Constitution-making and liberal democracy: The role of citizens and representative elites (2022)

Abstract

This article discusses the impact of citizen participation and elite cooperation in constitution-making on the deepening of an already existing electoral democracy. It argues that while direct citizen involvement in the drafting of constitutions may be desirable on normative grounds or necessary for pragmatic reasons, only cooperation among a plurality of elected political representatives at the constitution-making stage is likely to improve the liberal dimension of democracy after the enactment of the new constitution. Inclusive constitutional agreements at the level of representative elites not only establish legal limits on state action but may also provide opposition parties and citizens alike with the means to make institutional constraints on executive power and civil liberties effective. This effect is usually observed during the early years of life of the new constitution, when the balance of power among the political forces that created the constitution tends to remain stable. I find preliminary support for this argument analyzing aggregate data and selected case studies from all episodes of democratic constitution-making in the world between 1900 and 2015.

1. Introduction

Normative theories of constitution-making have made competing claims about the features of this process that have the potential to establish or improve a democratic regime. In one widely shared view, the involvement of citizens before, during, and after constitution-writing is supposed to enhance the sense of collective ownership over the new text, promote a democratic institutional design, and facilitate its enforcement. A different perspective emphasizes the importance of elite accommodation and cooperation to promote legality and consensus-building among the major political forces. Neither theory, however, has examined the relative weight of these arguments by analyzing conceptually and empirically the exact mechanisms by which the actions of citizens and elites may affect democratization.

In this article I discuss the impact of citizen participation and elite cooperation in constitution-making on the deepening1 of an already existing electoral democracy. I argue that while direct citizen involvement in the drafting of constitutions may be desirable on normative grounds or necessary for pragmatic reasons, only cooperation among a plurality of elected political representatives at the constitution-making stage is likely to improve the liberal dimension of democracy after the enactment of the new constitution. Inclusive constitutional agreements at the level of representative elites not only establish legal limits on state action but may also provide opposition parties and citizens alike with the means to make institutional constraints on executive power and civil liberties effective. This effect is usually observed during the early years of life of the new constitution, when the balance of power among the political forces that created the constitution tends to remain stable. I find preliminary support for this argument analyzing aggregate data and selected case studies from all episodes of democratic constitution-making in the world between 1900 and 2015.

This article is organized as follows. Section 2 offers a critical review of existing constitution-making theories and their relation with explanations that emphasize the primary role of citizens or elites in initiating or deepening a process of democratization. Section 3 analyzes the different features that promote cooperation between representative elites and popular participation during the adoption of a new constitution in an electoral democracy. Sections 4 and 5 offer quantitative and qualitative evidence that is consistent with the proposal that elite cooperation and not citizen participation in constitution-writing contributes to improving democracy in its liberal dimension. Section 6 presents conclusions.

2. Citizens, elites, and democracy in constitution-making theory

Most theories of constitution-making are predominantly normative, as they attempt to respond to the question of what rules political actors should choose if they intend to create a legitimate constitution. These theories, however, often imply that following their normative prescriptions at the constitution-making stage would result in a stable democracy or a stronger one, if it already existed. One view emphasizes the role of public participation in constitution-making in enabling citizens to monitor the actions of elected officials through their increased awareness of existing rules and rights. An alternative position stresses the importance of elite cooperation during constitution-writing for the creation of a consensus among the main political forces about the constitutional rules that would regulate democratic competition. As we will see, both theories share several limitations. However, the proposition highlighting the importance of elite accommodation during constitution-making provides the most appropriate starting point for examining the impact of constitutional change on democratization in comparative perspective.

The most traditional and still very influential theory of constitution-making emphasizes the role of the people as the founder of democratic constitutions through indirect channels of citizen involvement. This theory was born out of the great revolutions of the late eighteenth century and presupposes a radical rupture with the preexisting political and legal order. As a reflection of its historical struggle against oppressive monarchical government, a distinctive claim of the revolutionary theory of constitution-making was that only the people are the legitimate holders of constituent power.2 As Thomas Paine summarized it, “a constitution is not the act of a government but of a people constituting a government.”3

The idea of the people as a collective author of the constitution was subject to different conceptualizations in the American and French revolutionary traditions. Whereas in America “the People” was understood as a plural association of preconstituted territorial entities, in France, “the nation” was conceived as a single collective subject that was legally unbound in relation to existing institutions.4 However, in respect to how the People should express its will, revolutionary theories were inspired in a republican view that usually rejected direct forms of citizen participation.5 For this reason, the popular origins of constitutions often referred to a founding principle that could be satisfied by representative channels, such as the election of a constituent assembly or the ratification of the constitution through elected conventions.6

Recent formulations of participatory constitution-making have in mind either the adoption of a new constitution in societies affected by the complete breakdown of representative institutions or the involvement of ordinary citizens in important democratic reforms. In regard to how citizens should be involved, contemporary advocates of participatory constitution-making take the idea of popular authorship to its natural conclusion and claim the need for actual and direct and not just fictional or indirect citizen participation in constitutional change.7 Although this perspective does not presuppose discontinuity with the preexisting political and legal order, it does assume either the inability to resort to normal representative procedures, as in the adoption of a new constitution in some post-conflict societies, or the mere need to complement representative institutions with direct citizen involvement in significant democratic reforms that do not necessarily involve the drafting of a new constitution.8 Cooperation among the elite is not excluded from this analysis but plays a normatively less important role than the direct engagement by citizens in processes of deliberation and voting on constitutional change.9

In relation to the potential consequences of constitutional change for the inauguration or deepening of democracy, some advocates of participatory constitution-making take a more or less strict principled position. In this perspective, citizens have a right to participate in the making of the constitution because they must consent to the higher norm that will bind them in the future.10 In a similar vein, it has been argued that regardless of its effects on democracy or other outcomes, citizen participation in constitution-making enhances a collective sense of ownership over the constitution and thus its legitimacy.11 More often, however, arguments supporting citizen participation in constitution-making tend to emphasize, either implicitly or explicitly, its potentially positive consequences for democratization.

One general argument is that direct citizen involvement in the making of constitutions increases public awareness of accepted behavior under the new constitution, which, in turn, enables citizens to monitor elected officials and prevent transgressions.12 Some works have hypothesized that participatory processes, in particular the popular ratification of new constitutions, are likely to lead to the formal expansion of rights and reforms strengthening citizen influence and control over representatives.13 Others have proposed that participatory constitution-making may induce reformers to impose more constraints on government authority.14

A perspective that differs from both the traditional constituent power doctrine and contemporary theories of participatory constitution-making highlights the importance of legal continuity, gradualism, and, above all, the central role that negotiation and deliberation among political elites should have for the foundation of democratic constitutions.15 This theory has been inspired by some of the gradual and negotiated transitions to democracy that took place in countries such as Hungary and Poland in the late 1980s. The most prominent example is Arato’s “post-sovereign” constitution-making model.

According to Arato, the central features of a democratic constitution-making process should include: (i) a two-stage process with initial adoption of an interim constitution whose rules constrain the adoption of the final constitution; (ii) drafting of the interim constitution by a non-elected round table, and drafting of a final constitution by a freely elected constituent legislature; (iii) the use of existing amendment rules to provide legal continuity to the process; and (iv) enforcement of the interim constitution by a constitutional court. Direct forms of citizen involvement, particularly popular referenda, are not required and may even be counterproductive for the realization of this model.16

The post-sovereign theory has a clear normative goal, which is the rejection of the idea of a unified popular sovereign, characteristic of some versions of the constituent power doctrine. However, it also proposes that following the main components of the model would help not only to construct democratic legitimacy for the new constitution but also to facilitate the consolidation of a democratic regime by promoting consensus among the major political forces, a plurality of channels to express popular consent, publicity, and the rule of law.17 Moreover, this theory also has a practical, political dimension. Just as legal discontinuity, special conventions, and popular authorship suppose a revolutionary setting in which the old regime has lost legitimacy and its elite must be displaced from the new order, the ideas of legal continuity, parliamentary constitution-making, and indirect citizen involvement reflect the need for elite cooperation in some transitions to democracy. In particular, gradualism and multiparty round-table agreements are key elements when excluding and not protecting the interests of the outgoing authoritarian elite might obstruct the initiation of the transition or consolidation of the new democratic regime. Non-reliance on referenda, in turn, is intended to secure outcomes negotiated beforehand among elites, either in non-elected forums or in a newly elected parliament.

In spite of their different emphasis on citizen participation or elite negotiations, these theories of constitution-making have similar shortcomings as frameworks for the comparative analysis of the relationship between constitutional change and democratization. They portray constitution-making as merely a formal, legalistic process and fail to specify the actual mechanisms through which the actions of citizens and political elites at the drafting stage might have an impact on the implementation stage of new constitutions. These theories also lack generality because they are meant to apply to specific political environments, such as constitutions drafted as a consequence of revolutions, civil wars, or negotiated transitions to democracy. Nevertheless, a perspective emphasizing the key role of elite cooperation provides a critical vantage point for understanding how the dynamics of constitution-making might have an effect on democratization.

Although the interdisciplinary dialogue between comparative constitutional theory and comparative political science on this matter has been sparse, there is a long tradition of research in the social sciences that focuses on the central importance of decisions made by political elites for the inauguration of a stable democracy and, potentially, its future improvement.18 Seminal works in this research agenda have argued and provided a significant amount of evidence in support of the idea that a procedural compromise among the leaders of contending political forces is crucial for a democratic opening.19 One key role of this compromise is to create a set of rules of mutual security that make it unlikely that the subsequent competitive political process would result in outcomes highly adverse to the interests of any of the main political and social groups.20 Inspired by this line of reasoning, it has also been argued that elite settlements and pacts in which the main political actors commit to follow rules of mutual security are the very foundation of a self-enforcing or consolidated liberal democracy.21

The relevance of these propositions to constitution-making is clear. There is no doubt that in many cases mass mobilization can be crucial for promoting a democratic opening or for making possible reforms to a deficient democracy.22 It is also likely, particularly in the face of preceding events of mass action, that channels of citizen participation may make constitutional changes more legitimate in the eyes of the general public. Yet the drafting of constitutions has historically been (and still is) a predominantly elite affair. Representatives of political parties and leaders from the most important social groups have usually been the ones who decide how constitutions should be drafted and what content they will have. This content, in turn, depends on the distribution of political power among the main political forces.

When none of the political groups involved in a constitution-making process have the popular support or the institutional resources to make decisions alone, they are likely to cooperate in creating institutions that protect the interests of all the parties involved. These institutions usually take the form of rules that create legislative and judicial constraints on incumbent governments, reduce the power of electoral and legislative majorities, and establish rights that protect the interests of all the relevant groups in society.23 More specifically, a constitution created by contending forces will tend to include formal rules that provide a standard for the detection of constitutional transgressions and legal mechanisms to react against them.24 In other words, accommodation and compromise among different fractions of the political elite at the time of writing the constitution are likely to produce a constitutional design that enhances the principles of liberal democracy and provides legal means for their realization.25

To be sure, there is no reason to expect that liberal institutions would matter for the future democratic regime if the constitution is not implemented and observed in practice. What matters is an actual change in behavior, not just in formal rules. The role of elites is again crucial in this respect.26 When a plurality of representatives of organized political and social interests has participated in the constitutional agreement, more actors are likely to have both the incentives and the resources to enforce that agreement later on.27 After the constitution is enacted, it is generally the opposition political leaders who are more inclined to react when those in power renege on the initial constitutional compromise. And they would not act alone. Representatives of opposition political groups would, among other actions, mobilize public opinion or organize mass actions in defense of the constitution. This suggests, in turn, that during electoral competition it is crucial that political forces maintain a relatively even distribution of voter support. If one of the participants in the initial constitutional compromise gains control over the government and becomes a dominant actor in the electoral arena, it would be more difficult to monitor and sanction transgressions to the constitution.

Because the power of political elites depends on their social support, the actions of citizens matter for the effective implementation of a democratic constitution; in particular, it matters whether citizens are willing and able to vote against incumbents or engage in massive acts of social protest when the government infringes constitutional provisions. Yet these actions are not determined by the direct participation of citizens during constitution-writing. Citizen participation in some processes of public consultation and voting on a new democratic constitution may indeed be an empowering experience. It is also reasonable to suppose that among those motivated to participate and obtain information about the process, involvement in constitution-making will increase their knowledge about the content of the constitution and its importance in political life. It is doubtful, however, that participatory constitution-making, just by itself, would deepen democracy after the new constitution is in force. Although some arguments about the democratic benefits of participatory constitution-making have been tested with positive results, the underlying logic that links direct citizen involvement during the writing of the constitution with actual levels of democratization after the event is generally weak.

The hypothesis that makes the most sense is that active citizen involvement in constitution-making may provide reformers with an incentive to expand citizen rights.28 Ratification referenda, for instance, create what Elster calls a “downstream” constraint on the decisions that reformers can make.29 If reformers know or anticipate the preferences of those who have the power to accept or reject their proposals, they have every incentive to satisfy those preferences beforehand. Regardless of whatever else they include in the proposal, the expansion of citizen rights can be presented and is likely to be regarded by significant segments of the population as an improvement in collective welfare. If citizens express demands for the inclusion of various rights during processes of consultation, reformers may also decide to expand the list of rights to address some of these demands. However, there is no reason to think that citizen voting or consultation during constitution-making would lead to the creation of a set of rights that protects all major social groups. More importantly, even if reforms were designed in an impartial manner, majority voting or other forms of public participation during their adoption do not guarantee that they will be effectively implemented after the new constitution is enacted.

Unlike the case of reforms related to the expansion of rights, the rationale behind the proposal that citizen participation in constitution-drafting may lead to a strengthening of constraints on the executive power is not apparent. In the first place, ordinary citizens are not likely to have well-defined and fixed general preferences on this matter. Moreover, one can think of many cases of constitution-making against the background of a deep economic or political crisis, where people may be willing to support stronger rather than weaker executive authority.30 And, of course, even if there were circumstances under which citizens would demand increasing formal constraints on the executive, there is no reason to suppose that these constraints would in fact be implemented after the constitution is approved.

Finally, and crucially, the argument that direct citizen involvement during the drafting or approval of new constitutions will increase public awareness about existing rules and rights, facilitate the detection and sanction of transgressions, and thus prevent self-serving behavior by elected authorities rests on dubious assumptions about the preferences of citizens regarding the content of constitutions and their collective capacity to act in defense of legality. Citizen participation in the formulation, discussion, or promulgation of a new constitution does not generate consensus about the rules and rights that should be included in it. Citizens are often divided along ethnic, religious, ideological, or socioeconomic lines, and these divisions are not likely to disappear just because citizens participate, even through deliberation channels, during the writing of a new constitution.31 If it emerges at all, a general agreement among citizens about the rules and rights that should be respected by the state initially depends on successful negotiation of the content of the constitution among a plurality of leaders representing the diversity of interests in society.32 In addition, even if they were to agree on what rights should be universally protected, citizens do not normally have the ability to mobilize spontaneously against an incumbent government that transgresses the constitution. Aside from some episodic outbursts of protest, the capacity of the masses for sustained and effective mobilization usually depends on the leadership or organizational resources provided by the political and social elites who oppose incumbents.33

The preceding discussion suggests that while the direct participation of citizens in democratic constitutional replacements may be normatively desirable or politically convenient, only cooperation among representative elites at the constitution-making stage is likely to improve the liberal dimension of democracy after the enactment of the new constitution. Inclusive constitutional agreements at the elite level not only establish legal limits on state action but may also provide the organized political opposition and ordinary citizens with the means to make institutional constraints on executive power and civil liberties effective. In most cases, however, this effect should be observed during the early years of life of the new constitution, when the balance of power among the political forces that created the constitution tends to remain stable.

These propositions are sufficiently general to apply to a wide variety of political environments. There are several reasons, however, why it is worth examining them in relation to constitutional replacements that take place in a regime that is already democratic in at least the minimal, electoral sense. The first is that unlike some cases of revolutionary ruptures and democratic transitions, both citizen participation and elite cooperation may be required when a constitution is adopted after the establishment of free and fair elections. Whereas in revolutionary overthrow situations established elites tend to be displaced, in some transitions to democracy citizen mobilization may be disruptive to sustaining necessary inter-elite agreements. Democratic constitutional replacements, in turn, are often necessary to address public criticism of the working of inherited representative institutions. To achieve legitimacy and placate disaffected citizens these reforms are likely to demand direct public involvement during drafting and approval of the constitution. At the same time, even if suddenly discredited by a crisis, preexisting parties and political leaders in an electoral democracy usually express the preferences of a significant number of citizens because they have achieved power positions through free and fair elections. For this reason, they have a legitimate claim to inclusion in the negotiations and deliberations leading to the adoption of a new constitution.

Second, whereas cooperation among political elites can take place under authoritarian or democratic conditions, citizen participation is less likely to be genuine when civil liberties and competitive elections have not been fully reestablished. Determining when processes of public consultation are consequential for the design or implementation of a new constitution is always a difficult matter. Yet it is clear that those processes are more likely to be meaningful when the basic rights of assembly and freedom of expression are enforced. Voting in referenda can be informative as to popular preferences and may constrain the decisions of reformers only if we assume that citizens are free to participate and their votes are counted fairly. In other words, although elite manipulation of citizen participation is always possible, true citizen involvement in constitution-making is more likely when a regime is already at least minimally democratic.

Finally, studying constitutional change once an electoral democracy is in place is also appropriate because it allows a better appreciation of the difference that the adoption of a new constitution makes in terms of improving or not improving dimensions of democracy that go beyond the mere existence of free and fair elections. One of these dimensions, the most basic after equitable conditions of electoral competition are met, is the realization of liberal principles, such as the implementation of legislative and judicial constraints over the executive and the effective protection of citizen rights. This dimension is crucial not only because it improves a merely competitive regime but also because it relates to the success of the ideal of limited government of classic constitutionalism.34 By contrast, when the reference point prior to the creation of a new democratic constitution is an authoritarian regime, democratic conditions in a country are likely to improve in almost every dimension, not necessarily because a new constitution is in place.

3. Features of constitutional replacements in democratic regimes

The phenomenon of constitution-making in democratic regimes has been neglected by traditional analyses of constitution-making because it is often assumed that new constitutions only emerge as a consequence of the foundation of a new political order. Yet historically, this is not true. Although constitutional replacements in democratic regimes have not been as frequent as those that have taken place during the creation of a new state, a revolution, or a regime transition, they are sufficiently numerous to warrant scholarly attention. In addition, given the growing dissatisfaction with the performance and quality of contemporary democracies in various regions of the world, it is likely that this phenomenon will increase in the future.35

It is, of course, difficult to define exactly when a democracy is established or consolidated in the sense that an authoritarian reversal is no longer possible or highly unlikely. But one can agree, as most political scientists do, that an electoral democracy exists when the first free (without proscriptions) and fair (without open state intervention in favor of the government party) election takes place in a country that previously had a non-democratic form of government. From this perspective, an intra-democratic constitutional replacement takes place when a new constitution is adopted sometime after the founding election. Five years is a reasonable time. At this point, at least one competitively elected government has usually completed its term and representative legislative and executive institutions have clearly replaced the institutions of the authoritarian regime. Also, among both new and old democracies a period of five years provides a considerable and yet manageable time frame to compare political conditions before and after a new constitution is enacted.36 Using this criterion, a total of twenty-five new constitutions have been adopted in independent democratic states between 1900 and 2015.37Table 1 lists these cases, their geographic location, and the reason for change.38

Table 1.

Constitutional replacement in democratic regimes, 1900–2015.

CountryYearRegionSubregionReason for replacement
Denmark1915EuropeWesternDemocratization
Denmark1953EuropeWesternDemocratization
Finland2000EuropeWesternModernization/democratization
France1958EuropeWesternPolitical crisis
Greece1952EuropeWesternDemocratization
Iceland1944EuropeWesternDemocratization
Ireland1937EuropeWesternBalance-of-power shift/ democratization
Sweden1974EuropeWesternDemocratization/modernization
Switzerland1999EuropeWesternModernization
Hungary2011EuropeEasternBalance-of-power shift
Poland1997EuropeEasternDemocratization
Ukraine1996EuropeEasternDemocratization/state-building
Kenya2010AfricaSub-SaharanPolitical crisis
Thailand1997AsiaEastDemocratization
Sri Lanka1972AsiaSouthBalance-of-power shift
Nepal2015AsiaSouthDemocratization
Bolivia2009Latin AmericaAndeanPolitical crisis/balance-of- power shift
Colombia1991Latin AmericaAndeanPolitical crisis
Ecuador1998Latin AmericaAndeanPolitical crisis
Ecuador2008Latin AmericaAndeanBalance-of-power shift
Uruguay1952Latin AmericaSouthDemocratization
Uruguay1967Latin AmericaSouthDemocratization
Venezuela1999Latin AmericaAndeanBalance-of-power shift
Dominican Republic2010CaribbeanSouthDemocratization
Trinidad & Tobago1976CaribbeanSouthPolitical crisis/democratization
CountryYearRegionSubregionReason for replacement
Denmark1915EuropeWesternDemocratization
Denmark1953EuropeWesternDemocratization
Finland2000EuropeWesternModernization/democratization
France1958EuropeWesternPolitical crisis
Greece1952EuropeWesternDemocratization
Iceland1944EuropeWesternDemocratization
Ireland1937EuropeWesternBalance-of-power shift/ democratization
Sweden1974EuropeWesternDemocratization/modernization
Switzerland1999EuropeWesternModernization
Hungary2011EuropeEasternBalance-of-power shift
Poland1997EuropeEasternDemocratization
Ukraine1996EuropeEasternDemocratization/state-building
Kenya2010AfricaSub-SaharanPolitical crisis
Thailand1997AsiaEastDemocratization
Sri Lanka1972AsiaSouthBalance-of-power shift
Nepal2015AsiaSouthDemocratization
Bolivia2009Latin AmericaAndeanPolitical crisis/balance-of- power shift
Colombia1991Latin AmericaAndeanPolitical crisis
Ecuador1998Latin AmericaAndeanPolitical crisis
Ecuador2008Latin AmericaAndeanBalance-of-power shift
Uruguay1952Latin AmericaSouthDemocratization
Uruguay1967Latin AmericaSouthDemocratization
Venezuela1999Latin AmericaAndeanBalance-of-power shift
Dominican Republic2010CaribbeanSouthDemocratization
Trinidad & Tobago1976CaribbeanSouthPolitical crisis/democratization

Open in new tab

Table 1.

Constitutional replacement in democratic regimes, 1900–2015.

CountryYearRegionSubregionReason for replacement
Denmark1915EuropeWesternDemocratization
Denmark1953EuropeWesternDemocratization
Finland2000EuropeWesternModernization/democratization
France1958EuropeWesternPolitical crisis
Greece1952EuropeWesternDemocratization
Iceland1944EuropeWesternDemocratization
Ireland1937EuropeWesternBalance-of-power shift/ democratization
Sweden1974EuropeWesternDemocratization/modernization
Switzerland1999EuropeWesternModernization
Hungary2011EuropeEasternBalance-of-power shift
Poland1997EuropeEasternDemocratization
Ukraine1996EuropeEasternDemocratization/state-building
Kenya2010AfricaSub-SaharanPolitical crisis
Thailand1997AsiaEastDemocratization
Sri Lanka1972AsiaSouthBalance-of-power shift
Nepal2015AsiaSouthDemocratization
Bolivia2009Latin AmericaAndeanPolitical crisis/balance-of- power shift
Colombia1991Latin AmericaAndeanPolitical crisis
Ecuador1998Latin AmericaAndeanPolitical crisis
Ecuador2008Latin AmericaAndeanBalance-of-power shift
Uruguay1952Latin AmericaSouthDemocratization
Uruguay1967Latin AmericaSouthDemocratization
Venezuela1999Latin AmericaAndeanBalance-of-power shift
Dominican Republic2010CaribbeanSouthDemocratization
Trinidad & Tobago1976CaribbeanSouthPolitical crisis/democratization
CountryYearRegionSubregionReason for replacement
Denmark1915EuropeWesternDemocratization
Denmark1953EuropeWesternDemocratization
Finland2000EuropeWesternModernization/democratization
France1958EuropeWesternPolitical crisis
Greece1952EuropeWesternDemocratization
Iceland1944EuropeWesternDemocratization
Ireland1937EuropeWesternBalance-of-power shift/ democratization
Sweden1974EuropeWesternDemocratization/modernization
Switzerland1999EuropeWesternModernization
Hungary2011EuropeEasternBalance-of-power shift
Poland1997EuropeEasternDemocratization
Ukraine1996EuropeEasternDemocratization/state-building
Kenya2010AfricaSub-SaharanPolitical crisis
Thailand1997AsiaEastDemocratization
Sri Lanka1972AsiaSouthBalance-of-power shift
Nepal2015AsiaSouthDemocratization
Bolivia2009Latin AmericaAndeanPolitical crisis/balance-of- power shift
Colombia1991Latin AmericaAndeanPolitical crisis
Ecuador1998Latin AmericaAndeanPolitical crisis
Ecuador2008Latin AmericaAndeanBalance-of-power shift
Uruguay1952Latin AmericaSouthDemocratization
Uruguay1967Latin AmericaSouthDemocratization
Venezuela1999Latin AmericaAndeanBalance-of-power shift
Dominican Republic2010CaribbeanSouthDemocratization
Trinidad & Tobago1976CaribbeanSouthPolitical crisis/democratization

Open in new tab

It is noticeable that the vast majority of constitutions adopted in an electoral democracy are located in Western Europe and in South and Andean Latin America. The reason is that historically these regions have had a larger number of democratic regimes than other parts of the world. Although the political and legal contexts where democratic constitutional replacements took place have been widely diverse, there seems to be a common set of causes that explain them, the most frequent being the democratization of basic institutions, political crises, and drastic balance-of-power shifts among political actors. These causes are obviously not mutually exclusive, although one or another is usually predominant in any given case.

What features of these cases are likely to signal the level of elite cooperation and citizen participation that take place during the creation of a new constitution? The degree of elite cooperation is associated with certain characteristics that provide various political actors with mutual guarantees of inclusion and control over constitution-making, such as legal continuity, institutional checks, and, most importantly, a politically plural representative body responsible for drafting the constitutional text. The level of citizen participation, in turn, depends on the number of instances of direct involvement by ordinary citizens in the formulation of reform proposals and in voting to approve or reject those proposals. Table 2 illustrates the presence or absence of these features across the set of cases.

Table 2.

Features of constitution-making in democratic regimes, 1900–2015.

CountryYearLegal continuityInstitutional checksPlural constituent bodyPublic consultation/ submissions*Popular referendum**
Denmark1915YesYesYesNoYes
Denmark1953YesYesYesNoYes
Finland2000YesYesYesNoNo
France1958YesYesNoNoYes
Greece1952NoYesYesNoYes
Iceland1944YesNoYesNoYes
Ireland1937YesNoNoNoYes
Sweden1974YesYesYesNoNo
Switzerland1999YesYesYesYesYes
Hungary2011YesNoNoYesNo
Poland1997YesYesYesYesYes
Ukraine1996YesNoYesNoNo
Kenya2010YesYesYesYesYes
Thailand1997YesYesYesYesNo
Sri Lanka1972NoNoNoYesNo
Nepal2015YesYesYesYesNo
Bolivia2009YesYesYesYesYes
Colombia1991NoNoYesYesYes
Ecuador1998YesYesYesNoYes
Ecuador2008NoNoNoYesYes
Uruguay1952YesYesYesNoYes
Uruguay1967YesYesYesNoYes
Venezuela1999NoNoNoYesYes
Dominican Republic2010YesNoYesYesNo
Trinidad & Tobago1976YesNoNoNoNo
CountryYearLegal continuityInstitutional checksPlural constituent bodyPublic consultation/ submissions*Popular referendum**
Denmark1915YesYesYesNoYes
Denmark1953YesYesYesNoYes
Finland2000YesYesYesNoNo
France1958YesYesNoNoYes
Greece1952NoYesYesNoYes
Iceland1944YesNoYesNoYes
Ireland1937YesNoNoNoYes
Sweden1974YesYesYesNoNo
Switzerland1999YesYesYesYesYes
Hungary2011YesNoNoYesNo
Poland1997YesYesYesYesYes
Ukraine1996YesNoYesNoNo
Kenya2010YesYesYesYesYes
Thailand1997YesYesYesYesNo
Sri Lanka1972NoNoNoYesNo
Nepal2015YesYesYesYesNo
Bolivia2009YesYesYesYesYes
Colombia1991NoNoYesYesYes
Ecuador1998YesYesYesNoYes
Ecuador2008NoNoNoYesYes
Uruguay1952YesYesYesNoYes
Uruguay1967YesYesYesNoYes
Venezuela1999NoNoNoYesYes
Dominican Republic2010YesNoYesYesNo
Trinidad & Tobago1976YesNoNoNoNo

Open in new tab

Table 2.

Features of constitution-making in democratic regimes, 1900–2015.

CountryYearLegal continuityInstitutional checksPlural constituent bodyPublic consultation/ submissions*Popular referendum**
Denmark1915YesYesYesNoYes
Denmark1953YesYesYesNoYes
Finland2000YesYesYesNoNo
France1958YesYesNoNoYes
Greece1952NoYesYesNoYes
Iceland1944YesNoYesNoYes
Ireland1937YesNoNoNoYes
Sweden1974YesYesYesNoNo
Switzerland1999YesYesYesYesYes
Hungary2011YesNoNoYesNo
Poland1997YesYesYesYesYes
Ukraine1996YesNoYesNoNo
Kenya2010YesYesYesYesYes
Thailand1997YesYesYesYesNo
Sri Lanka1972NoNoNoYesNo
Nepal2015YesYesYesYesNo
Bolivia2009YesYesYesYesYes
Colombia1991NoNoYesYesYes
Ecuador1998YesYesYesNoYes
Ecuador2008NoNoNoYesYes
Uruguay1952YesYesYesNoYes
Uruguay1967YesYesYesNoYes
Venezuela1999NoNoNoYesYes
Dominican Republic2010YesNoYesYesNo
Trinidad & Tobago1976YesNoNoNoNo
CountryYearLegal continuityInstitutional checksPlural constituent bodyPublic consultation/ submissions*Popular referendum**
Denmark1915YesYesYesNoYes
Denmark1953YesYesYesNoYes
Finland2000YesYesYesNoNo
France1958YesYesNoNoYes
Greece1952NoYesYesNoYes
Iceland1944YesNoYesNoYes
Ireland1937YesNoNoNoYes
Sweden1974YesYesYesNoNo
Switzerland1999YesYesYesYesYes
Hungary2011YesNoNoYesNo
Poland1997YesYesYesYesYes
Ukraine1996YesNoYesNoNo
Kenya2010YesYesYesYesYes
Thailand1997YesYesYesYesNo
Sri Lanka1972NoNoNoYesNo
Nepal2015YesYesYesYesNo
Bolivia2009YesYesYesYesYes
Colombia1991NoNoYesYesYes
Ecuador1998YesYesYesNoYes
Ecuador2008NoNoNoYesYes
Uruguay1952YesYesYesNoYes
Uruguay1967YesYesYesNoYes
Venezuela1999NoNoNoYesYes
Dominican Republic2010YesNoYesYesNo
Trinidad & Tobago1976YesNoNoNoNo

Open in new tab

Legal continuity between the old and the new order makes the process more predictable and enables the courts to supervise compliance with the rules. It is achieved when the existing constitution (in its original or amended form) or a norm of constitutional status regulates the adoption of the new constitution, as was the case of the 1999 Swiss Constitution, the 2009 Bolivian Constitution, and the 1997 Polish Constitution.39 In total, twenty (80 percent) of the episodes of democratic constitution-making included in the database preserved legal continuity.

Institutional checks allude to the intervention of various independent collective institutions or instances in the activation and regulation of the process. This feature protects the interests of all participants in the process because it prevents any single state actor from proposing a change while also having direct influence in the drafting and approval of the new constitution. Checks of this type are lacking when a unicameral legislature proposes and enacts a new constitution without an intervening election (e.g., Hungary, 2011), or when the executive alone convenes a special constituent convention, regulates its election and procedures, and participates in the drafting of the new text (e.g., Venezuela, 1999). Institutional checks were observed in fifteen cases (60 percent) of democratic constitution-making included in the database.

The key feature that signals the presence of elite cooperation is a politically plural constitution-making body. Such a body exists when (i) two or more than two political parties or groups achieved representation in it through free and fair elections, and (ii) the collaboration of at least two of these parties or groups is necessary according to the decision rule established for the approval of the constitution. This indicator is correlated with legal continuity and institutional checks, suggesting that they are all tapping into the same phenomenon.40 Nevertheless, since only some of these features may be present in a single episode, one should always prefer the collaboration among a plurality of political forces in the constitution-making body as a measure of elite cooperation because it is a factual, not formal, indicator.41 It was observed in eighteen of the twenty-five constitutions (72 percent) included in the database.

Speaking generally, it may be said that the most common form of citizen participation in democratic constitution-making occurs through the election of representatives who will propose or decide on the content of revisions. Using this perspective, a recent important work on participatory constitution-making considers the special free and fair election of delegates to the constitution-making body as a key indicator of genuine popular participation at the drafting stage.42 This implies, however, a rather confusing concept of citizen input in constitutional change. The election of a constitution-making body is an indirect form of citizen involvement, via the representative process. To be precise, and to distinguish popular participation from elite representation, the former should be restricted to instances of direct citizen involvement that can take place through non-electoral mechanisms, such as public consultations and proposal submissions, or in the form of voting in referenda.43

Citizens can contribute to the formulation of reform proposals in various forms of public consultation before the formal initiation of the process, after its activation but before the writing of the initial draft, and after the initial draft is completed but before its final approval.44 Some of these channels involve forms of collective deliberation, such as public forums that take place before the process is activated in order to determine the content of the future reform agenda. Ordinary citizens and civil society groups can also have influence on the reform agenda, particularly after the process is initiated and during the writing and approval of a constitutional draft, by submitting reform proposals or comments.45 There are no cases in which all these channels of citizen involvement in the formulation of reform proposals are present. In some instances, however, citizens participate both before the writing of the initial draft and before its final approval, as in the making of the 1997 Polish Constitution, the 2015 Nepalese Constitution, the 2008 Ecuadorean Constitution, the 1999 Venezuelan Constitution, and the 1976 Constitution of Trinidad and Tobago.

Citizen participation through voting takes place in popular referenda on constitutional reform proposals. Constitutional referenda can be implemented before or after the drafting process is completed. In the first case, voting is used to decide a particular matter before a new constitution is actually drafted, such as choosing between a monarchical or republican form of government, as was the case with the referendum held in Greece in 1946. They may also be called to authorize replacing the constitution through the election of a constituent assembly when this procedure is not foreseen in the existing constitution, as in Colombia’s 1990 referendum. The most common form of referendum, of course, is a referendum implemented to ratify or reject the new text after it has been voted on in a representative body.

As we can see from Table 2, citizen participation through non-electoral channels has been observed in twelve cases (48 percent) and through voting in referenda in sixteen (64 percent). These mechanisms are, however, less frequently used together. Only seven of the twenty-five episodes of democratic constitution-making under analysis include both forms of citizen involvement. In fact, if we dichotomize any form of citizen participation in the formulation of proposals and voting in referenda during constitution-making, the correlation between them is not only insignificant but also negative. This suggests that the two mechanisms are often seen in practice as different and not necessarily complementary forms of citizen involvement in democratic constitution-making.

One final question is how elite cooperation and citizen participation relate to each other. As already argued, there is a conceptual difference between representation and decision-making at the elite level and participation at the citizen level. This difference also holds on empirical grounds. There is no significant association between the existence of a political plural constitution-making body and different forms of direct citizen involvement. Moreover, there are several cases, such as Ecuador, 2008; France, 1958; Hungary, 2011; Ireland, 1937; Sri Lanka, 1972; Trinidad and Tobago, 1976; and Venezuela, 1999, where there was centralized control by the executive or a single political party over the constitution-making body while some form of citizen consultation or voting was used during the writing or approval of the new constitution.

4. Quantitative analysis

As already noted, the number of constitutions adopted in the world within an established electoral democracy is not negligible and may increase in the future. Yet these are relatively few cases, a mere twenty-five if we use the criterion of observing a new constitution at least five years after the inauguration of democracy. We can expand the sample to forty-three observations if we include all constitutions in the world enacted at or after the third year of free and fair elections in a country. The observations included in this larger set include some constitution-making events that take place while the transition to democracy in a broad sense may not be over, such as the adoption of South Africa’s 1996 Constitution or Brazil’s 1988 Constitution. Nonetheless, all the constitutions in this analysis will still have been enacted after holding democratic elections.46

A universe of forty-three observations is still admittedly small for testing causal claims using statistical techniques. It is suitable, however, for a basic statistical analysis to observe whether empirical patterns exist and whether these patterns are consistent with existing hypotheses relating citizen participation and elite cooperation during the drafting stage of new constitutions to levels of democracy after their implementation. In order to check whether the patterns observed are an artifact of a restricted sample I will also provide an additional test using a larger database of ninety-four observations, which represent all the constitutions adopted in the world in a democratic year between 1900 and 2015.47

4.1. The liberal dimension of democracy

I have argued that only elite cooperation at the constitution-making stage is likely to improve the liberal dimension of democracy by making institutional constraints on executive power and citizen rights effective during the early years of life of the new constitution. The empirical analysis of this proposition should focus, then, on the actual implementation of liberal principles after a new constitution is enacted in the context of an existing electoral democracy.

In order to measure the implementation of liberal principles and the level of liberal democracy after the enactment of the new constitution I have used four indexes taken from the Varieties of Democracy (V-DEM) project.48 The first and most important is the liberal democracy index, which measures the quality of a regime that already qualifies as an electoral democracy by the effective limits placed on the exercise of executive power through the protection of civil liberties, checks and balances, rule of law, and an independent judiciary. Because this variable captures both changes in the electoral and liberal dimension of democracy, I relied on three additional indices that focus more specifically on legislative and judicial constraints on executive power and the liberal component of a political regime more generally. The first two indicate the extent to which in a political regime the national executive abides by the constitution and respects court rulings and how effective the legislature is in exercising oversight over the executive. The third averages these measures along with an index that captures respect for the rule of law and individual liberties. All these indices take values from 0 to 1, where 0 is the minimum and 1 the maximum score. I will also report in the text whether results change if we use other indices that also capture changes in democratization and executive constraints, such as the polity and executive constraints scores of the Polity IV database.49

I have proposed that the impact of elite cooperation during constitution-making on liberal democracy should be observed during the early years of life of the new constitution, when the level of popular support and institutional resources that organized political forces had at the drafting stage are likely to persist. For this reason, the dependent variables under analysis consist of the average value of each index during the first five years of life of the new constitution (i.e., from t + 1 to t + 5). This is a reasonable time span because electoral cycles across parliamentary and presidential systems oscillate, on average, between four and five years.50

4.2. Elite cooperation and citizen participation in constitution-making

Following the discussion in Section 3 of this article, I measure the impact of elite cooperation during constitutional change using a dummy variable that takes a value of 1 when the collaboration between two or more than two freely and fairly elected parties was necessary to pass the constitution according to the decision rule, and 0 otherwise.51 A value of 0 in this variable indicates that a dominant party was able pass the constitution unilaterally or that the executive or a single political force appointed the constitution-making body.

Also following the discussion in Section 3, popular participation in constitution-making is measured taking into account that citizens might be involved in constitutional change through non-electoral and electoral mechanisms. The first alternative is captured using a dummy variable that takes the value of 1 if ordinary citizens were involved in the formulation, discussion, or submission of reform proposals at any stage in the process. The voting alternative is measured through a dummy variable coded as 1 if citizens participated in popular referenda either at the beginning, in authorizing the process or deciding on a particular issue, or at the end, to ratify the new constitutional text.

The most important variables added to control for alternative explanations are the lagged value of the dependent variables and the age of democracy before the enactment of the new constitution. The success of democracy in the present is likely to be explained by its success in the past. For this reason, I will use the average of the indices of liberal democracy, legal constraints, judicial constraints, and liberal component during five years before the new constitution was adopted as a standard of comparison to determine whether actual practices changed as a result of constitutional change. The age of democracy is also a key control variable because it is very likely that the pre- and post-constitutional levels of democracy in any dimension are associated with the time passed since its inauguration. Additional variables included are the GDP per capita the year before the new constitution was adopted, the level of ethnic fractionalization in the country, the year the new constitution was enacted, and the region of the world to which the country experiencing the change belongs.52 Just as democratic success in a given country may very well be linked to a high level of economic development, internal social divisions may determine democratic failure. Finally, temporal and regional trends may also be responsible for the level of democracy observed in a given country.53 I will also report in the text whether results change as a consequence of adding other relevant control variables, such as population size and civil unrest, at the time of adopting a new constitution.

4.3. Results

Given the structure of the data and the nature of the dependent variables, I used an OLS (Ordinary Least Squares) regression with robust standard errors clustered by country to provide a basic test to the proposed hypotheses. Table 3 shows the results of this analysis.

Explanatory VariablesDependent variables
Liberal democracy (1)Legislative constraints (2)Judicial constraints (3)Liberal component (4)
123123123123
Elite Cooperation.120** (.052).108** (.047).089** (.038).127** (.062).121** (.057).125*** (.037).066 (.042).098* (.055).074** (.036).082* (.041).081* (.044).075** (.031)
Citizen Consultation.015 (.052).061 (.068).010 (.041)-.035 (.058)-.019 (.068).008 (.044).016 (.049).077 (.067).002 (.041)-.002 (.045).050 (.047).024 (.036)
Citizen Voting-103** (.044)-.060 (.051).012 (.031)-.084 (.053)-.029 (.054).031 (.035)-.060 (.045)-.041 (.054).005 (.035)-.063 (.040)-.016 (.041).013 (.026)
Lagged DV t-5.707*** (.128).499** (.193).380*** (.099).368* (.206).257 (.279). 180* (.107).862*** (.112).657*** (.133).559*** (.103).563*** (.156)180 (.196).283*** (.101)
Democratic Age (ln)-.030 (.023)-.044 (.028)-.027** (.012)-.016 (.030)-.037 (.037)-.024* (.014)-.035* (.018)-.039 (.024)-.037*** (.011)-.020 (.017)-.024 (.022)-.024** (.001)
GDPPC t-1(ln)______.084 (.050).117*** (.039)______.079 (.052).100*** (.034)______.063 (.046).103*** (.034)______.105** (.039).102*** (.031)
Enactment Year______-.002 (.001)-.000 (.001)______.002 (.001)-.001. (.001)______-.002 (.001)-.001. (.001)______-.002 * (.001)-.000. (.001)
Ethnic Fractionalization______-.152 (.191)-.232** (.110)______-.156 (.197)-.158 (.101)______-.166 (.163)-.209** (.097)______-.190 (.144)-.185** (.091)
Africa______.029 (.077).118* (.066)______.161 (.113).207*** (.075)______.120 (.093).144** (.064)______.131* (.073).137** (.056)
Asia______.090 (.149).142 (.091)______.136 (.144).164 (.111)______.108 (.141).197** (.096)______.122 (.132).129 (.087)
Europe______.025 (.075).027 (.049)______-.002 (.071).028 (.047)______.022 (.055).025 (.045)_____.051 (.043).040 (.041)
Constant.250** (.093)3.481 (2.322)-.216 (1.901).491*** (.130)3.926 (2.673).953 (1.719).209* (.108)3.940 (2.484).535 (1.840).387*** (.131)3.405* (1.767).781 (1.578)
Adjusted R20.440.550.590.180.250.350.640.630.630.330.490.54
N403573403573403573403573
Explanatory VariablesDependent variables
Liberal democracy (1)Legislative constraints (2)Judicial constraints (3)Liberal component (4)
123123123123
Elite Cooperation.120** (.052).108** (.047).089** (.038).127** (.062).121** (.057).125*** (.037).066 (.042).098* (.055).074** (.036).082* (.041).081* (.044).075** (.031)
Citizen Consultation.015 (.052).061 (.068).010 (.041)-.035 (.058)-.019 (.068).008 (.044).016 (.049).077 (.067).002 (.041)-.002 (.045).050 (.047).024 (.036)
Citizen Voting-103** (.044)-.060 (.051).012 (.031)-.084 (.053)-.029 (.054).031 (.035)-.060 (.045)-.041 (.054).005 (.035)-.063 (.040)-.016 (.041).013 (.026)
Lagged DV t-5.707*** (.128).499** (.193).380*** (.099).368* (.206).257 (.279). 180* (.107).862*** (.112).657*** (.133).559*** (.103).563*** (.156)180 (.196).283*** (.101)
Democratic Age (ln)-.030 (.023)-.044 (.028)-.027** (.012)-.016 (.030)-.037 (.037)-.024* (.014)-.035* (.018)-.039 (.024)-.037*** (.011)-.020 (.017)-.024 (.022)-.024** (.001)
GDPPC t-1(ln)______.084 (.050).117*** (.039)______.079 (.052).100*** (.034)______.063 (.046).103*** (.034)______.105** (.039).102*** (.031)
Enactment Year______-.002 (.001)-.000 (.001)______.002 (.001)-.001. (.001)______-.002 (.001)-.001. (.001)______-.002 * (.001)-.000. (.001)
Ethnic Fractionalization______-.152 (.191)-.232** (.110)______-.156 (.197)-.158 (.101)______-.166 (.163)-.209** (.097)______-.190 (.144)-.185** (.091)
Africa______.029 (.077).118* (.066)______.161 (.113).207*** (.075)______.120 (.093).144** (.064)______.131* (.073).137** (.056)
Asia______.090 (.149).142 (.091)______.136 (.144).164 (.111)______.108 (.141).197** (.096)______.122 (.132).129 (.087)
Europe______.025 (.075).027 (.049)______-.002 (.071).028 (.047)______.022 (.055).025 (.045)_____.051 (.043).040 (.041)
Constant.250** (.093)3.481 (2.322)-.216 (1.901).491*** (.130)3.926 (2.673).953 (1.719).209* (.108)3.940 (2.484).535 (1.840).387*** (.131)3.405* (1.767).781 (1.578)
Adjusted R20.440.550.590.180.250.350.640.630.630.330.490.54
N403573403573403573403573

Open in new tab

Table 3.

Democratic constitution-making, liberal democracy, and liberal principles.

Explanatory VariablesDependent variables
Liberal democracy (1)Legislative constraints (2)Judicial constraints (3)Liberal component (4)
123123123123
Elite Cooperation.120** (.052).108** (.047).089** (.038).127** (.062).121** (.057).125*** (.037).066 (.042).098* (.055).074** (.036).082* (.041).081* (.044).075** (.031)
Citizen Consultation.015 (.052).061 (.068).010 (.041)-.035 (.058)-.019 (.068).008 (.044).016 (.049).077 (.067).002 (.041)-.002 (.045).050 (.047).024 (.036)
Citizen Voting-103** (.044)-.060 (.051).012 (.031)-.084 (.053)-.029 (.054).031 (.035)-.060 (.045)-.041 (.054).005 (.035)-.063 (.040)-.016 (.041).013 (.026)
Lagged DV t-5.707*** (.128).499** (.193).380*** (.099).368* (.206).257 (.279). 180* (.107).862*** (.112).657*** (.133).559*** (.103).563*** (.156)180 (.196).283*** (.101)
Democratic Age (ln)-.030 (.023)-.044 (.028)-.027** (.012)-.016 (.030)-.037 (.037)-.024* (.014)-.035* (.018)-.039 (.024)-.037*** (.011)-.020 (.017)-.024 (.022)-.024** (.001)
GDPPC t-1(ln)______.084 (.050).117*** (.039)______.079 (.052).100*** (.034)______.063 (.046).103*** (.034)______.105** (.039).102*** (.031)
Enactment Year______-.002 (.001)-.000 (.001)______.002 (.001)-.001. (.001)______-.002 (.001)-.001. (.001)______-.002 * (.001)-.000. (.001)
Ethnic Fractionalization______-.152 (.191)-.232** (.110)______-.156 (.197)-.158 (.101)______-.166 (.163)-.209** (.097)______-.190 (.144)-.185** (.091)
Africa______.029 (.077).118* (.066)______.161 (.113).207*** (.075)______.120 (.093).144** (.064)______.131* (.073).137** (.056)
Asia______.090 (.149).142 (.091)______.136 (.144).164 (.111)______.108 (.141).197** (.096)______.122 (.132).129 (.087)
Europe______.025 (.075).027 (.049)______-.002 (.071).028 (.047)______.022 (.055).025 (.045)_____.051 (.043).040 (.041)
Constant.250** (.093)3.481 (2.322)-.216 (1.901).491*** (.130)3.926 (2.673).953 (1.719).209* (.108)3.940 (2.484).535 (1.840).387*** (.131)3.405* (1.767).781 (1.578)
Adjusted R20.440.550.590.180.250.350.640.630.630.330.490.54
N403573403573403573403573
Explanatory VariablesDependent variables
Liberal democracy (1)Legislative constraints (2)Judicial constraints (3)Liberal component (4)
123123123123
Elite Cooperation.120** (.052).108** (.047).089** (.038).127** (.062).121** (.057).125*** (.037).066 (.042).098* (.055).074** (.036).082* (.041).081* (.044).075** (.031)
Citizen Consultation.015 (.052).061 (.068).010 (.041)-.035 (.058)-.019 (.068).008 (.044).016 (.049).077 (.067).002 (.041)-.002 (.045).050 (.047).024 (.036)
Citizen Voting-103** (.044)-.060 (.051).012 (.031)-.084 (.053)-.029 (.054).031 (.035)-.060 (.045)-.041 (.054).005 (.035)-.063 (.040)-.016 (.041).013 (.026)
Lagged DV t-5.707*** (.128).499** (.193).380*** (.099).368* (.206).257 (.279). 180* (.107).862*** (.112).657*** (.133).559*** (.103).563*** (.156)180 (.196).283*** (.101)
Democratic Age (ln)-.030 (.023)-.044 (.028)-.027** (.012)-.016 (.030)-.037 (.037)-.024* (.014)-.035* (.018)-.039 (.024)-.037*** (.011)-.020 (.017)-.024 (.022)-.024** (.001)
GDPPC t-1(ln)______.084 (.050).117*** (.039)______.079 (.052).100*** (.034)______.063 (.046).103*** (.034)______.105** (.039).102*** (.031)
Enactment Year______-.002 (.001)-.000 (.001)______.002 (.001)-.001. (.001)______-.002 (.001)-.001. (.001)______-.002 * (.001)-.000. (.001)
Ethnic Fractionalization______-.152 (.191)-.232** (.110)______-.156 (.197)-.158 (.101)______-.166 (.163)-.209** (.097)______-.190 (.144)-.185** (.091)
Africa______.029 (.077).118* (.066)______.161 (.113).207*** (.075)______.120 (.093).144** (.064)______.131* (.073).137** (.056)
Asia______.090 (.149).142 (.091)______.136 (.144).164 (.111)______.108 (.141).197** (.096)______.122 (.132).129 (.087)
Europe______.025 (.075).027 (.049)______-.002 (.071).028 (.047)______.022 (.055).025 (.045)_____.051 (.043).040 (.041)
Constant.250** (.093)3.481 (2.322)-.216 (1.901).491*** (.130)3.926 (2.673).953 (1.719).209* (.108)3.940 (2.484).535 (1.840).387*** (.131)3.405* (1.767).781 (1.578)
Adjusted R20.440.550.590.180.250.350.640.630.630.330.490.54
N403573403573403573403573

Open in new tab

For each dependent variable, the regression shown in model 1 includes the main explanatory variables along with the lagged score of the dependent variable and the accumulated years of electoral democracy before the adoption of the new constitution. Model 2 incorporates the rest of the control variables, and model 3 applies the analysis to all episodes of constitutions adopted in a democratic year between 1900 and 2015. The results of all the tests are consistent with the proposition that elite cooperation at the constitution-making stage improves liberal democracy and facilitates the implementation of liberal principles during the early years of life of the new constitution.

In particular, when the agreement between two or more than two freely and fairly elected parties is required to pass the constitution there is a statistically and substantially significant improvement in the implementation of institutional constraints on executive power and citizen rights during the first five years after enactment of the constitution.54 By contrast, citizen participation, either in consultations and proposal submissions or voting, does not generally have a significant impact on liberal democracy or liberal principles after the new constitution is in force. The direction of the effect of voting in constitutional referenda is consistently negative. Yet only in one basic model (model 1 of liberal democracy) is this effect statistically significant and it disappears once we add the rest of the controls.55

Among the control variables, the lagged scores of liberal democracy, executive constraints, and liberal component generally show an expected positive impact. The level of GDP per capita the year before a constitutional replacement takes place also appears to be associated with an improvement in the implementation of liberal principles, suggesting that richer countries are more likely to make institutional constraints over the executive and citizen rights effective. There is also evidence that as a democracy becomes older it is less likely to experience further improvements in its liberal dimension and that a high level of ethnic fractionalization may be an obstacle to deepening democratization. Similar results (not shown) obtain if we add additional control variables, such as population size and a measure of civil unrest based on the addition of general strikes, riots, and antigovernment demonstrations at the time of change. Moreover, although they are not as precise as the V-DEM indices to capture the liberal dimension of democracy and specific constraints on executive power, the same effects are observed if we use the polity and the executive constraints scores of the Polity IV database as dependent variables.56

These findings do not suggest that public participation in democratic constitution-making is worthless. Comparing political conditions before and after the adoption of a new constitutional text, cases such as the adoption of Colombia’s 1991 Constitution show that cooperation among fractions of the political elite and citizen involvement in both consultation and voting during the process can be a fruitful combination for the strengthening of liberal democracy. Other cases, such as Brazil’s 1988 Constitution or South Africa’s 1996 Constitution, illustrate the benefits of mixing elite cooperation with citizen involvement in the formulation of reform proposals. Even pure voting in referenda may enhance liberal democracy if it follows an inclusive elite agreement, such as the adoption of Italy’s 1948 Constitution with the support of Christian Democrats, Communists, and Socialists. What the preceding analysis clearly indicates, however, is that no form of citizen participation is likely, by itself, to improve levels of democracy after the new constitution is in force.

5. Exploring causal mechanisms: Kenya, 2010 and Bolivia, 2009

The results of the above analysis may not be conclusive proof that a causal relationship between constitution-making and democratization exists. Its negative findings, however, strongly suggest that participatory constitution-making does not contribute to improving the liberal dimension of democracy after the adoption of a new constitution. A way to complement and make this analysis more persuasive is by examining particular cases to show how the dynamics of elite cooperation at the constitution-making stage are linked to democratic developments once a new constitution is in force.

For comparative analysis, I have selected the making of Kenya’s 2010 Constitution and Bolivia’s 2009 Constitution. These constitutions were made in the context of an electoral democracy that persisted after they were enacted. The countries where these episodes took place shared a number of key background conditions. Yet outcomes were different in terms of the levels of liberal democracy observed before and after constitutional replacement.57

Both Kenya and Bolivia are middle-income countries with a high level of ethnic fractionalization, a presidential form of government, and a relatively recent authoritarian past. In both cases the new constitution was adopted in the context of preexisting free and fair elections and after a political crisis that called the working of representative institutions into question. In both cases, also, this crisis made necessary a high degree of direct popular participation during constitution-making through channels of consultation and voting. Furthermore, in both Kenya and Bolivia the approval of the constitution required the collaboration of the incumbent and the main opposition party so that, according to the operationalization provided in this article, they qualify as cases of elite cooperation. In spite of these shared characteristics, the liberal components of democracy improved in Kenya but relatively declined in Bolivia during the early years of life of the new constitution in each country.

Given the similarities between these cases, it seems difficult to account for the divergent outcomes by looking at the features of the constitution-making process. I will argue, however, that a careful analysis of the political context in each case reveals that the reason for the contrasting results resides in the fact that the balance of political forces that sustained a cooperative constitutional agreement in Bolivia was more fragile and short-lived than in Kenya. In both episodes the agreement was to a large extent induced and constrained by exogenous factors. In neither case did the agreement reflect a genuine consensus on norms of democratic behavior. The difference was that whereas in Bolivia the incumbent party soon became a hegemonic electoral force able to implement and interpret the constitution at will, in Kenya the electoral and institutional power that the incumbent and main opposition party had at the time of passing the constitution remained relatively stable after the first round of elections. This made it possible for the leaders of the political opposition in Kenya, at least in the short term, to prevent the incumbent party from reneging on the constitutional agreement.

Kenya transitioned from a one-party state to a multiparty democracy in 2002, when the Kenya African National Union party, which had governed the country since independence, lost both the presidency and a parliamentary majority. Although the demand for a new constitution emerged in the 1990s, it was only with the beginning of the transition to democracy in 2001 that Parliament enacted a Review Act outlining the steps for the adoption of a new constitution. These included initial drafting by a small review commission, revisions to the draft by a national convention, and ratification by Parliament.58 Immediately after the 2002 election, however, a conflict between government and opposition over the structure of executive-legislative relations derailed the process. The government decided to draft a presidential constitution without support from the opposition and submitted it for popular approval in 2005. The proposal was defeated.

The outcome of the 2007 presidential and parliamentary elections formed the immediate background for the 2010 Constitution. The incumbent president, Mwai Kibaki, was reelected by a very narrow margin over his opponent, Raila Odinga, leader of the Orange Democratic Movement (ODM). At the same time, in the legislative election the president’s party, the Party of National Unity, won only 43 seats while the ODM became the largest party in Parliament, with 99 of 210 seats. Odinga claimed that presidential elections had been rigged and demanded that a new election be held. Opposition supporters led street protests and violence rapidly spread across the country, leading to the killing of more than 1000 people.59 Preexisting inter-ethnic conflict in part fueled this level of violence. Whereas the incumbent president belonged to the Kikuyu tribe, the largest (but not majoritarian) ethnic group, Odinga belonged to the Luo people, the fourth largest ethnic group in the country. On the brink of a civil war that no party had the capacity to win, with the help of international mediation and under strong international pressure, government and opposition signed a pact that led to a power-sharing coalition government that committed to enact a new constitution on a consensual basis.

Like the 2001 Review Act, the 2008 Constitution of Kenya Review Act established a replacement procedure that contained several features aimed at preventing any single institution or actor from manipulating the process. Legal continuity was secured by following the reform procedure established in the existing constitution, as amended between 1997 and 2008. According to this procedure, a committee of experts was responsible for collecting public views and submitting a draft to a parliamentary committee, which in turn would revise the draft before submitting it to approval by the National Assembly (Parliament). The latter could only pass the proposal if it received support by a qualified majority of 65 percent of the total membership. This design, along with party fragmentation in the approval body, contributed to a high degree of collaboration among the representatives of the main political groups.

The making of the new constitution in Kenya also involved a relatively high degree of citizen participation. Before the writing of the draft, the committee of experts held public consultations to identify contentious issues. Later, before submitting the proposal to the parliamentary committee, it released the draft for a month of public consultations that led to the writing of an amended version in response to public views.60 In a last stage, after involvement by the parliamentary committee and the final vote in Parliament, the text was submitted to a popular referendum, where the constitution was approved by 67 percent of the valid votes cast.

Although the new Kenyan Constitution maintained a presidential structure of government, it constrained the powers of the executive in several ways. In the first place, the president would be elected by majority runoff instead of plurality rule, thus potentially increasing the probability of party fragmentation and reducing the president’s ability to count on the support of a legislative majority.61 Interference in the electoral process by the president was limited by the proposal to create an Independent Electoral and Boundaries Commission. The government powers of the executive were also curtailed by a process of political decentralization that allowed for the popular election of governors. At the same time, both legislative and judicial controls over the government were strengthened and the number of citizen rights expanded.62 Finally, additional checks and balances were introduced by means of a Constitutional Court and a bicameral legislature that replaced the preexisting unicameral assembly.

In terms of the actual implementation of the new constitution, although Kenya continued to be affected by strong ethnic rivalries and violence, public corruption, and a weak party system, the country experienced a relative democratic improvement. According to the Polity IV index, Kenya increased its score from 7 in 2009 to 8 in 2011 and 2012 and to 9 between 2013 and 2015. As a result of the new constitution, the Independent Electoral and Boundaries Commission was founded in 2011 and became responsible for organizing the coming 2013 elections. Although the presidential race was relatively close and the loser challenged the outcome in the courts, the opposition abided by the results. Institutional constraints on the executive improved after 2010, particularly due to the increased judicial controls over the administration. Due to ethnic and religious tensions, as well as terrorist threats, civil rights have remained fragile, but their protection was stronger than in the previous period. For these reasons, and according to V-DEM indexes, the liberal democracy score improved from 0.31 in 2009 to 0.35 in 2015.

This outcome was made possible by the relative stability of the balance of power among the political forces that participated in the approval of the 2010 Constitution. The two main contenders in the 2013 presidential election, Uhuru Kenyatta and Raila Odinga, were members of Kibaki’s coalition government, participants in the constitution-making process, and strong supporters of the 2010 Constitution. Although Kenyatta won the presidential election, his supporting coalition, the Jubilee Alliance, did not win a majority in either the assembly or the senate.63 At the same time, the main opposition party, the ODM, remained a significant opposition force with institutional influence and capacity to mobilize its constituents in defense of the constitution.64 However, it seems dubious whether this balance will hold in the future. Violence has reemerged during the 2017 election and the incumbent president was reelected with an almost absolute majority in the legislature. The opposition protested against electoral irregularities and although it won a successful legal challenge in the Supreme Court, it refused to participate in the election rerun.

The constitution-making process in Bolivia between 2006 and 2009 had several features in common with the Kenyan case in terms of fostering the democratic legitimacy of the new constitution by means of citizen participation. The adoption of a new constitution in Bolivia derived from the extensive popular mobilizations that took place between 2000 and 2003, demanding changes in public policies and deep reforms to the exclusionary nature of existing representative institutions. In 2003, the incumbent president was forced to resign in the midst of widespread social protests against his government, which included a demand to call a constituent convention to replace the 1967 Constitution. As a response to these events, the provisional government and the parties represented in Congress organized a constitution-making process that was meant to involve citizens in a wide variety of ways.

The Constitution was amended in February 2004 to allow Congress to convene an independently elected constituent convention and regulate its internal procedures. Based on this reform, in 2006 Congress passed a law establishing the system to elect constituent assembly delegates; the decision-making process of the assembly, the relationship between the constituent assembly and the Congress; and a final ratification of the constitution by referendum. Along with the election of delegates to the convention, Bolivians were also called to decide on the autonomy of regions in a referendum that would be binding on the convention. After the convention was installed, it organized deliberative forums in the different regions of the country and both ordinary citizens and civic organizations were allowed to submit reform proposals to the different committees on various aspects of the new constitution.65

What sets the Bolivian case apart from the Kenyan case is the precarious cooperation achieved between government and opposition and the short-lived nature of the balance of power that made it possible. At the time of amending the Bolivian Constitution to regulate its replacement (2004) the main party advocating reform, the Movimiento al Socialismo (MAS), was a minority political group. The situation started to change in 2005, when Evo Morales, the MAS candidate and member of the Aymara’s ethnic indigenous group, won the presidential election with more than 50 percent of the vote.

Morales’s party reached an absolute majority in the chamber of deputies but fell short of a majority in the senate. In this context, the 2006 congressional law regulating the constituent convention was passed with the support of PODEMOS, the main opposition party at the time, and included important safeguards to preserve a consensual process, such as the requirement for a two-thirds majority to pass the constitution. Yet after winning a majority of seats in the convention, the president and his party attempted to violate the convening law and adopt the constitution by majority rule. After a series of violent civil confrontations, the government backtracked from its attempt to impose the constitution and its final text was negotiated with opposition forces in Congress. It is clear, however, that the agreement was forced by time constraints and the need to obtain congressional approval for the ratification referendum.66

The tug of war between government and opposition to control the process was reflected in the design of the new constitution. While citizen rights were significantly expanded, the powers of the executive were strengthened in various ways.67 The threshold of votes to elect the president was reduced from 50 to 40 percent and the consensual Bolivian system that required Congress to intervene if no candidate reached that threshold was eliminated.68 The president was allowed to compete for one consecutive reelection and his legislative powers were increased by the power to submit urgency bills to Congress for approval. At the same time, both the formal independence and powers of the judiciary were enhanced. After the promulgation of the new constitution, however, the legal institutional constraints over the executive failed to be implemented.

Attacks on the judiciary began in 2007, when the government initiated several impeachment processes against members of the Supreme Court in response to rulings that supposedly benefited members of the opposition. Similar actions were followed in May and August 2009, immediately after the new constitution was ratified in a popular referendum.69 The government also followed the strategy of inducing the exit of members of the Constitutional Court and the Judicial Council until both bodies were de facto unable to function.70 These actions paved the way for filling positions in these bodies with government allies after 2009. After being reelected in December 2009, the president also violated a key point in the constitutional agreement, namely, the inability to run as presidential candidate for more than two periods. In 2013 the president’s party passed a congressional law with the interpretation that the period initiated after the 2009 election was the president’s first and not second term, so that he could be reelected again in 2014.

Although Bolivia continued to be an electoral democracy, its liberal dimension declined after the new constitution was adopted. According to Polity IV, Bolivia had a score of 7 in 2014, above the minimum of 6 to qualify as a democracy. Six years earlier in 2008, however, Bolivia had a higher level of democracy with a score of 8. According to V-DEM, in turn, the index of liberal democracy in Bolivia went from a score of 0.46 in 2008 to a score of 0.44 in 2014. The main factor that explains a decline in the effective implementation of the institutional constraints on executive power created in 2009 lies in the rapid change of the initial balance of forces in favor of the incumbent government. In the 2009 election, the president was reelected and his party reached a qualified majority in both the lower and upper houses, sufficient to approve constitutional changes. The PODEMOS opposition alliance, one of the main actors during the constitution-making process, disappeared and was replaced by Plan de Progreso para Bolivia, which won less than 30 percent of the seats in both chambers of Congress. As the same imbalance has persisted since the incumbent president’s reelection in 2014, the prospects of liberal democracy in Bolivia have remained low.71

6. Conclusion

A long tradition in constitutional theory has emphasized the democratizing potential of participatory constitution-making. An alternative perspective, relatively recent in constitutional studies but with roots in a well-established research agenda on democratization, stresses the critical role of elite accommodation during constitution-writing for the inauguration and consolidation of democracy. I have argued that a focus on the actions of representative elites provides the most promising starting point to understand the link between constitution-making and democratization in a comparative perspective. In particular, I proposed that only cooperation among a plurality of elected political representatives at the constitution-making stage is likely to improve the liberal dimension of democracy after the enactment of the new constitution, at least in the short term. Analyzing the effects of citizen participation and elite cooperation during constitution-making in the context of existing electoral democracies, this article has shown quantitative and qualitative evidence that is consistent with this hypothesis.

The arguments and findings of this article do not contradict the widely cherished idea that citizen participation is an essential principle of democratic constitution-making, or that it may be politically convenient when constitutional change is a response to popular demands. Yet they provide reasons to be skeptical about the democratizing effects of citizen involvement in constitution-making, in isolation from the actions and decisions of political representatives at the elite level. Clearly, more work is needed to examine the factors that make citizen participation and elite cooperation in constitutional change more or less likely and test their separate and joint effects on democracy within a comprehensive theoretical framework containing insights from both comparative constitutional law and comparative political science. This article is a first step in that direction.

I would like to thank Zach Elkins, Lourdes Jiménez Brito, Saúl López Noriega, Mariano Sánchez-Talanquer, Georg Vanberg, and two anonymous reviewers for their comments on a previous version of this article.

© The Author(s) 2020. Oxford University Press and New York University School of Law. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

© The Author(s) 2020. Oxford University Press and New York University School of Law. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

You might also like

Latest Posts

Article information

Author: Catherine Tremblay

Last Updated: 10/12/2022

Views: 5934

Rating: 4.7 / 5 (67 voted)

Reviews: 82% of readers found this page helpful

Author information

Name: Catherine Tremblay

Birthday: 1999-09-23

Address: Suite 461 73643 Sherril Loaf, Dickinsonland, AZ 47941-2379

Phone: +2678139151039

Job: International Administration Supervisor

Hobby: Dowsing, Snowboarding, Rowing, Beekeeping, Calligraphy, Shooting, Air sports

Introduction: My name is Catherine Tremblay, I am a precious, perfect, tasty, enthusiastic, inexpensive, vast, kind person who loves writing and wants to share my knowledge and understanding with you.