What is the constitutional fourth branch?
This symposium explores the role of “fourth branch” institutions, and specifically the role of independent electoral commissions (IECs) in protecting and promoting constitutional democracy. It does so by focusing on the global South, and Asia in particular. It aims to go beyond the “usual suspects” in comparative constitutional law,Footnote 1 and put the constitutional experiences of countries such as Indonesia, Kenya, Myanmar, Malaysia, and Sri Lanka at the centre of a decolonized constitutional project and understanding,Footnote 2 supplementing them with an examination of more-often-studied systems such as Australia and India.
These countries represent some of the leading systems with IECs in the Asian region, and beyond that, include examples of “best practice” in the design of IECs.Footnote 3 The Kenyan IEC, for example, drew heavily in its design on South Africa's IEC and South Africa's broader fourth branch constitutional architecture, which is arguably the leading global model of fourth branch institutional design.Footnote 4 The Australian electoral commission is one of the longest-running, effective fourth-branch institutions worldwide.
Asia is also an especially interesting site to examine the focus of IEC's, given the recent growth of IEC's in the region. As Langford et al note, there has been a significant growth in the number of IEC's in recent years: the number of IECs worldwide rose from 113 to 130 between 2006 and 2019, with 15 out of 24 Asian states possessing an IEC – tracking now the global average.Footnote 5
These developments reflect parallel developments in constitutional theory. Constitutional theory dating back at least to Locke and Montesquieu classified the work governments did into three branches: the legislative, which produced general laws, the executive, which carried out the legislature's instructions, and the judiciary, which resolved disputes about the meaning and application of the laws the legislature produced.Footnote 6 The three branches did everything a government had to do, or so constitutional theorists argued.
This tripartite structure sometimes emerged organically through gradual development. Sometimes it was created by revolution against governments in which the three functions were merged into a single institution, the absolute and arbitrary monarch. Simplifying enormously: At first constitutional theorists assumed that replacing absolutism with a tripartite structure was all that needed to be done. Eventually, though, they came to see that they had to theorize the possibility of challenges to tripartite governments as well.
The core of the theory that emerged was the understanding that stable government required that a fourth function be performed – the preservation of a decently functioning government itself. Put another way, theorists, of whom Hans Kelsen was the most prominent, came to understand that merely providing good government was not enough to guarantee that the government would persist. To use old-fashioned language, factious evildoers might take advantage of the government's structures to put themselves in place and restore absolute (in today's terms, authoritarian) rule.
With a fourth function identified, theorists began to consider whether a fourth branch might be needed, and if so, how it would be structured. By the end of the twentieth century, many agreed that such a branch was needed, and could consist of a number of institutions for protecting the constitution. For Hans Kelsen, these institutions were “guardians” of the constitution; Khaitan calls them collectively the guarantor branch.Footnote 7
As Tushnet argues in his contribution here, in modern constitutional democracies the primary threat to the constitution comes from the very political parties that have become central to the smooth functioning of the legislative and executive branches. Well-organized parties can win elections and dominate all three traditional branches (the judiciary through the influence of political parties on judicial appointment and removal). The result can be a return to authoritarianism (captured in the phrase “one person, one vote, one time”), or what political scientists have labelled competitive authoritarianism (where elections are held on a playing field so tilted that anti-authoritarian parties have essentially no chance of winning), or dominant-party democracy (where the playing field is tilted enough to make opposition victory quite unlikely but not impossible).
Fourth-branch institutions respond to the threat political parties pose to stable democracy.Footnote 8 They must be designed to be above party in some sense – or, in Khaitan's term “post-partisan”. Some scholars suggest they must also have specialized, as opposed, to general legal and political functions. A specialized constitutional court, for example, might be considered a prime candidate for inclusion in the fourth branch, but Khaitan questions (we think rightly) whether the jurisdiction of such courts is so broad as to disqualify them from being included in the category. The same is true for election management bodies, examined in this symposium. Whether they count as truly “fourth branch” institutions will depend on their insulation from partisan electoral pressures, and the degree to which they have specialized functions.Footnote 9
What are their formal institutional features?
One of the central questions the symposium addresses is the formal institutional design, structure and jurisdiction of independent electoral commissions.
Contributors note certain commonalities in the structure and design of IECs in the jurisdictions they examine. They also note variations in the method of appointment and degree of formal and de facto independence that IECs enjoy. As the summary account of the reasons for having fourth-branch institutions suggests, the role of IECs looks quite different in dominant party democracies compared to competitive ones, and in countries that incorporate two party or multi-party influence on a commission's composition and functioning. One pervasive question is, under what conditions can an IEC perform its core function of preserving the constitution in a competitive authoritarian or dominant-party system?
Another common theme of the papers is the degree to which the status of IECs as fourth branch constitutional institutions depends on them having formally entrenched constitutional standing or status, or some de facto equivalent political form of entrenchment. Khaitan, for example, argues that a defining feature of IECs as constitutional bodies is their entrenched status within a constitutional order. No doubt drawing on his UK knowledge and experience, he suggests that entrenchment is inevitably a question of degree, and may be formal or informal, or legal or political.Footnote 10 But he suggests that there must be some agreement among political parties that fourth-branch institutions enjoy independence and protection, even in the face of decisions that go against a party's interests. Attention to the Australian experience also suggests how this can play out: the Australian Electoral Commission, as Kildea and Murray show, has no formal constitutional status, but has played a longstanding and central role in Australian elections, which is not open to partisan dispute or contestation.Footnote 11 But the larger empirical study, by Langford et al, finds little support for this hypothesis, or that constitutionalization per se is a predictor of the effectiveness of an IEC.Footnote 12
Khaitan likewise suggests that fourth branch institutions are defined by reference to their special claim to “expertise”. They engage in a distinct mode of reasoning; decide substantively by weighting the particular “promise” they are enforcing especially heavily; and do not engage in a true all-things-considered judgment. As we have noted, Khaitan argues that constitutional courts cannot claim access to some special expertise. We think, in contrast, that the widespread practice of creating constitutional courts distinct from apex “ordinary” courts suggests that constitution-designers have a different understanding than Khaitan does about how judges on constitutional courts reason.
What is their function?
Another way of defining fourth branch institutions, however, is more “functional” in nature. IECs play a range of what Stacey and Miyandazi call “constitutive” and “regulative” roles: they organize and oversee elections, including special elections, plebiscites, and referendums, prescribe electoral rules and regulations, enforce those regulations, and play a role in public education surrounding elections. Their regulatory functions are also potentially quite broad: This may include setting, applying and determining the content of rules relating to candidate and party eligibility, prohibited speech and campaign finance regulations, rules and regulations for election observers, and sometimes electoral boundaries.Footnote 13 By ensuring the integrity of elections, IECs also help prevent the risk of electoral violence and play a deeper role in stabilizing constitutional democracy.Footnote 14
As one of us (Tushnet) notes in his opening contribution to the symposium, every constitutional system must maintain itself; or be self-perpetuating in the face of at least modest external shocks.Footnote 15 If it undergoes severe shocks, constitutional design may not be sufficient to protect democracy or to ensure that institutions are self-perpetuating. There are a couple of modes of doing this: culture or “civic virtue” and structure or institutional design. Civic virtue might be perpetuated through religion, or a commitment to markets, but neither seem fully adequate to preserving or stabilizing constitutional democracy.
Another approach, therefore, is to rely on political forces, or a web of institutional checks and balances. Madison, for example, envisaged a constitution that would be politically self-stabilising or enforcing, or where “faction would check faction” as a means of guarding against constitutional erosion or breach. Modern political parties have emerged in ways that create conflicts and convergences of interest that undermine the efficacy of Madisonian mechanisms.Footnote 16 And this creates the need for a range of structural “checking” institutions.
One leading institution of this kind is a constitutional court. But courts tend to transform political disputes into legal ones, which is not always the best or most logical way of protecting and promoting democracy.Footnote 17 They may also lack expertise in certain areas. Involving courts in certain kinds of checking and balancing function may also risk drawing them into “high” or “mega” political disputes that then undermine their broader independence and capacity to perform their core constitutional function, of upholding the rule of law and protecting individual rights and liberties.Footnote 18
Constitutional courts, therefore, will often be necessary but not sufficient to create a truly effective web of constitutional checks and balances. Or at the very least, their effectiveness will be increased where they operate in conjunction, or collaboration, with a range of independent fourth branch institutions – including IECs.Footnote 19
Well-functioning IECs will tend to be non-legalistic in approach, have relevant expertise, and be capable of overseeing elections without jeopardising any broader constitutional functions.Footnote 20 In fact, this is their core constitutional function. And expertise of this kind can be especially valuable in a range of settings: in large polities, where there are issues of sheer scale and capacity to deliver ballots and election infrastructure to hundreds of millions of voters; in contexts of reserved seats, where there are technical questions about which lists candidates should be allocated to; and in countries with particular problems of election fraud, manipulation and interference.
Take Indonesia: The General Election Commission or KPU employs 2800 commissioners nationally, and 6 million poll workers on election day, to supervise elections in which there are 193 million eligible voters. In 2019, they also supervised simultaneous elections involving 20 000 legislative seats and 800 000 polling stations, involving different levels of office.Footnote 21
Another argument for IECs as part a broader web of independent institutions is that they provide a form of “insurance” against the risk that would-be authoritarian actors may capture all independent oversight bodies.Footnote 22 In Kenya, for example, the Electoral Commission became increasingly non-independent from 2014 onwards, but the Court became more independent, in ways that allowed it to check potential electoral regularities in the 2017 presidential elections.Footnote 23
One could also imagine the opposite dynamic occurring. On the basis of its 2008 constitution, the President of Myanmar appoints a broad variety of executive and oversight institutions, including the Union Election Commission, the Court of Audits, and the Commander-in-Chief of the Armed Forces, whereas the Constitutional Court is appointed in equal parts by the President and the Speakers of the Upper and Lower Houses of Parliament.
Khaitan frames many of these same ideas in quasi-originalist/purposive constitutional terms, that is, by describing constitution framers’ understandings about these institutions’ purposes. He claims that fourth branch institutions make specific promises to the constitutional people – not just generally self-enforcing, but to ensure the credibility of specific constitutional promises. The temporal dimension to certain promises is key. Some promises, like legality, are guaranteed by the traditional separation of powers.Footnote 24 But other promises, he suggests, cannot be guaranteed within that model, and it is the need to make these promises credible that gives rise to the role of fourth branch institutions.
It is also possible, as Butt and Siregar note, for a country to disaggregate these various functions further and give them to different fourth branch institutions. In Indonesia, for example, the Constitution provides for three separate electoral institutions: the General Election Commission (Komisi Pemilahan Umum or KPU) with responsibility for running national elections, the Election Supervisory Board (Bawaslu, Bagan Pengawas Pemilahan Umum) with responsibility for election planning logistics and oversight, and the Electoral Honour Council (Dewan Kehormatan Penyelenggara Pemilu or DKPP) which hears allegations of breaches of electoral law, and complaints about the other institutions.
Determinants of success
What, however, determines whether IECs are in fact “well-functioning”, or successful in performing these various roles or functions? We offer the following thoughts as a provocation rather than as a firm conclusion.Footnote 25
Taking the contributions to this symposium as a whole, we are struck by the roles contingency, personality, and randomness play in the stories of success and failure.Footnote 26 An IEC whose design remains unchanged can work well when it is headed by one person, badly when it is headed by another. Civil servants sometimes do well, sometimes do badly, when they run IECs. The same institution run by the same people can succeed when it meshes with a constitutional court, and fail when the constitutional court is weaker.
We might think that contingency and randomness argue against the proposition to which we as constitutional lawyers are committed, that institutions matter. Perhaps, though, institutions matter, but in such complex ways – with variation along so many dimensions – that we cannot prescribe general design principles for constructing well-functioning IECs. Perhaps the best we can do is identify matters of interest – that constitution designers should think about whether they want their IEC to be multimember or headed by a single commissioner, or whether they want to compose it through bipartisan or multiparty selection of IEC members. Not only will the actual choices be determined by local conditions and, importantly, by the political circumstances at the moment of design; even more, how those choices will play out in the short and long runs may be largely unpredictable.
Langford et al engage in a large-n empirical analysis of the effectiveness of IECs across time.Footnote 27 They first note the increase in IECs worldwide from 113 to 130 from 2006 to 2019, of which 70 and 90 respectively were in democratic constitutional systems. Next, they examine three factors as potentially helpful to predicting the effectiveness of IECs in preventing electoral violence and promoting confidence in electoral outcomes: the commission's independence, the constitutionalization or formal entrenchment of its role, and its capacity (as measured by staffing and resources). Of these factors, they further find that IEC capacity is the most important predictor of positive electoral outcomes.Footnote 28 For constitutionalization, they find a weak positive correlation between electoral violence and the decision to entrench an IEC's role – suggesting a potential reverse causal story, whereby violence leads to the impetus to adopt such a commission, but without significant changes in the relevant outcome in the short-run.Footnote 29
The attitudes of political elites toward an IEC and their conception of the IEC's role are linked to IEC capacity and independence. In some countries, elites seem to see the benefits of an independent and effective IEC in protecting electoral integrity, whereas in others they see it as an obstacle to their instrumental political goals. This may also be reflected in their attitudes toward independence and funding, in ways that suggest multiple pathways of influence over ultimate electoral outcomes.Footnote 30
Another factor that several contributors point to is political leadership and engagement with civil society: in her careful ethnographic study of the Sri Lankan Electoral Commission, for example, Samararatne notes the importance of the charisma and personal leadership of the Commissioner in ensuring free and fair elections, and his commitment to working with civil society as a key to the Commission's success.Footnote 31 TN Sesahn appears to have played a similar role in India, with at least a short-term legacy (perhaps to be dissipated with the re-establishment of dominant-party government).
In their equally nuanced study of the electoral commission in Myanmar, Renshaw and Lidauer note the degree to which a difference in composition and leadership has affected the role and performance of the Union Election Commission. In 2015, the UEC was led by a former military leader who argued that while he had close friendships with the government and military, “if they ask me to help them win the election, I would say ‘sorry’”.Footnote 32 In 2016, in contrast, the NLD government appointed a former civil servant with close ties to the government and a much less independent or robust conception of the UEC's role. Indicating that structural categories like “civil servant” may not capture a great deal of what we are interested in, India's IEC became a dynamic force in political life with the accession of TN Seshan, a career civil servant, to the head of the Indian election commission. What is important, in all these cases, is the degree to which different individuals, appointed by and with close ties to governments, saw their role in terms of a fourth branch independent checking-function.
Related to leadership will be the historical experience and trajectory of an IEC, and how that informs the role conception or approach of current commissioners. In Malaysia, for example, Shah notes that attempts by the electoral commission to assert a robust and independent role in electoral redistricting in the late 1950s led in 1960 to the removal of the Electoral Commission Chairman and then a radical reduction in the Commission's role and jurisdiction. From 1960 to 2018, this also arguably encouraged a quite deferential and limited role on the part of the Commission in regulating the practices of Malaysia's dominant party.Footnote 33 It was only in 2018 after the change of government that quite different appointments were made to the Commission in ways that might be expected to lead to a gradual shift in the Commission's approach and role conception.Footnote 34
Another related factor is the degree to which IECs work collaboratively with constitutional courts, and vice versa. In Indonesia, for example, the Electoral Commission has consistently implemented the Court's orders, including controversial decisions about the need to ensure simultaneous presidential and legislative elections.Footnote 35 It has done so at considerable human cost: Attempting to count votes and meet relevant deadlines in simultaneous elections led to the deaths of 600 poll workers from exhaustion.Footnote 36 The Constitutional Court, in turn, has repeatedly cited the KPU's decisions in judgments regarding elections, thereby buttressing its role and authority.Footnote 37
How do they interact with other institutions?
In any democratic constitution, horizontal checks and balances can also intersect with vertical checks and balances. For electoral commissions, as Michael Pal notes in his article, this means their role intersects with federalism guarantees and institutions.Footnote 38
Indeed, one of the most pressing constitutional design questions in a federal system with fourth branch institutions is whether those institutions should have national jurisdiction, or jurisdiction only in respect of federal practices – such as federal elections. Countries such as Australia and Canada have a decentralized model, whereby each state has its own fourth branch institution responsible for the oversight of state elections. But India provides a leading example of a model of unitary electoral administration, whereby the national election commission also oversees both national and state elections. Some countries have complex hybrids. In Brazil, for example, local entities can opt in to either federal or state election administration. In Australia, there are also mechanisms for cooperation among various state commissions, or within a model of self-rule.
As Pal notes, there are practical reasons for a unified model: National commissions may be better staffed and funded.Footnote 39 They may be part of a post-independence nation-building project. Designers may have an interest in making appointments, in ways that involve an attempt to extend their own influence. But the theoretical concern is that election administration is not a technical or side issue but the way that the right to vote is brought to life. Here, the challenge becomes balancing the danger of tyranny by a national versus local majority: National commissions may help protect against the danger of a local majority oppressing a local minority, though not in every case.Footnote 40 The reverse may also apply: state commissions may in some cases help protect against the danger of national over-reach.Footnote 41
Another question relating to the interaction between IECs and other institutions concerns the legal and democratic accountability of these institutions themselves. Bhat, for example, argues that while clearly powerful and effective on many levels, the Indian Electoral Commission has progressively expanded its informal or de facto powers in ways that create a serious question of legal and political accountability. While the Commission's model of “architectural, nudge and notice”-based regulation has been highly effective in many cases, it has occurred without legislative authorization or substantive judicial oversight. Bhat therefore questions whether greater judicial oversight of the commission's functions would be appropriate.Footnote 42
Are IECs always democracy promoting?
A final question is whether IECs are always likely to be democracy promoting or may be captured by non-democratic regimes or dominant parties in ways that mean they end up playing an anti- rather than pro-democratic function.
It is clear that some electoral commissions may have limited formal or de facto independence, and because of that play a largely quiescent role in supervising elections. This kind of “passive” virtue on the part of a non-independent “I”EC may also contribute to legitimating elections that are not in fact free and fair.Footnote 43 Indeed, this may be quite common in some dominant party democracies and competitive authoritarian regimes.
In Malaysia, for example, Shah suggests that between 1960 and 2018 the sheer dominance of the ruling party, the Barisan Nasional [BN] at both the federal and state levels allowed the government to fully control the appointment of EC members. It was thus quite predictable that it would play a limited role in supervising BN electoral practices.
It is also possible, however, for non-independent ECs to play an even more active role in undermining electoral democracy. As one of us has recently noted (with David Landau), would-be authoritarian actors may seek to subvert or capture democratic norms or institutions in a variety of ways, and sometimes so as to invert their purpose.Footnote 44 This form of “abusive borrowing” can also apply as much to ECs as to other constitutional democratic institutions – that is, transform an EC from an institution designed to uphold electoral integrity into one that compromises it.
While it is still early days, there are some intimations that this is what has been occurring in Myanmar. The military regime used the election commission to lead the country into a controlled transition starting from 2010, and to oversee a second of round elections which they wanted to be “clean” in 2015. However, they deposed of UEC after the third elections in 2020 when electoral results became too detrimental to their interests. During the last term, the UEC has been dominated by civil servants loyal to the NLD government. Nevertheless, it has issued a series of decisions with deeply problematic democratic consequences: they have censored the electoral platforms of a range of non-NLD candidates, continued to disqualify Muslim candidates for failure to prove their citizenship, and cancelled elections at unprecedented scale in a range of states experiencing ethnic violence. It also upheld the disenfranchisement of Rohingya which had become stateless and lost their right to vote prior to the 2015 elections. However, they have embraced technology less than the previous UEC and apparently produced poorer voter lists, providing an avenue for the military regime to argue there was electoral fraud, despite clear evidence of intentional manipulation.Footnote 45
At the very least, it seems important to note the current limits to IEC independence in many Asian jurisdictions. As Langford et al note, a defining feature of Asian electoral commissions may in fact be their unusual combination of quite high capacity coupled with low de facto political independence.Footnote 46 And if this is true, it may be some time before see Asian IEC's playing a central role in the process of democracy protection or promotion.
