II. The constituent power theory: Its core problems

There are several problems with the constituent power theory.Footnote ³⁰ This section outlines two prime reasons to reject the theory in favor of a new yardstick for normative legitimacy.

First, the constituent power theory is detached from the realities of constitution-making and cannot be applied on its own terms to the bulk of constitution-making instances. This is obviously undesirable for a theory that is the primary yardstick to measure normative constitutional legitimacy, and that is often invoked by leaders to defend their constitutions.

The constituent power theory proceeds on the assumption that a constitution is a collective expression by a society’s populace at the moment of drafting. Constitutions are only normatively legitimate when they successfully reflect such an understanding. Nonetheless, constitutions are hardly ever the product or expression of the people’s collective will. They are generally products of different variants of political competition at the moment of drafting and reflect elite preferences.Footnote ³¹ Versteeg’s large-N analysis shows how, in most countries around the world, public opinion on topics of key concern differed from the constitutional content.Footnote ³² Elster famously noted that the bulk of the actual constitution-making process occurs behind closed doors,Footnote ³³ where power is frequently wielded in partisan ways on behalf of ‘some’ rather than ‘all’ the people.Footnote ³⁴ This has been the case in almost all recent constitution-making incidents in the Arab world, such as in Egypt, Iraq, Libya and Tunisia. Brown describes how, in these cases, important political actors were excluded and threats of force were sometimes part of the constitution-making process.Footnote ³⁵ In the much-lauded case of Tunisia, threats of force formed an ‘implicit but quite clear backdrop’ for constitutional bargaining.Footnote ³⁶

The aforesaid dynamics are unfortunately present even in highly participatory constitution-making processes. Iceland’s failed constitution-making process is a case in point. Gylfason points out how the 1944 Constitution ‘was drawn up in haste with minimal adjustment of the 1874 constitution as part of Iceland’s declaration of independence from Nazi-occupied Denmark’.Footnote ³⁷ People blamed some of the structural flaws of the 1944 Constitution for Iceland’s 2008 financial crash.Footnote ³⁸ After the crash, protestors took to the street demanding constitutional reform, leading the existing parliamentary government to resign. The new government implemented a plan to draft a new constitution. Although the entire constitutional drafting process is beyond the scope of this article, it was highly participatory and included (among other participatory tools) crowdsourcing constitutional content via social media. However, Hudson’s empirical research reveals how only 10 percent of the total submissions from the public seemed to impact the draft constitution.Footnote ³⁹ Moreover, political elites eventually stalled the constitution-making process and prevented the aforesaid draft from coming into force because they did not want to change the status quo, which they believed was profiting them financially and politically.Footnote ⁴⁰

Iceland at least provides an example where public participation impacted a draft constitution (even though, despite a major constitutional moment, the public did not get the new constitution it wanted). In most cases, public views do not even make their way into the draft. Take, for example, South Africa’s 1991–95 constitution-making process, which has been lauded for its degree of public participation and innovative mechanisms adopted to conduct the same. Some 48 per cent of the population felt they had been part of the constitution-making process.Footnote ⁴¹ Nevertheless, this extensive public participation program had a negligible impact on the constitutional text.Footnote ⁴² Even the most progressive provisions of the new constitution were elite driven rather than a product of public participation.Footnote ⁴³

The usual story is that if the public’s views on matters of ‘actual importance or relevance’ run counter to elite preferences and wants, they are swept under the carpet.Footnote ⁴⁴ Elites will, however, sometimes window-dress the constitution with aspirational rights provisions to appease the public and/or international actors. This happened in Iraq,Footnote ⁴⁵ RwandaFootnote ⁴⁶ and AfghanistanFootnote ⁴⁷ in the early 2000s. The Afghanistan story highlights another reality of constitution-making. Afghanistan is an example of an imposed constitution: one forced upon the country by an occupying power, rather than expressing or channelling the will of the people. In Afghanistan, not only was the Taliban (understandably) excluded from the constitution-making process, but so were moderate Islamic groups and warlords who exercised most on-the-ground control.Footnote ⁴⁸ Western forces and their on-the-ground allies forced the latter two groups to accept terms through backdoor coercion.Footnote ⁴⁹ The final 2004 Afghanistan constitution was what Western forces and their on-the-ground allies wanted. Cases like Afghanistan are a recurring phenomenon in constitution-making.Footnote ⁵⁰ Though perhaps not completely dominating the constitution-making space, it is a reality that cannot be ignored and is here to stay.

Moreover, even where constitutions cannot be classified as imposed, transnational networks of constitutional advisors and international organizations now play a substantial role in constitution-making processes worldwide.Footnote ⁵¹ At times, these transnational networks play a more important role in constitution-making and contribute more to the content of a constitution than the local populace. A trend has also developed in constitution-making practice of significant borrowing from other countries, wholesale grafting and using universal templates.Footnote ⁵² These variants of constitution-making only reinforce the reality that constitutions are not a product of the people they govern or an expression of their collective will.

As noted earlier, the problem with the constituent power theory is not that the normative criteria it propounds are theoretically irrational. The problem is that they have a marginal connection to actual practice. The theory that several scholars uphold as the preeminent measure of constitutional legitimacy is barely applicable to real-world constitution-making on its own terms.

Considering the constituent power theory’s detachment from reality, there is another crucial reason to reevaluate its use as the predominant yardstick for the normative legitimacy of constitutions. This is its usefulness in legitimizing authoritarian ends. Seeing the track record of their democratic counterparts, would-be-autocrats realize that they do not have to do much to legitimize their authoritarian constitutions. They can take advantage of momentary bursts of popularity, mislead the populace to hold biased viewpoints and appeal to the ‘will of the people’ to overhaul constitutional orders in authoritarian ways.

For instance, in Venezuela in 1999, Hugo Chávez went outside the constitutional order to unilaterally redraft the existing constitution into an authoritarian one. A political outsider, Chávez came to power in the wake of an economic crisis that the two major political parties, embroiled in their own corruption scandals, could not handle. One of Chávez’s electoral promises was to draft a new constitution and establish a constitutional order that would allow for effective social and participative democracy. The existing constitution contained a revision clause that required negotiation with the legislature. Upon Chávez’s election, the aforementioned two major political parties still controlled the legislature. Hence, instead of negotiating with his political opponents, Chávez issued a unilateral decree calling for a referendum on convening a Constituent Assembly.Footnote ⁵³ Some groups appealed this decree to the Supreme Court, but the Supreme Court upheld it on the grounds that the constituent power theory gives people the right to remake their constitutions.Footnote ⁵⁴ The referendum passed decisively with over 82 per cent of the vote.

The electoral rules written by Chávez for the selection of Constituent Assembly members were designed to tilt the scales in favour of his loyalists – commonly known as the Chavistas. Consequently, the Chavistas, who were internally diverse and represented various segments of society, won about 93 per cent of Constituent Assembly seats (though only 60 per cent of the vote share). As a result, they had no need to negotiate with the handful of opposition members who made it into the Assembly. Later, the Assembly closed state institutions such as the legislature, in which the opposition still had a say, and transferred their powers to itself.Footnote ⁵⁵ Again, the Supreme Court acquiesced to these moves. It held that the Constituent Assembly, as the direct manifestation of the people’s constituent power, was a supra-institution capable of wielding such authority.Footnote ⁵⁶ Ironically, finding the Supreme Court a potential roadblock, the Constituent Assembly also shut down the Court soon afterwards.Footnote ⁵⁷ Venezuela’s new authoritarian constitution was drafted in consultation with trade unions, civil society organizations, professional associations and neighbourhood organizations.Footnote ⁵⁸ These organizations submitted 624 proposals, and more than 50 per cent of their recommendations were accepted for inclusion in the constitutionFootnote ⁵⁹ – significantly more than in cases of highly lauded participatory processes such as those in South Africa and Iceland.

Hungary’s 2012 authoritarian constitution-making process (carried out within the existing legal order) also shared some characteristics in its use of the constituent power theory to legitimize an authoritarian constitution.Footnote ⁶⁰ Likewise in Turkey, where Recep Tayyip Erdoğan ‘dismembered’ the constitution using three large-scale amendment packages in the 2010s, he appealed to constituent power rhetoric. He sought public approval via referendums even though he did not need to,Footnote ⁶¹ as he could have passed the amendment packages legislatively as well.Footnote ⁶² Erdoğan had explicitly stated that he would not pass the constitutional changes through the legislature if the referendums failed.Footnote ⁶³ However, as shown by Elkins and Hudson, referendums on such amendment packages or entire constitution drafts, especially when passed by would-be-autocrats, rarely fail.Footnote ⁶⁴

Many other jurisdictions, such as Russia (1993), Kazakhstan (1995), Belarus (1996), East Timor (2002), Bolivia (2008), Ecuador (2009) and Georgia (2017), have seen elites use the ‘constituent power theory’ and appeal to the ‘will of the people’ to unilaterally reorganize the state in authoritarian ways (be it through participatory constitution-making processes or otherwise).Footnote ⁶⁵ As Landau and Dixon argue, the ordinary usage of the theory can result in authoritarian outcomes.Footnote ⁶⁶

III. Can the constituent power theory be reimagined or improved?

The constituent power theory is highly entrenched as the baseline standard to evaluate the normative legitimacy of constitutions. Hence, it would be intellectually dishonest to move on to suggest an alternative without considering whether the constituent power theory can be adjusted to apply on more realistic terms to modern constitution-making and prevent its misuse. Aware of the realities highlighted in the previous section, scholars have already attempted to reimagine and/or expand the theory.

To address the problems of constitutions being largely the work of elites and reflecting their preferences, Colón-Ríos advocates for a baseline framework. He contends that high levels of public participation at various stages of constitution-making are essential for legitimacy,Footnote ⁶⁷ and all attempts should be made to receive citizen input through the most participatory methods possible.Footnote ⁶⁸ Any barrier to public participation during constitution-making should cast doubt on the constitution’s legitimacy.Footnote ⁶⁹ Likewise, scholars such as Bannon and Saunders have called for similar requirements for imposed constitution-making processesFootnote ⁷⁰ and those with international involvement.Footnote ⁷¹

The issue is whether these suggestions actually help to overcome the problem. It is highly questionable whether any type of public participation can meaningfully capture the will of the people.Footnote ⁷² This is especially true in societies with low public turnout, lack of infrastructure, high levels of illiteracy and security issues.Footnote ⁷³ These factors are particularly relevant since the bulk of modern-day constitution-making is taking place in societies with such characteristics.Footnote ⁷⁴ At the bare minimum, material inequalities among the general populace result in groups with more access to resources having a disproportionate impact.Footnote ⁷⁵ To add to all this, Bannon, Moelher, Muller and others note that public opinions on the constitution are typically less about their views on the constitutional content and more about the elites dominating the constitution-making process.Footnote ⁷⁶ Because ordinary citizens (including those in advanced countries) often find it difficult to evaluate the constitution and constitution-making process, they look to elites for cues.Footnote ⁷⁷

Consider the differences in the outcomes of the 2005 and 2010 Kenyan constitution-making processes and the accompanying approval referendums. The first process started when the government could no longer ignore demands for constitutional change by opposition parties, civil society and the international community.Footnote ⁷⁸ The resulting draft was a product of significant public participation (though limited elite involvement).Footnote ⁷⁹ This draft was vehemently opposed by an influential group of political elites, aligned with the incumbent president, who saw no need to alter the status quo. Footnote ⁸⁰ This constitution-making process ended in 2005 when a referendum rejected the draft constitution.Footnote ⁸¹ This result was arguably less a reflection of what the people wanted than of political elites’ campaigns for and against the new constitution.Footnote ⁸² The final voting was completely along political lines.Footnote ⁸³ In contrast, the second constitution-making process succeeded and was approved by a referendum in 2010. The second process succeeded because elites from different major political groups (both incumbent and opposition) were on board – unlike the first time – and urged the populace to approve the new constitution.Footnote ⁸⁴

However, even if this were not the case and there were a way to capture the genuine will of the people, it would have limited significance in practice, considering the realities of how political elites treat public input. As discussed earlier, notwithstanding the widespread participation in South Africa and Iceland, minimal input found its way into the draft. In Rwanda in 2003, only 7 per cent of public views were analysed, and in Afghanistan in 2004 and Iraq in 2005, almost none were.Footnote ⁸⁵ All these three countries carried out participatory processes facilitated by international organizations. Adding any requirement for a certain degree of citizen input to make its way to final constitutional texts could result in a scenario where very few constitutions would qualify as legitimate, and would still deem Venezuela’s comparatively authoritarian constitution more legitimate than South Africa’s comparatively democratic one. Such a requirement would also not bode well for the eventual enforcement of the constitution. After all, those in power after the constitution’s promulgation have to agree to enforce its terms and provisions.

Likewise, to tackle the paradox that arises where constituent power is used to legitimize authoritarian ends, scholars have argued that constituent power should not be completely unlimited. Scholars such as Roznai and Stacey assert that the constituent power theory should be limited by principles such as liberalism, the rule of law and constitutionalism.Footnote ⁸⁶ Comparably, Braver, and to some extent Partlett,Footnote ⁸⁷ argue that the constituent power theory should only be invoked after exhausting all existing legal channels of constitutional change or reform within a constitutional system.Footnote ⁸⁸ Braver’s suggestion was targeted explicitly to cases like Venezuela, where would-be autocrats such as Chavez ignored the total reform/revision clause of the original constitution by invoking the constituent power theory. Nevertheless, as noble as they might be, these solutions run their course beyond a point in solving the problem of legitimizing authoritarian constitutions (while still not addressing the first problem of the constituent power theory).

When it comes to solutions such as those suggested by Roznai and Stacey, most modern authoritarian constitutions do prima facie contain principles associated with liberalism, the rule of law and constitutionalism (or the like).Footnote ⁸⁹ The authoritarian constitutions discussed above almost always created ostensibly independent courts with the power of judicial review, contained core civil and political rights guarantees and emphasized a commitment to the rule of law. It is only in the margins and intricate workings of authoritarian constitutions that their problematic nature is noticeable. For example, in Hungary, the new constitution eliminated the actio popularis review.Footnote ⁹⁰ Instead, it provided a constitutional complaint mechanism similar to Germany’s, where individuals may only challenge laws that affect them personally.Footnote ⁹¹ In Turkey, the amendment packages transformed the parliamentary system into a presidential system.Footnote ⁹² In Venezuela, the new constitution removed the single five-year term limit for presidents and allowed them to stay in office for up to two consecutive terms of six years each.Footnote ⁹³ All these modifications appear harmless when viewed individually and arguably have functioned normally in other national contexts. However, in each of these countries’ contexts, they were highly problematic (especially when seen together with other comparable changes).Footnote ⁹⁴ This is precisely how the modern-day legal autocrat works: using constitutional innovations to their own advantage.

Solutions such as Braver’s, which require exhausting existing legal channels of constitutional change or reform before invoking the constituent power theory, similarly fail to prevent abuses of the constituent power theory. In cases such as Hungary and Turkey, the respective would-be autocrats operated within the existing legal orders using options provided by the legal orders themselves.Footnote ⁹⁵ Although it must be mentioned, and as touched upon in Part V, in Hungary the aforesaid involved a degree of creative legal maneouvring.Footnote ⁹⁶ Further, in response to Braver, Landau points out how an existing legal system might have ‘bad’ rules or rules that can be very easily wielded in abusive ways.Footnote ⁹⁷ Today’s autocrats have mastered using the law to their advantage.Footnote ⁹⁸ Thus, Braver’s reimagination of the constituent power theory might not be sufficient to prevent legitimizing authoritarian constitutions.

Additionally, as Verdugo argues, a major concern with Braver, Roznai and Stacey’s solutions is that these solutions ultimately ‘[empty] the constituent power idea from its core elements’.Footnote ⁹⁹ After all, the virtue of the constituent power theory is that it validates the people’s sovereign will to decide how they want to be governed. Imposing additional restrictions on how and what the people can decide detracts from the theory’s core aspects. This is precisely why this article suggests an alternative standard rather than trying to expand (or reimagine) the constituent power theory. In conclusion, it is hard to see a version of the constituent power theory that can fully overcome both hurdles discussed above. Perhaps the only remaining option is to look beyond the constituent power theory.

IV. The theory of equitable elite bargaining

Given the main problems of the constituent power theory, an alternative theory should succeed in doing three things if it is to gain mainstream acceptance. First, it should recognize that constitutions are a product of elite contestation, and constitution-making processes depend heavily on elites and their preferences.Footnote ¹⁰⁰ Second, it should not be a fair-weather theory, but must be capable of being applied on equal terms to all instances of constitution-making, including difficult cases such as imposed constitution-making and internationalized constitution-making. Third, it should not be susceptible to being used to legitimize authoritarian (or any other comparable suboptimal) ends.

As an alternative framework to evaluate normative constitutional legitimacy, I propose what I term ‘the theory of equitable elite bargaining’. This theory is built on the premise that society contains certain individuals and groups who exercise disproportionate political power and influence compared with ordinary citizens, and who drive the process of constitution-making.Footnote ¹⁰¹ It accepts this reality and attempts to channel it in the best way possible to avoid legitimizing not only authoritarian constitution-making but also other suboptimal outcomes such as partisan elite-driven constitution-making. According to this theory, a constitution is normatively legitimate if it is the product of an equitable bargain between elites from most major political groups in society at the moment of constitution-making.

An alternative theory to measure normative legitimacy could merely have reverse-engineered the reality of constitution-making processes that resulted in democratic outcomes. A common thread in most constitution-making processes that resulted in democratic outcomes was an equitable (or even equal) bargain between the different sources of political power at the moment of constitution-making.Footnote ¹⁰² Thus, such an alternative theory could have laid down that a legitimate constitution is one that is an equitable (or even equal) bargain between the different sources of political power at the moment of constitution-making. As Dixon and Ginsburg’s ‘insurance model’ informs us, such a hypothetical theory would, in most cases, result in acceptable outcomes.Footnote ¹⁰³ However, there is a slight problem with this hypothetical theory. Although such a hypothetical theory would be reconcilable with the realities of constitution-making, it could nevertheless be abused by would-be-autocrats who take advantage of temporary surges in power at the expense of weak and divided political parties. This is precisely what happened in many of the authoritarian constitution-making incidents discussed in this article.

Further, such a hypothetical theory would not prevent power from being wielded in highly partisan ways, as happened in constitution-making episodes in the Arab world. It would also not prevent the problem of ‘disgruntled losers’, which is covered in more detail later on.Footnote ¹⁰⁴ More importantly, it could legitimize constitutions that lock out vital interests in society. This would be particularly relevant in cases where certain groups are not formally represented in politics at a particular time. A historical example is the American constitution-making process, which excluded Native Indians and African Americans. The theory of equitable elite bargaining goes further than requiring the balance of power in a society to be respected. Instead, it requires an equitable bargain between most major political groups. Adding this extra requirement can help ‘reduce’ the possibility of suboptimal outcomes while still recognizing the realities of how constitutions are made. As the remainder of this article will show, the theory of equitable elite bargaining can work in most real-life circumstances of constitution-making and simultaneously reduce possibilities for legitimizing anti-democratic outcomes.

To clarify how this theory could be applied in practice, five key concepts should be clarified: equity, inclusivity, duration of application, plenary power of elites, and civic society and public participation.

First, equity: in this theory, ‘equitable’ should be understood in line with its simple dictionary meaning – fair and impartial. The theory of equitable elite bargaining does not require elites from every major political group to have an equal voice – it simply requires equity. While this term is highly subjective, elites from different groups should be involved so they have a meaningful voice in the constitution-making process. How this would work out in practice would be highly context dependent and is influenced by other clarifications provided below. Ultimately (subject to the exceptions covered below), equity would require that most major political groups in a society have a chance at a realistic say in deciding both the process of drafting a constitution and the eventual constitution and its terms. The hypotheticals in the next section can help to illustrate what equitable might look like in varying situations.

Second, inclusivity: ‘major’ should be understood broadly and on the side of inclusivity rather than exclusivity. Political groups should not be understood as restricted to organized political parties, but can also include other groups – such as a revolutionary front – that, in a given moment, exercise political influence. It could also include groups that do not always have formal political representation, such as Indigenous groups. Political groups should also be read to include groups that, in a given time and space, are comparatively weaker, discredited, unpopular, exercising de facto control and the like. At the same time, the theory of equitable elite bargaining should not be understood to require elites from every major political group in a country to be involved in the equitable bargain for the process to be legitimate. There can surely be situations where elites from a particular major group are excluded for prudent reasons. This could be the case when, despite all good-faith attempts to involve elites from a specific group, they refuse to join or refuse to bargain equitably. It could also be the case when elites from almost every other major group in the country genuinely and in good faith believe that a particular group should be excluded for the sake of the future stability of the country (for example, that Nazis should not have been involved in drafting the German Basic Law). Such scenarios should be exceptional and have extremely compelling reasons. This should not be normalized as a standard practice. Further, it should not be due to a political maneouvre by certain political groups to lock out others. When in doubt, it is better to err on the side of inclusivity than exclusivity. While exceptions can be made, in cases where elites from a major political group (or groups) are excluded for whatever reason, the degree of normative constitutional legitimacy would be impacted.

Third, duration of application of the theory: equitable bargaining should take place at all stages of constitution-making: triggering the drafting process, drafting the new constitution and ratification/approval of the constitution. The entire process of constitution-making must be designed to facilitate the equitable bargain. An exception to the above rule is allowed when, during drafting negotiations, elites from a major political group (or groups) agree voluntarily, explicitly and without the use of force to walk away, take part in a non-equitable way (or one that is unfavourable to them) or make some other mutually negotiated arrangement. Such situations can even arise due to a quid pro quo arrangement, as long as the arrangement is voluntary, explicit and without the use of force.

Fourth, plenary power: elites from major political groups can trigger the constitution-making process at any stage during the operation of a polity, if such a decision is a product of an equitable bargain between at least most major political groups. There is no need for a constitutional moment. Furthermore, elites may set the rules of the entire process without restrictions, as long as they result from an equitable bargain. The new constitution can be drafted either within or outside the existing legal order. In the latter case, during the initial stages of the process and based on an equitable bargain, elites can decide the status of existing institutions. The theory of equitable elite bargaining puts no limits on these decisions beyond the requirement that every decision must be the product of an equitable bargain between elites from most major political groups. If, at any stage, the original rules of the process need to be modified, this can only happen as an equitable bargain between elites from the original major political groups.

Fifth, civil society and public participation: the theory of equitable elite bargaining does not prohibit or dissuade participation in constitution-making (of any and every kind) by civil society and/or the public. In most cases, these would be crucial for a constitution’s sociological legitimacy; they can also accord other benefits. Building on the previous clarification, the theory of equitable elite bargaining does not put limitations on the plenary power of elites; elites from most major groups may, therefore, equitably decide whether, for what reasons, to what extent, in what manner and at what stages they want to involve civil society and the general populace in constitution-making. However, involving or not involving these actors does not impact the normative legitimacy of the constitution. As detailed in Part VII, such an approach ensures that civil society and/or public participation, if and when conducted, are meaningful, not a box-ticking exercise and not a means to legitimize undemocratic outcomes.

V. The theory of equitable elite bargaining in practice

How would the theory of equitable elite bargaining operate in practice? Take a hypothetical parliamentary democracy, Atlantis, which is deeply divided between three major ethnic groups. Two of these ethnic groups are represented in the electoral arena by political parties 1 and 2, respectively. Both parties have equal electoral sway in the country, which for all practical purposes is a two-party system. The third ethnic group (an Indigenous group) is not formally represented in politics by a specific political party. However, its Indigenous leadership, 3, has a significant voice in the country and, from time to time, conveys support for leaders from both parties 1 and 2. Most of Atlantis’s political actors and population desire a new constitution because their previous constitution was drafted under colonial rule. For a constitution-making process to be legitimate, the entire process should be a bargain between elites from 1, 2 and 3. Elites from 1, 2 and 3 can equitably decide on the constitution-making process and determine the content of the constitution through this process. Elites from 1 and 2 cannot proceed with a constitution-making process that excludes elites from 3 at any stage. Nor can 1 and 3, or 2 and 3 do the same. Be it via elections, elections coupled with reservations, or elite agreements, elites from 1, 2 and 3 must have an equitable say. However, a political group or its elites can be excluded entirely from a constitution-making process when they refuse to participate or equitably bargain despite all good-faith attempts to involve them. The only other scenario that can see non-equitable participation is when, after the initial equitable negotiation to agree to the terms of the constitution-making process, one (or more) of the groups agrees voluntarily to participate in the remainder of the process in a non-equitable manner.

This leads us to the question of exceptions permitted by the theory. How exactly do these exceptions work? Assume that in the previous hypothetical situation, 1, 2 and 3 mutually agree to draft a new constitution for Atlantis via a drafting body wherein any decision must be approved by at least three-quarters of the members present and voting. The bulk of the members of this drafting body will be elected via elections. However, 3 can voluntarily agree not to compete in this election if 25 percent of the seats in this drafting body are reserved for its nominees selected via an internal process. The theory of equitable elite bargaining allows for this, as this is a voluntary decision by 3 in exchange for a guaranteed 25 per cent voice in the process – sufficient to exercise a veto. In contrast, the theory of equitable elite bargaining would not allow elites from 1 and 2 to lock out elites from 3 and, despite their objections, settle on electoral rules that would result in a minuscule percent of 3’s preferred members in the drafting body. This would not be an equitable outcome for 3. Likewise, elites from 1 and 2 cannot, at a later stage, agree to change the drafting body’s voting rules to require only three-fifths of the members present and voting without equitably bargaining with elites from 3. Any change in the originally accepted rules must happen via an equitable bargain with elites from 3.

Alternatively, elites from 1, 2 and 3 could mutually agree to have the entire constitution drafted by a technical committee of domestic and foreign experts and then ratified via a public referendum. They could also add another requirement that the technical committee or a similar small body engages in widespread public consultation before any draft is put up for a referendum. Such outcomes diminish the elites from all three groups in the process, but are mutually and voluntarily agreed upon, and hence justified. Likewise, the theory of equitable bargaining would be compatible with a situation where, in elections to the drafting body according to rules equitably set by elites from all three groups, one or more groups receive very few seats (similar outcomes, under a different set of facts, were recently seen in Chile’s recent failed constitution-making process). This is because they voluntarily agreed to these rules as part of an equitable bargain. The onus is on elites to defend their interests in the equitable bargaining processes, and not agree to terms that can lead to their exclusion.

Let us consider the application of this theory in another situation. Atlantis has a presidential system but does not have deep societal divisions. However, it suffers from other problems. The existing political order – consisting of two parties, 1 and 2, between whom power has been shared for most of modern history – is undergoing an institutional crisis, as both parties are embroiled in corruption scandals. They have also been unable to address the country’s economic problems. Large sections of society demand a change. Against this backdrop, a leftist political outsider, Hercules, is elected president. Hercules’ chief electoral promise is to draft a new constitution to reset the political order and establish a social democracy. Hercules can draft a new constitution, but must involve elites from 1 and 2. They cannot avoid bargaining with elites from 1 and 2, no matter how participatory the constitution-making process or how egalitarian and human rights-protective their draft constitution is. However democratic the rules of the process – whether relating to a referendum to approve the creation of a constitution-drafting body, election of the drafting body or ratification of the final text – if they result in a scenario that blocks elites from 1 and 2 from the process, the process should not be considered legitimate. There are several ways that 1 and 2 can be involved in the process, which would depend on the realities on the ground in Atlantis. In some cases, this could be as simple as setting rules to elect members of the drafting body that ensure that elites from 1 and 2 are not excluded. In other cases, it might go beyond electoral rules and even extend to the rules regarding decision-making in the drafting body. For example, this could mean a special type of decision-making that requires the final constitution to be approved by at least X per cent of members from 1 and 2 (or, conversely, members who are not aligned with Hercules’s political group). In more complicated cases, it might require Hercules to enter into a political agreement with elites from 1 and 2 outside the formal legal order that lays down the rules of the constitution-making process in a way acceptable to 1 and 2 (even if they accept begrudgingly).

Some might consider this unfair, especially if Hercules is reforming the system in a way that aligns with their own politics. They might defend the constituent power theory, which might allow Hercules to draft a new constitution. But what happened in Atlantis is similar to how Chávez captured Venezuela and shares some commonalities with Viktor Orbán’s capture of Hungary. According to the theory of equitable elite bargaining, Chávez would have needed to negotiate with Venezuela’s opposition-controlled legislature – something that was also required by the revision/total-reform clause in the existing constitution.Footnote ¹⁰⁵ In Hungary, the new constitution was drafted within the existing legislature and passed as an ordinary piece of legislation.Footnote ¹⁰⁶ This allowed Orbán’s Fidesz party, which had 68 per cent control of the legislature, to block the opposition completely. The theory of equitable elite bargaining would demand that the opposition, which still had 32 per cent of seats in the legislature, be given some form of voice or veto power in the constitution-making process. Curiously, the Hungarian Constitution originally contained a provision that required a four-fifths majority of the legislature to decide the rules to draft any new constitution – a provision that Orbán and his Fidesz party had changed via a simple constitutional amendment.Footnote ¹⁰⁷ The theory of equitable elite bargaining does not mandate staying within the existing order to draft the constitution. Still, in the cases of Venezuela and Hungary, the rules prescribed by the existing constitution could have promoted more equitable elite bargaining – at least initially. Per the theory, these rules would play out the same way if, instead of Chávez or Orbán, there were more democratic actors in their place. Having such baseline rules in place can prevent any actor – good or bad – from legitimizing their goals unilaterally.

The application of this theory can undoubtedly be more complicated in tougher cases. Suppose Atlantis is governed by political parties 1 and 2, who do not adequately represent the citizens, and the country then breaks out in protests. Could the protestors simply take control of the political institutions by force and create their new constitutional order? This is a factual pattern seen in revolutionary constitution-making. The constituent power theory might say yes. However, in accordance with the theory of equitable elite bargaining, the protest leaders would need to negotiate equitably with 1 and 2 to create their new constitutional order. The same would apply if, instead of 1 and 2, Atlantis were ruled by an authoritarian government, 3. In fact, scholars have argued that bargaining – even with authoritarian groups – improves outcomes, and that failure to do so may result in instability or even a return to a form of authoritarianism.Footnote ¹⁰⁸

Let us further complicate this hypothetical situation. It is known that 3 is a genocidal authoritarian regime that promotes terrorism. The regime poses a threat to the people of Atlantis and other countries. Atlantis is invaded by a coalition of countries and international organizations named 4, which overthrows 3. The coalition intends to implement a new constitution for Atlantis with the aid of its allies on the ground, groups 1 and 2. In an ideal world, any constitution in Atlantis should be entirely homegrown and involve only domestic actors. This process should certainly not include 4. However, the realities of politics suggest that this seldom happens. Sometimes groups like 4 even have a major role to play in ensuring short-term stability in a country. Recognizing this reality, the theory of equitable elite bargaining accounts for situations such as those described herein. Thus, the theory of equitable elite bargaining is not incompatible with 4’s involvement in a constitution-making process (though it does allow 4 to voluntarily not be a part of the process and walk away). Nonetheless, 4 cannot unilaterally impose a constitution on Atlantis and needs to involve most other major political groups in Atlantis (beyond merely 1 and 2). At the same time, the theory of elite democratic bargaining would certainly not go so far as to advocate that 1, 2 and 4 must negotiate with 3, no matter what. However, they should certainly negotiate with a diverse set of Atlantian elites, whose involvement is ultimately vital to running the country.

In a case like Afghanistan, this would have required that Western forces and their on-the-ground allies not exclude Islamists and warlords who exercised significant political and social influence. This might have resulted in a more stable Afghanistan, lessening the chances of the Taliban returning to power. Even in Afghanistan, including more moderate factions of the Taliban would have been beneficial. On the other hand, being more in line with the theory of equitable elite bargaining might be one reason why the German Basic Law has endured. The Basic Law was drafted under foreign occupation but involved Germans representing diverse territorial, social and political interests.Footnote ¹⁰⁹ The drafting process excluded Nazis, but succeeded in bringing together most other important elite actors, who were vital to Germany’s future operation.Footnote ¹¹⁰ This is important, and is what separates the constitution-making incident in Germany from Afghanistan.

VI. Equitably elite bargained constitutions: Why are they normatively legitimate?

Until now, this article only provided a descriptive account of its suggestive theory. A pertinent question remains to be answered: why is a constitution drafted in accordance with the theory of equitable elite bargaining normatively legitimate? And what intrinsic standards of justifiability accords to it its normative worth?

At a primary level, this section will argue that consequentialist reasons justify a constitution drafted in accordance with the theory of equitable elite bargaining as normatively legitimate. In addition, at a secondary level, constitutions drafted in accordance with the theory of equitable elite bargaining can also gain additional normative legitimacy for representation-based reasons. Due to the limited scope of this article, this section will refrain from defending the status of consequentialist and representation-based justifications for normative legitimacy. Both approaches have been passionately criticized and championed; however, it is safe to assume that both these approaches are mainstream ones when it comes to according normative legitimacy to facets of political life.Footnote ¹¹¹ I proceed with that assumption. This is a shortcoming of this section’s analysis – which might need more detailed attention on another occasion. Nevertheless, I hope that this section, coupled with the next section wherein I deal with potential criticisms and concerns (and invariably justify certain inclusions and omissions in this theory), outlines a degree of theoretical backing to consider the possibility of adopting the theory of equitable elite bargaining as a new standard for measuring normative constitutional legitimacy.

At a primary level, consequentialist reasons justify a constitutional drafted in accordance with this article’s theory as normatively legitimate. Scholars from as far back as Thomasius in the late seventeenth century have argued that consequentialist considerations can be valid justifications for a constitution’s normative legitimacy.Footnote ¹¹² As stated by the likes of Raz, Wellman and Arneson, this essentially implies that a legitimate constitution can be conducive to positive societal outcomes.Footnote ¹¹³ In present times, scholars such as Harel, Shinar and Verdugo have increasingly seen the merit of such arguments, especially considering the realities of different constitution-making situations.Footnote ¹¹⁴ Constitutions in compliance with the theory of equitable bargaining can be defended as legitimate because, more often than not, they result in two major beneficial societal consequences. These are: (1) the effective enforcement of the constitution;Footnote ¹¹⁵ and (2) institutional rules and arrangements that advance democracy.Footnote ¹¹⁶ Few would argue that these two characteristics are unimportant to constitutions (and modern-day societies).Footnote ¹¹⁷

Constitutions based on equitable elite bargaining are usually enforced effectively because those who consented to the constitution and have a vested interest in its success remain major political players in the immediate years after the constitution’s promulgation.Footnote ¹¹⁸ This also eliminates the problem of disgruntled losers who would be determined to regain power, including using force.Footnote ¹¹⁹ Further, on obtaining power, such disgruntled losers would seek to completely remake a constitution in ways that might result in highly unwanted outcomes. Both of these advantages can be seen in South Africa. As discussed earlier, very little input from the public made it into the constitution. What explains the initial compliance with the South African Constitution and the absence of disgruntled losers is the effective bargain carried out between the elites of the African National Congress and the National Party.Footnote ¹²⁰ In contrast, Afghanistan is an example of where, because of the lack of an equitable elite bargain, the 2004 constitution was not effectively enforced and the country was faced with the problem of disgruntled losers in the form of the Taliban.Footnote ¹²¹

Moreover, constitutions drafted using this theory as a benchmark result in the creation of institutional rules and arrangements that help to advance democracy. Negretto and Sánchez-Talanquer show how constitutions produced through negotiated bargains between diverse political elites result in the creation of limits on state action and provide opposition parties and citizens with the means to make these limits effective.Footnote ¹²² Likewise, Bejarano and Segura demonstrate that when a variety of political forces present in society participate in negotiated bargaining during constitution-making, the result is a balanced constitutional text that disperses power, and thereby deepens democracy.Footnote ¹²³ This is because, in most cases, power distribution during constitution-making means that no single actor/group has absolute sway, typically resulting in the adoption of institutions and rules that protect against the arbitrary abuse of power by a single actor/group.Footnote ¹²⁴ In fact, Eisenstadt and Maboudi demonstrate how the absence of group inclusion during the constitution-making is almost always ‘doomed to failure’ from a democratic standpoint.Footnote ¹²⁵ Other scholars have reached similar findings,Footnote ¹²⁶ including in relation to individual constitutional elements such as constitutional courts, accountability institutions and human rights.Footnote ¹²⁷

Colombia offers a case in point. In 1991, elites representing different major political groups rallied around the perennial problem of drug wars and violence. They established a new constitutional system based on a multi-group equitable bargain.Footnote ¹²⁸ Compared to the previous constitution, which was highly exclusionary and centralized, the new constitution expanded channels for political participation and representation and created a balance among the different branches of government.Footnote ¹²⁹ It also devolved power to regional and local government, thereby adding extra checks on the power of the central government.Footnote ¹³⁰ This system was largely self-enforcing in the following years and facilitated the rotation of powers. The institutions and rules put in place ensured that elites and their successors did not renege on their commitments.

Similar outcomes can be seen in most of the cases discussed in this article that saw a degree of equitable elite bargains, such as Germany, South Africa and Kenya. In all these instances, the new constitutions that came into force improved levels of democracy. This would even be the case in a situation where a single political group dominates a society. As opposed to a situation wherein such a dominant group unilaterally drafts a constitution without involving any other political group, the requirement of equitably bargaining with other groups (despite them being comparatively much weaker) can help to enhance the democratic features of the ensuing constitution (even if not to the degree for which one would hope). At the same time, constitution-making processes characterized by a lack of bargaining between major political actors tend to produce institutional arrangements that concentrate power in the executive, at the expense of institutions that limit the use of power.Footnote ¹³¹ This results in conditions that are not conducive to the rotation of powers. In the previously mentioned cases of Afghanistan, Egypt, Hungary, Iraq, Libya, Tunisia, Turkey and Venezuela, a lack of equitable elite bargaining led to these exact problems.

Though consequentialist reasons might be sufficient to accord a constitution drafted in accordance with the theory of equitable elite bargaining legitimacy, in most instances other reasons can also strengthen its claim to normative legitimacy. Specifically, a secondary reason for constitutions drafted in accordance with the theory of equitable elite bargaining to be considered legitimate is that, to a fair degree, they are representative of broader societal interests (or, at a bare minimum, diverse societal interests). In modern-day politics, for better or worse, people’s interests are largely represented by political elites and groups.Footnote ¹³² Therefore, a constitution drafted as a consequence of an equitable bargain between elites from most major political groups would be representative ‘to the extent’ that such groups are illustrative of broader societal voices and values.Footnote ¹³³ This is so while ensuring that constitutions represent a plurality of ‘societal interests’ rather than only some interests – which was a faultline in many practical applications of the constituent power theory. Although not to the degree and manner with which ardent proponents of representativeness might be content (because elites might look out for their own interests rather than those they represent), given the realities of constitution-making, equitable elite bargaining is one of the more effective mechanisms to ensure broad representativeness without having other net negatives for normative constitutional legitimacy.

But why is this the case? When it comes to ensuring broad societal representativeness, we are in a Catch-22 situation. If, to ensure representativeness, we necessitate compulsory public deliberation and consultation, we are back to square one, wherein elites decide what goes into the constitution (and they can choose to ignore the outcomes of these processes). In such a case, elites can use compulsory participation of this kind as a requirement of normative legitimacy to unfairly impact it.

Constitutional legitimacy is not a yes or no binary, but a spectrum.Footnote ¹³⁴ Any mandatory inflexible public participation requirement as a precondition of normative legitimacy could compromise the need to bargain equitably with elites from most major groups. Elites from certain groups could use the justification of public participation to increase a constitution’s normative legitimacy instead of equitably bargaining with elites from most major political groups, as was done in cases of authoritarian constitution-making. Such scenarios could put the entire rationale behind this theory at risk. Furthermore, when they do not care about the public views, elites could undertake public participation as a box-ticking exercise, compromising the constitution’s normative legitimacy.

On the other hand, to ensure broad societal representativeness, if we require compulsory elections to drafting bodies or constitutional approval by referendums as a component of normative legitimacy, it can result in: (1) polarization and/or majoritarian outcomes;(2) locking out minority interests, or interests not formally represented in politics in a given time and space; and (3) being utilized to legitimate undemocratic constitutions.Footnote ¹³⁵

Thus, arguably the way we can ensure that a constitution is fairly representative of broader societal interests without other net negatives is by ensuring it is the product of an equitable elite bargain between most major political groups in a society. Moreover, as touched upon above, a constitution drafted in accordance with the requirements of the theory of equitable elite bargaining can increase representativeness in the future by opening up channels for participation and representation in the political system. Hence, this theory has the potential to ensure a degree of representativeness in the present without locking out future generations.

VII. Responding to potential critics and sceptics

The theory of equitable elite bargaining might be subjected to several objections, criticisms or concerns. This section will touch upon some of the most obvious ones and attempt to allay these concerns.

A primary issue could be raised by those who are not opposed to the theory of equitable elite bargaining in principle but believe that it (or its elements) can be read into the constituent power theory. It could be contended that the earlier versions of the constituent power theory can be improved to accommodate the requirements of the theory of equitable elite bargaining as they involved small groups of people channelling the will of the people. Such a claim gains added relevance considering that the theory of equitable elite bargaining hopes to gain some normative power (albeit at a secondary level) from representation-based reasons – which many argue is one of the main justifications for the constituent power theory as well.Footnote ¹³⁶ There are three major issues with such a viewpoint.

First, we would be back to square one, wherein constitutions (even if according to ‘the people’ benefits or being representative of them) actually reflect elite preferences and choices rather than the will of the people, while being wrongly legitimized in their name. This was precisely why there was a need to add some ‘legitimization from below’ to the earlier versions of the constituent power theory. Second, such a reimagination of the constituent power theory drains it of its primary contents – the will of the people. This is only aggravated in those cases when the equitable elite bargain is not accompanied by any public involvement whatsoever (say, by-elections to drafting bodies or approval via referendums) or the constitutional contents depart from public opinion. Third, when faced with significant public support for a particular version of a constitution, a requirement that the constituent power is channelled via an equitable bargain between elites from most major political groups will struggle to hold up. The examples of the failure of revision/total reform clauses in constitutions are a case in point. Revision/total reform clauses are often added to constitutions to prevent future constitutional redrafting from being unilateral. They generally demand compromises between different legislative groups in a manner that would be comparable to the requirements of the theory of equitable elite bargaining. In both Hungary and Venezuela, the existing constitutions contained types of revision/total reform clauses, following which, as explained earlier, could have been seen as a practical application of the theory of equitable elite bargaining. However, in both Hungary and Venezuela, modifying or not following them was seen as permissible by the constituent power theory as large sections of the populace wanted to go down a different route (albeit due to the misleading rhetoric of would-be-autocrats).

Another concern could come from those who acknowledge the realities of constitution-making and the shortfalls of the constituent power theory but still feel that the theory of equitable elite bargaining is theoretically troublesome because it does not centre ‘the people’.Footnote ¹³⁷ They would argue that though equitable elite bargains might be necessary, constitutions would still need some form of consent from the governed to be considered normatively legitimate.Footnote ¹³⁸ Advocates of this view might instead prefer a model similar to Arato’s two-stage post-sovereign conception of constitution-making, inspired by the South African constitution-making process.Footnote ¹³⁹ The first stage requires various political actors, along with social groups such as civil society organizations, to mutually agree on an interim constitution in a series of roundtable talks. In the second stage, a permanent constitution is drafted in a highly participatory manner by an elected parliament doubling as a drafting body. This permanent constitution is subject to the principles agreed to in the interim draft. To some, Arato’s theory might capture the merits of my approach while also including public (and civil society) involvement.

In response, I would first state that the theory of equitable elite bargaining does not preclude public (or civil society) involvement in constitution-making. The practical application of the theory could take the form of Arato’s two-stage model. However, as it pertains to according normative legitimacy, the theory simply looks at people’s role in constitution-making in line with realities.Footnote ¹⁴⁰ The manner and outcomes of public participation come down to elite preferences.Footnote ¹⁴¹ It is important to understand that in a case such as South Africa, the eventual outcome resulted from the equitable bargain between elites from the African National Congress (ANC) and the National Party (NP).Footnote ¹⁴² Over-simplifying things somewhat, the reason for the particular structure of the South African process was for the parties to ‘insure’ their respective interests.Footnote ¹⁴³ None of the components of the constitution-making process independently contributed to the success of the process, nor arguably would they have succeeded without the bargain between the ANC and NP.

In most cases, public involvement would voluntarily be incorporated into constitution-making processes by elites themselves – as they have often done in recent years. This is because public involvement in constitution-making is important to a constitution’s social and international legitimacy, and can have several other sociological benefits depending on the context, such as aiding state-building, citizenization, transparency, deadlock resolution and violence reduction.Footnote ¹⁴⁴ At the same time, in certain circumstances, public participation in constitution-making may come at a cost: it can impede sensitive negotiations, distract attention from important developmental issues, produce incoherent and unworkable texts, risk polarization and majoritarian excesses and make bargaining harder.Footnote ¹⁴⁵ In many new and post-conflict societies, where the lion’s share of modern-day constitution-making is occurring, meaningful public participation could be extremely hard (or impossible) due to low literacy rates, security issues and a lack of resources and infrastructure.Footnote ¹⁴⁶ Thus, it might be prudent to eschew or limit public participation in particular cases.Footnote ¹⁴⁷ Hence, it is better and more realistic if public involvement in constitution-making is introduced in accordance with the terms of an equitable bargaining process and not as a default. This would ensure that public participation is carried out for the goals it can serve in a specific circumstance and, as cautioned in the previous section, not as a means to unduly influence the constitution’s legitimacy or as a box-ticking exercise.

Further potential criticism could come from those who believe that this theory’s mainstream usage will result in moderated outcomes vis-à-vis human rights. Such a claim might be correct,Footnote ¹⁴⁸ but I do not see these as negatives. Chilton and Versteeg’s empirical research has shown how codifying rights in a constitution does not necessarily result in improved rights enforcement.Footnote ¹⁴⁹ Rights are only enforced if elected officials see a reason to do so or if not doing so would result in costs for elected officials.Footnote ¹⁵⁰ Most autocracies today have codified a large number of rights in their constitution, often exceeding 75.Footnote ¹⁵¹ Conversely, most stable democracies have a small number of rights, seldom exceeding 30.Footnote ¹⁵² Very few democracies feature in the top 100 in the rankings of the number of rights codified in a constitution.Footnote ¹⁵³ I am not trying to make a causal claim here; I am simply stating that moderated outcomes are not inherently concerning. Authoritarian regimes frequently use rights to window-dress their constitutions to deflect criticism.Footnote ¹⁵⁴ Institutional arrangements created as a result of equitable elite bargains would be more conducive to rights protection than simply having constitutions laden with a plethora of rights. Even if constitutions created in accordance with the theory of equitable elite bargaining result in moderated outcomes, those in power would be interested in upholding these outcomes.

This theory might also be criticized on the ground that, in real-world situations, it would demand too much and make constitution-making processes arduous, especially in societies characterized by weak political groups or low trust between different political groups. I concede this to be true. Nonetheless, I do not think it is a reason to abandon the theory. The response should be to find ways to facilitate smoother equitable bargains – a relatively unexplored topic in scholarship. Nonetheless, I believe the arduousness associated with this theory is a cost worth bearing for the improved outcomes that this theory’s mainstream usage would facilitate. Moreover, as Elkins et al. have shown, on average constitutions endure for nineteen years.Footnote ¹⁵⁵ A major reason why they are so short-livedFootnote ¹⁵⁶ is not that people’s preferences change, but rather because previous constitutions were often unilaterally drafted and/or did not gather buy-in from elites representing different political groups.Footnote ¹⁵⁷

Although the endurance of a constitution is not necessarily good, its immediate failure is not ideal either. Most of the values associated with constitutionalism have been known to improve with age.Footnote ¹⁵⁸ Constitutions drafted as a result of equitable elite bargaining would stand a greater chance of survival during their initial vulnerable years when the balance of power among the political forces that created the constitution tends to remain stable. Once a constitution survives for an initial vulnerable period, it is more likely to endure into the future.Footnote ¹⁵⁹ This is precisely why many older constitutions appear unlikely to be abandoned.

Lastly, although this theory largely remedies the core problems of the constituent power theory and its various iterations, it is not completely immune to results that go against the values of constitutionalism. There might be exceptional cases. For example, elites from different political groups in society might put aside their disagreements and cooperate to create a special kind of authoritarian society. Political elites might also put in place a system with little to no room for human rights, or one that does not protect the rights of unpopular minorities. Thus, in many ways this theory relies on the good motives of elites. These are possibilities that plague not only the theory of equitable elite bargaining but any other theory of constitutional legitimacy, including the constituent power theory. To prevent such scenarios, Bernal, a similar critic of the constituent power theory, opines that for any constitution to be considered legitimate, it must contain essential elements related to ideals of democratic constitutionalism.Footnote ¹⁶⁰

While not completely opposed to this idea, I would exercise caution in expanding this theory for reasons similar to adding a mandatory public participation requirement. In most cases, an equitable bargain between most major political groups (even when good motives are not at play) will, by default, produce constitutions incorporating many vital elements of democratic constitutionalism. This is because political groups in general (weak or strong) would not be willing to agree to terms that do not provide their political futures with some degree of insurance. Such insurance generally happens via certain elements of democratic constitutionalism.Footnote ¹⁶¹ Adding Bernal’s requirements (or other similar mechanisms) to the theory of equitable elite bargaining can allow elites to diminish the need to bargain equitably with elites from most major political groups. It could also divert attention away from the core bargaining points, or make bargaining more difficult by introducing new bargaining points. Additionally, forcing constitutional provisions that political elites might oppose in particular cases will do little for their enforcement. At the same time, as discussed earlier, suggestions such as Bernal’s do not adequately address the problem of undemocratic outcomes, considering that modern-day autocrats often operate with such constitutional provisions. Perhaps we should not let the perfect be the enemy of the good, but should be ready to accept some small risks – which are present not only with the theory of equitable elite bargaining, but also with any other similar alternative.