a written constitution. In other words, the chapter aims at exploring the effect of framers’ decisions to leave clear choices on controversial issues outside the formal constitution by, for example, deferring these issues to future political deliberation, or by including ambivalent and opaque wording, or even conflicting provisions, in the constitutional text. Such permissive strategies were rarely employed by constitutional drafters when institutional issues, concerning the structure and mechanisms of government, were at stake, yet were often used in order to mitigate ideational conflicts, namely when the constitution was written under conditions of deep division over the basic values and norms that should underpin the state.

The chapter takes first steps in examining the impact of constitutional permissiveness in the area of one particular type of ideational conflicts – religious conflicts.

How should we evaluate constitutional performance? What should count as “success” in constitutional design? Is there a universal benchmark against which all constitutions, regardless of local circumstance, can be evaluated? Or is constitutional design as idiosyncratic as a person's choice in neckties? These questions, which are the focus of this volume, are necessarily raised by the emergent transnational practice of constitutional advice-giving and criticism. They are implicated every time a scholar, consultant, human-rights activist, or international organization expresses a position on a proposed constitution, whether in Somalia, Tunisia, Nepal, or the United Kingdom. They are thus necessarily questions for the governments and international organizations that fund such practices. And they are equally questions for the national publics engaged in the act of constitutional creation, who are often on the receiving end of international advice about what they should be doing. Finally, they ought to be puzzles for the growing coterie of scholars and jurists engaged in the comparative analysis and critique of new constitutions, a scholarly literature that often employs explicitly normative criteria in evaluating constitutional design. If we wanted to err on the side of grandiosity, we might even say they are questions implicated every time one decides that a constitution, as a going concern, merits our continued fidelity.

The contributors to this conference have been asked to respond, from a variety of perspectives, to the seemingly simple question of what counts as constitutional success (a term we will use interchangeably with constitutional performance here). By posing this concededly naïve question, we hope to draw attention to a normative terrain that has received surprisingly little attention from scholars and practitioners who assume, often implicitly, that there is a convergent consensus on what counts as “success” in constitutional design, and that therefore it is meaningful to praise or to blame a constitution for meeting or falling short of this desideratum. In so doing, we hope to provoke more careful debate among legal and political theorists about the plural possible meanings of constitutional success or quality. The chapters assembled in this book, we think, provide a series of important landmarks and provocations in that debate rather than a singular, definitive answer to our threshold question.

Who decides what a constitution's purpose should be?

A UNIVERSAL TEST

Scholars and practitioners have long struggled to identify what qualities make a constitution successful. The question is incredibly complex, not least because there is very little consensus on what constitutions should look like, or on the areas that constitutions should deal with. In an effort to resolve this issue, there is a natural tendency amongst many scholars to favor a “universal test,” that is, a single set of criteria that can be applied universally to measure the performance of all constitutions. While such an approach is certainly tempting, not least of all because of its simplicity, there are at least two sets of reasons that we should question its applicability in all cases.

The first is related to the fact that any attempt to establish external criteria will necessarily have to remain as general as possible. Constitutions are enacted in countries with widely varying starting points and are also subject to be influenced by a multiplicity of factors. A constitution can be enacted for the purpose of ending violent conflict (e.g. Bosnia), alleviating poverty (e.g. Ecuador) or transforming a previously discriminatory society into a more egalitarian one (e.g. South Africa), while others are designed merely to streamline government in what is already an effectively administered state (e.g. Switzerland). Some constitutions are well constructed but are applied in states that barely survive in unstable regions and that are subject to overbearing and negative foreign influence (e.g. Yemen), while others still are applied in countries that exist within a broader state of peace and economic development (e.g. Spain).

In that context, the only way to measure constitutional performance through the use of a universal test would be to ignore each country's nuances and idiosyncrasies and to remain focused on general objectives. The alternative would simply be unworkable: If we were to apply a narrow set of criteria to evaluate constitutions, the result would be that many of those individual tests would simply be inapplicable to a large number of contexts. For example, a universal test that measures constitutions’ success in promoting racial equality would be inapplicable in Tunisia, which is racially homogeneous.

An enterprise that seeks to compare the written provisions of all existing constitutions in terms of varying real-world outcomes is necessarily a highly complex one. Unfortunately there is yet a further complexity. The real workings of all constitutions depend in part not only on the institutional patterns they specifically provide but also on the nature of the party systems and the electoral arrangements of their polities, yet political parties and much of the electoral rules for most states are not provided for in their constitutions. Political regime A and political regime B might have identically worded constitutions but, nevertheless, vary greatly in their actual operations because of differing party systems and electoral rules. And, of course, party systems and electoral rules reciprocally influence one another. Each to some degree “causes” the other.

In the light of these considerations, I want to very tentatively and incompletely sketch how three general types of separation of power constitutions each may operate differently depending on which kinds of party systems and electoral rules are at play. Suppose one wished to compare parliamentary, presidential, and quasi- or semi-presidential constitutions to discover which type of constitution led to greater political stability, democracy, or other real-world outcome. Or suppose we pursued a narrower goal, say, asking in which of these constitutional types was judicial review more likely to succeed. Our findings might well be confounded if two parliamentary constitutional regimes operated in quite different ways because one enjoyed a two-party and the other a multi-party system. Showing the two together as if they were the same and then comparing them with presidential regimes to see which yielded more complete democracy or interest representation or whatever would yield very unsatisfactory results. You can compare apples to oranges if all the apples are pretty much the same as one another and all the oranges are about the same. You can't if some of the apples are very different from one another as are some of the oranges. If two-party presidential regimes actually are quite different from multi-party presidential regimes, it may not be easy to compare presidential constitutions with parliamentary constitutions that may themselves vary greatly in their actual operations because of differences in party systems.

INTRODUCTION

It is my impression that the notion of “constitutional performance” refers to an evaluative rather than to a statistical process. But, of course, everything depends on how we define the terms we want to use. In my view, in order to determine whether a Constitution is performing “well” we need to initiate a complex theoretical reasoning, which I shall begin to explore. Alternatively, however, we may stipulate that (say) the idea of “constitutional success” means something like (say) constitutional stability – this being a historically important candidate for playing this role, as I shall suggest. In this way, we gain the possibility of measuring the concept but – this is my opinion – at the risk of undermining the attractiveness, complexity and richness of the notion of “constitutional performance.” Of course, it is absolutely important to compare Constitutions according to (say) the stability they achieved, and to measure such things. But it is not clear to me that, in this way, we will be measuring “constitutional performance.” In association with the notion of “constitution,” the idea of “performance” appeals to a complex process, which – it is my impression – should not be simply reduced to a statistical analysis.

In what follows – and based on previous studies about constitutionalism in the Americas – (Gargarella 2010, 2013), I will explore this discussion with some more detail, and propose an approach to “constitutional performance” that is both contextually and normative sensitive, and which is based on the teachings of nineteenth-century constitutional scholars. I will illustrate the potential of this alternative understanding through four examples related to early constitutional history in the Americas.

WHEN IS A CONSTITUTION DOING WELL? THE EXAMPLE OF THE CHILEAN CONSTITUTION OF 1833

After the independence years – around 1810 in Latin America– most countries in the region began to explore ways to (re)organize their institutional structure, usually with the help of a new Constitution. The conditions within which these new Constitutions grew were extremely difficult for a variety of reasons, including the fact that the new societies were socially divided and characterized by profound inequalities; the presence of strong caudillos who represented, in many occasions, a serious “internal” threat to the new projects; and also the existence of the “external” threat posed by foreign countries (most typically Spain, trying to re-built its lost Empire).

INTRODUCTION

On many of the external metrics outlined in the Introduction to this volume, the United Kingdom ranks among the most successful nation states: It is a functioning democracy with robust rights protection; its public sector is one of the least corrupt; and its citizens benefit from an array of public goods, including health care, education, and a variety of other social welfare programs. From an internal perspective, the flexible British constitution appears well matched to the wants of its citizenry – as J.A.G. Griffith wrote more than thirty-five years ago, “the constitution is no more and no less than what happens. Everything that happens is constitutional. And if nothing happened that would be constitutional also” (Griffith 1979: 19). In short, there is no obvious gap between the aspirational and the actual.

This exceptional performance suggests we should look to the United Kingdom as a model for constitutional design. But, of course, the British constitution sits uneasily in contemporary constitutional studies. As an uncodified set of written laws and political practices with ancient roots (e.g., Magna Carta 2015) and modern fruit (e.g., The Fixed-Term Parliaments Act 2011), the constitution is difficult to pin down with precision and certainty. Even senior judges and politicians struggle: Lord Scarman described the constitution as “hidden, and difficult to find” (1993: 319), and Lord Callaghan suggested it has a certain “back of an envelope” quality to it (Hennessy 1995: 6). Academics agree that the specific substantive elements that make up the constitution are subject to debate (Hennessy 2007: 346) as well as that different understandings of the constitution – in Scotland and England, in particular – have coexisted for centuries (Feldman 2005: 347).

Yet, even if it were possible to extrapolate a modern model from this historically contingent and idiosyncratic system, an antecedent question niggles: How confident should we be in this positive assessment of the United Kingdom's constitutional success? A closer look at life in the British state uncovers roiling constitutional politics. Complaints abound of constitutional uncertainty, constitutional malaise, constitutional anomie – even constitutional crisis (Bogdanor 2015; Webber 2014). This discourse reveals a higher level of dissatisfaction than might be expected in a country that otherwise seems a paragon of constitutional success.

In the late 1990s, former Turkish president and prime minister Suleyman Demirel was asked to comment on an ongoing crisis between the civilian and military leadership in his coup-prone country. Demirel replied with a joke. There was an experiment in an English zoo, he explained, to place lambs and wolves in one cage in order to teach them how to live together. The zoo director was asked if the experiment was working. The director replied: “Yes, but from time to time, we have to replace the lambs.”

A transition from military to civilian rule often forces military and civilian leaders to live together as wolves and lambs in one cage for a period of time. Among other things, they must agree on a framework for the transition, which often, but not always, includes the amendment or replacement of the existing Constitution to govern the new civilian regime. During the constitutional-design process, the military often attempts to entrench into the revised Constitution provisions that protect its institutional autonomy and perpetuate its influence in civilian governance.

Constitutional entrenchment of the military's prerogatives creates a wealth of theoretical and empirical questions about constitutional performance. Depending on context, constitutional entrenchment can be a reliable method of protecting the autonomy and privileges of the military and its civilian allies, at least in the short term. Even from the perspective of other relevant actors, entrenchment may be a second-best outcome. Although immediate and complete ouster of the military from domestic politics may be possible in some transitions, in others, any attempts to swiftly force the military back to the barracks can prompt a backlash from the military, which might dig in, rather than give in, and derail the transition process. And in some cases, some level of military involvement may be desired in the initial stages of the transition so that the military can maintain stability and act as an arbiter of constitutional bargains between competing political groups.

Entrenchment of the military's prerogatives, however, can also undermine long-term democratic development. As a result of entrenchment, the military may emerge as a de facto, if not de jure, separate branch of government from the transition process. For democracy to persist, the military must ultimately retreat to the barracks and become subordinate to democratically elected civilian leaders.

Between 2010 and 2013 Iceland engaged in an unprecedented experiment in peacetime constitutional redrafting. The process included innovative participatory methods, including a National Forum, an elected Constitutional Assembly of non-professional politicians, and the well-known use of online crowdsourcing for twelve successive drafts of the constitutional proposal (which earned the latter its international, and perhaps overhyped, reputation as the “crowdsourced constitution”). Begun in the aftermath of a cataclysmic banking and financial crisis, which was estimated to have destroyed assets equivalent at the time to seven times the country's annual GDP, and spurred on by the new political personnel that came to power after the “Pots-and-Pans” revolution of 2008, the constitutional process ultimately led to a proposal that was approved by a two-third majority of the voters in the fall of 2012. After further amendments by legal experts, the proposal was offered to the Althingi (the Icelandic Parliament) as a bill to be discussed and voted on in the spring of 2013. To many commentators’ surprise, however, the bill was shelved at the eleventh hour for reasons that seem more political than substantive (Gylfason 2013b). The crowdsourced constitutional proposal is now considered dead by its opponents and “on ice” by its advocates (2014). It is unclear at this point whether it still stands a chance to be adopted as such in the near future.

The question that interests us here, however, is whether the crowdsourced proposal meant to replace the 1944 constitution was a good constitution – not merely in the sense of having been produced by the right kind of process but in a thicker, more substantive sense, which is entirely independent of the process that led to it. Previous work on the Icelandic experiment has focused on procedural aspects of the constitutional process. For example, in past work I argued that the Icelandic constitutional process aimed to be inclusive of the full diversity of the Icelandic people's views and to the extent that it succeeded, could be expected to have produced a substantively sound constitution, that is, a constitution tracking relevant facts and values of the Icelandic people in the twenty-first century and channeling its collective wisdom (Landemore 2015).

INTRODUCTION

Is there a criterion that can be used to evaluate constitutions written in Pretoria in 1996, Philadelphia in 1787, and Baghdad in 2005? Can the organic documents produced by democrats, oligarchs, tyrants, and theocracies all be judged by the same metric? This chapter, in taking up these questions, develops a benchmark for constitutional success with a general, albeit not universal, scope. That benchmark is independent of local criteria or prejudices, and relies instead on an analysis of the minimal conditions for constitutional success. Focusing my inquiry in that fashion, I contend that the most plausible touchstone of constitutional success is the avoidance of self-defeating constitutional design. It is a constitution purged of elements that conduce, perhaps unintentionally, to the breakdown or collapse of the state that the constitution endeavors to underwrite.

To situate this inquiry, it is useful to recall the dichotomy that Tom Ginsburg and I posit between metrics for evaluation of constitutional “success” that are internal and ones that are external in character. An internal criterion is one applied by a member of the polity engaged in constitutional creation or evaluation. An external criterion, by contrast, is one employed by an outsider such as a comparative constitutional scholar or a foreign advisor to constitutional drafters. So defined, the internal/external distinction turns on who applies the standard, not what the standard is. Human rights norms, for example, might be invoked both by participants to a drafting process as an internal criteria, or instead by transnational advocacy groups as a generally applicable external standard. Moreover, there is often leakage between internal and external criteria. Evaluative criteria deployed by participants in a constitution-making process are not autochthonic. They can leach their content and orientation from global intellectual currents (Ginsburg et al. 2008). Nor are external criteria necessarily loosed entirely from the gravitational force of the local circumstances of their intellectual production. Political rationality, in my view, lacks any transhistorical, acontextual form. Nevertheless, the distinction is useful insofar as it draws attention to the fact that criteria of constitutional success play different roles in different circumstances. And different standards might fit those distinct circumstances better than a single norm.