The problem of sustaining a democratic constitution – the Madisonian argument
Constitutional theory dating to Montesquieu identified three branches of government, each with a specific function: the legislature enacted general rules, the executive enforced the rules, and the judiciary resolved disputes about the rules’ meaning and application. Every government had to have these branches in some form; that is, the branches were necessary elements in a governance structure. In addition, the branches were exhaustive: that is, taken together they did everything a government could do.Footnote 1
One desirable feature of a governance structure is that it be reasonably stable. In ordinary times the structure must grind out policies, execute them, and deal with ensuing problems in a routine way.Footnote 2 Modest shocks – a small-ish war or some moderate economic disruptions or nagging failures of governance in limited domains – might lead to departures from the ordinary course, but after the shock dissipates or the failures remedied the structure should return to something like its prior state, perhaps modified a bit because people have learned that some adaptations should be built into the structure to be available as needed.Footnote 3
Further, well-designed governments require some mechanisms for adapting to changes in the policy problems the world throws up, changes in the preferences of the nation's people (sometimes the result of changes in the very composition of that people), and changes in the available technologies of governance.
What, though, ensures that a constitutional system achieves an appropriate combination of reasonable stability with adaptability? Or, put another way, what can we do to ensure as best we can that a governance structure can sustain itself over time? Classical constitutional theory identified two possibilities. Call them “civic virtue” and “structures.”
The difficulty with civic virtue as a guarantor of regime stability is obvious enough: constitutional and political theorists have not been able to come up with institutions (mechanisms) that have any substantial chance of reliably reproducing civic virtue in the citizenry. The idea that structures of governance could do a decent job of ensuring their own reproduction and adaptability goes back almost as far as the civic republican tradition. For present purposes we can associate it with James Madison.
Begin with adaptability. An orderly mechanism for amending the constitution allows the system to adapt to changes in the technologies of governance. Absent special considerations, the ordinary process of policy-adoption through legislation and executive decision-making is sufficient to ensure adaptation to changes in preferences and the array of policy problems the nation faces.Footnote 4
Now turn to the threats to stability. Consider first the threat posed by self-aggrandizement: legislators would seek to enforce the law themselves, for example by imposing sanctions on identified individuals. Executive officials would seek to legislate by decree and use their resources to force individuals to comply. And courts, though the weakest branch, might aggressively expand their jurisdiction and “interpret” laws in ways that effectively transformed them.
Constitution-designers could write provisions purporting to prohibit these attempted incursions, either specifically – as in a constitutional ban on bills of attainder – or generally, as in the famous formulation in the Massachusetts constitution.Footnote 5 But, Madison feared, these provisions would be mere “parchment barriers.”Footnote 6 He argued that competition among the branches was a better mechanism for guaranteeing that each branch did only what it was designed to do. As Madison put it, “Ambition must be made to counteract ambition. The interest of the man must be connected with the constitutional rights of the place.”Footnote 7 Executive officials would have the resources (mostly military) to resist legislative attempts to enforce the law, and legislators would have the resources (mostly financial) to resist executive efforts to legislate. And, because the interests of the occupants of government position were connected to the rights of the position, executive officials and legislators would have incentives to use those resources.
The mechanisms Madison identified to ensure regime stability were quite ingenious, and perhaps could work in principle. Reality defeated them, though. The argument about competition among the branches was defeated by the rise of political parties organized on a national scale. As a prominent article puts it, the United States now has a system of separation of parties, not separation of powers.Footnote 8 And this is true more or less everywhere, though the nature of party systems varies among nations, with different systems having different implications for the separation of powers.
We can see the difficulty most easily in parliamentary systems.Footnote 9 Where the party system is reasonably well-organized – with a handful of parties contesting each election with an eye to forming either a majority or a coalition government – the executive and legislature will collaborate rather than compete. The majority party or the governing coalition will decide whether acting through legislation or by executive decree best advances the government's program, without regard to Montesquiean formalities. So-called separation-of-powers systems, in which separate elections are conducted for the chief executive and the legislature, can use the Madisonian mechanism in a world of political parties, but only when government is divided, that is, when the chief executive is from one party and the legislature is controlled by other parties. Other configurations of party power are possible, but all, I believe, end up posing real problems for the Madisonian account.
Preserving democracy in a party political world: the conceptual argument for IPDs
Writing against the background of German parliamentary development, Hans Kelsen saw that the Madisonian mechanism of competition among the branches could not ensure regime stability. He argued that guaranteeing stability was a fourth function of governance structures. Its existence had perhaps been obscured by the Madisonian argument, which took the function to be performed as a by-product of the creation of branches whose primary purposes were the Montesquiean ones. The Madisonian argument did not require that politicians be committed in principle to preserving the constitution. Put another way, they did not have to make credible commitments to doing so, because institutional competition would do the job. Where that competition fails – that is, in a party-political world –we are forced to fall back on the politicians’ commitment to preserving the constitution.Footnote 10 But, in such a world politicians are committed only to advancing their party programs, and cannot credibly promise that their commitments to the constitution will prevail over their programmatic goals.Footnote 11 We need some other institution, and for Kelsen that institution was to be the “guardian of the constitution.”Footnote 12
As is well known, Kelsen thought that an institution he called the constitutional court could serve as that guardian. Its function was to preserve the constitution in conditions of party government, and that function dictated many of its characteristics. Those characteristics are well-known, and I simply highlight two: the constitutional court was to be removed from, or above, party politics, primarily by means of the mechanisms for selecting its members, and its work would be focused on the constitution understood as a legal document that affected the way politics operated.
Kelsen may have been right in seeing party politics as a threat to the constitution, and in seeing the need to create something to serve as a guardian of the constitution. He may have been wrong, though, in thinking that a single such institution, the constitutional court, would be sufficient. Perhaps we should take seriously the plural in the title of the South African Constitution's Chapter Nine: “Institutions Supporting Constitutional Democracy.” In doing so we might move the Madisonian argument to a higher level: not competition among institutions above party politics, but competition among institutions implicated in party politics but each in a slightly different way. A constitutional court might fail to protect the constitution against a specific threat, but perhaps the nation's ombudsperson will do so – or the nation's auditor general, or its public prosecutor. A nation with enough institutions supporting constitutional democracy might be able to protect against specific threats as they arise. Not in any systematic way, though: it is not that the constitutional court is systematically going to a better job of protecting rights than an ombuds office, or a public prosecutor a better job at attacking corruption than a court,. Rather, the hope is that with respect to any specific threat to the constitution, at least one of the many institutions available to support the constitution will manage to do a good job, perhaps (often) for quite idiosyncratic and unreproducible reasons.
Preserving constitutional democracy with IPDs: the functional argument
In a party-political world neither the legislature nor the executive could be trusted to perform the function except under specific, and probably unusual, conditions. Absent those conditions, legislatures controlled by one party would not investigate threats to the constitution posed by an executive of their own party, and would exaggerate the threats posed by an executive of the opposition party – and conversely with executive oversight of threats posed by the legislature. Call this a problem of conflict or convergence of interest.
The tradition deriving from Kelsen came to understand that there were risks associated with assigning the guardianship function to the ordinary judiciary. The frequent association of high politics with party politics posed a threat to the independence of the judiciary that constitutionalism required. And judges on ordinary courts have a generally legalistic cast of mind, which might impair their ability to arrive at constitutionally appropriate resolutions of constitutional conflicts. Call these problems of independence and expertise.
These problems emerge from a consideration of the reasons for requiring a fourth branch of government. Once they are brought into the open, though, we can see that other tasks within a constitutional system have similar characteristics. The most obvious are supervision of electoral fairness and rooting out high-level corruption.Footnote 13 This Article focuses on the former.
The problems of conflict of interest, independence, and expertise suggest the possibility that Kelsen had a too-limited vision of the tasks the constitutional court – or some other guardian of the constitution – might take on. After discussing in more detail the tasks that might be affected by those problems, I turn to the question of whether constitutional designers should consider expanding the constitutional court's remit to include the larger set of tasks, including supervision of electoral fairness and oversight of anti-corruption efforts – or whether, and if so why, designers should consider creation of a multi-institutional fourth branch of government.
Begin with conflicts and convergences of interest. We must be careful in spelling out what amounts to a troubling conflict or convergence of interest. Consider substantive programs that provide benefits to constituents who might then be drawn to affiliate themselves with the party whose platforms include preserving or extending those programs. The Workers’ Party in Brazil, for example, enacted a generous family allowance program, financed from the then substantial revenues from resource extraction, which cemented the affiliation of poor families to the Party – and had a non-trivial effect on reducing economic inequality within Brazil. These programs aim at entrenching the party in power, or, put another way, they reflect the governing party's self-interest in staying in power.
This sort of self-interest cannot be enough to take the policy out of the hands of ordinary party politics, though. Political parties’ programs consist of promises that, if fulfilled, will enhance the parties’ electoral support. As political scientist E.E. Schattschneider put it long ago, “New policies make a new politics.”Footnote 14 Or, put another way, all political platforms aim at securing a party's self-entrenchment.Footnote 15 To treat self-entrenchment as problematic in itself would be to remove all policies from the domain of democratic control. What we need to do is identify political domains where conflicts or confluences of interests occur no matter what the array of political power is. The Madisonian mechanisms can work in other domains, but not in these.
Consider the family-allowance example. One policy argument against that policy is that it is unsustainable.Footnote 16 And, as economists say, if something cannot go on forever it will stop.Footnote 17 We might describe the family-allowance program as self-limiting: its political advantages have a “natural” lifetime, after which the opposition can take power. Most policy domains deal with policies like this – ones that might give the party-in-power some temporary political advantages, but those advantages are not sustainable.
Sometimes, though, policies can have a permanent self-entrenching effect. Manipulation of election rules, including gerrymandering, can do so, for example. These practices can erect impenetrable walls around the party-in-power. Here the conflict-of-interest explanation of why some sort of constitutional guardianship is necessary seems apt.
The confluence-of-interest explanation also has some bite as well, though the mechanism comes in two variants. The general description is that opposition parties do not challenge what the party-in-power does. The more obvious mechanism is that opposition parties don't want to oppose manipulation of election rules because they hope to take power and then manipulate those rules themselves. Here the mechanism is this: parties-in-power try to manipulate election rules to entrench themselves permanently, but their efforts are not always entirely successful. They might miscalculate and leave room for the opposition to “sneak in,” or they might adopt election rules that would be permanently entrenching under ordinary circumstances but extraordinary events might intervene. A second mechanism is that opposition parties accept permanent minority status in exchange for pay-offs from the party-in-power – salaries and other financial benefits flowing directly from holding office, for example.
There is a final large topic to deal with. Kelsen's argument led to the creation of constitutional courts. Why are other fourth branch institutions needed? Many constitutions do expressly assign responsibility for supervising elections to the constitutional court. Sometimes constitution-designers make that choice because of concern that the nation lacks enough trained professionals with sufficient stature to staff more than one guardian institution. Sometimes, though, the choice appears to be driven by design concerns rather than personnel ones.
The functional logic underlying the fourth branch suggests reasons for creating several fourth branch institutions. For, in addition to concerns about conflicts and confluences of interest, constitution-designers need to attend to questions of expertise. Some fourth-branch institutions might require forms of expertise not readily available to the constitutional court, and some fourth-branch tasks might place undesirable burdens on that court.
The latter concern is most obvious in connection with using the constitutional court to supervise elections by determining who qualifies for the ballot, counting votes, and the like – a subset of what Ran Hirschl calls “megapolitics.”Footnote 18 The core of the constitutional court's work requires that it navigate through domains that combine law and high politics. In doing so the constitutional courts always runs the risk of coming to be seen as directly implicated in party politics. Supervision of elections exacerbates that risk: supporters of a major presidential candidate who is plausibly subject to disqualification under the constitution (but who has plausible arguments against disqualification) and is in fact disqualified by the constitutional court might well see the court's decision as party-political. Similarly with excluding parties from the ballot and with drawing constituency boundaries.
There are other risks associated with making megapolitical decisions. They might induce an overly strategic cast of mind in the judges, which might affect them even in their core work: no matter how hard they try, judges might find it quite difficult to exclude from their thinking that interpreting the constitution to require a major candidate's disqualification will have significant political effects.Footnote 19 Or, in the other direction, the judges might be overly legalistic, discounting political consequences too much. These effects might be especially troubling if they have the effect of reducing support for the constitutional court when it seeks to guard the constitution against the effects of party politics on constitutional structures and individual rights.
We should not place too much weight on the proposition that constitutional courts cannot call upon expertise when it is required. Public-law litigation in the United States and public interest litigation in India have shown that courts have the capacity to do many things that fall outside what Lon Fuller called “the forms and limits of adjudication.”Footnote 20 They can appoint experts to provide advice, and special hearing officers to obtain case-focused information, for example. So, concerns about expertise need not preclude assigning fourth-branch tasks to the constitutional court – with one important caveat: the US and Indian legal cultures turned out to be reasonably receptive to augmenting the courts with ad hoc “institutions” to provide the courts with expertise; other legal systems might be much more resistant to doing so. In the latter cultures constitutional courts may find themselves authorized to deal with problems as to which they lack sufficient expertise and yet reluctant to augment their capacity. So, creating additional fourth branch institutions in these “legalistic” cultures might be particularly desirable.
The conclusion here is that constitution-designers need not create a fourth branch composed of several institutions. In some settings authorizing the constitutional court to perform all the tasks identified by the logic of conflicts of interest as qualified by concerns about expertise. In other settings, though, creating several fourth-branch institutions might well be the more sensible choice.Footnote 21
Some risks associated with non-court IPDs
Election administration is of course a prime candidate for an IPD, precisely because the partisan stakes in election administration can be quite high: drawing district boundaries can have dramatic effects on outcomes, decisions about which parties can appear on a ballot can of course affect result. Even ensuring that those who cast ballots are eligible to do so can become the focus of deep partisan division. As the other articles in this issue show, the design issues associated with constitutional courts – how to balance independence and accountability – arise in connection with election management bodies as well.
One feature of election administration is that it can call upon a kind of expertise not readily available to courts. Experts in election administration may well be more sensitive to the political implications of various choices, and appropriately so (because they must be accountable to the larger political system) than experts in law. Yet, just as excessive legalism might be a pathology of constitutional courts, so too might there be pathologies, or at least subtle political “tilts”, to decisions by electoral management bodies.
A brief examination of some features of election management bodies in India and South Korea illustrates one such pathology – the possibility that the election management body will take a distinctive, and not politically neutral, view of how politics should be conducted. In both nations the election management bodies are regarded as having been highly successful in maintaining independence, sometimes even when a single party dominates politics. In both, though, some of the bodies’ decisions seem to seek to impose a kind of order – sometimes associated in India with a “middle class” perspective – on a politics otherwise characterized by intense popular involvement, sometimes to the point of raucousness. Their vision of politics is not politically neutral.Footnote 22
India
Initially the Commission's primary task was election administration narrowly defined: maintaining election rolls, ensuring that voters could cast their votes freely and without intimidation (a not insubstantial requirement in a society characterized at least in part by a social hierarchy in which deference to “higher ups” was common), and ensuring that votes were honestly counted. Given India's size, both physically and more important in population, these “mere” administrative chores were quite substantial.
By all accounts the Election Commission's performance over its first decades was impressive. India's elections were indeed reasonably free and fair, with election rolls reasonably accurate given the size of India's population, although of course there were incidents of failures to register voters and of voter intimidation.Footnote 23 Political elites and the public appear not to have attributed these problems to the Election Commission. The Election Commission garnered substantial popular support and respect: A report in 1999 found that the Commission had a higher level of public trust than any other institution, including the courts and the police.Footnote 24
The Commission was able to do a good job of election administration during this period in part because of the political context. The Congress Party was dominant politically, drawing on its legacy from the independence struggle and on policy stances that were broadly supported. That context changed in the 1970s and 1980s, first with the emergence of serious challenges to Congress Party rule, then with the imposition of an emergency regime, which resulted eventually in a system of coalition government that lasted into the early twenty-first century. A system of coalition government made Madisonian mechanisms more available to control election fraud and manipulation, with the potential for reducing the need for an independent Election Commission. Yet, the end of dominant-party rule opened space for the Election Commission to develop its own agenda for running elections, moving beyond mere administration into more substantive control of political campaigns.
One effect of the change in the party system was that the parties agreed upon a “Code of Conduct” for conducting election campaigns. Initially adopted through multi-party negotiations in the state of Kerala in 1960 – importantly, a state with vigorous party competition even then –, the Code gained national prominence within a decade, and became part of the Election Commission's public-facing communications.Footnote 25 Compliance with the Code is voluntary, at least nominally: the Election Commission cannot sanction a candidate or party merely for doing something “prohibited” by the Code.
A new Chief Election Commissioner arrived at the same time that the new party system emerged. TN Seshan, whose work in the civil service culminated in the position of Cabinet Secretary (a policy advisory and coordinating position, not a policy making one), took office in 1990 at a time when the Congress Party was providing support in Parliament to a government headed by Chandra Shekhar without a formal coalition agreement. Seshan was a dynamic figure who continued the Commission's election administration work. More important, invoking the idea that politics should be conducted differently once votes are being cast – for example, a candidate should not be able to adjust her campaign platform after some voters had cast their votes in light of the platforms they had been presented with –, Seshan used that power as a method of getting parties to comply with campaign regulations he favored. As the period between the opening of a campaign and the first days of voting contracted and the length of the voting period grew, the Commission's ability to control campaign behavior increased as well.
Widespread illiteracy meant that parties appeared on the ballot – and so in their campaign activities – with symbols; today Congress's symbol is the palm of a hand, that of the governing Bharatiya Janata Party a flowering lotus. The Election Commission allocates symbols to registered parties. Registration requires a party to submit its party constitution, which must “conform to the spirit of the Constitution.” Seshan used the registration power to induce compliance with campaign regulations.
Historian David Gilmartin describes several of Seshan's interventions. Seshan challenged “the announcement or undertaking of any government policies during election campaigns that might be construed as government attempts to use its power to unduly influence the voters.” A government promise to expand the list of Scheduled Castes, announcements of new development plans, even visits by government ministers to areas with by-elections – all came under Seshan's critical eye. He attempted to block the government from adopting a new cotton export program during a campaign. Seshan vigorously “enforced” – via publicity – Code policies aimed at reducing purely emotional appeals to voters, through loudspeakers, banners, and posters. According to political scientist Christophe Jaffrelot, the 1996 elections “were no longer marked by innumerable rallies, a plethora of posters, and the use of blaring mobile loudspeakers or video vans….” Gilmartin notes that Seshan's initiatives took place against a background of public concern that “muscle and money power” rather than deliberation about policy were determining election outcomes. “Suspicion of politics… was particularly pervasive in the urban middle class,” and Seshan tapped into that concern – a concern that may well have resulted from the end of Congress Party dominance. As Jaffrelot observes, “While this newly introduced discipline cut back on the festive aspect of the elections, it also reduced the funding needs of parties, which was expected to impact the degree of corruption.”Footnote 26
Gilmartin reprints a cartoon in which Seshan says, “Promise of prosperity, jobs, exports, border security… He is clearly trying to influence the voter? Can't be allowed!” He also quotes a prominent politician apologizing for an evident Code violation by saying, “Election rules do not permit me to promise you anything. So I cannot do so. I shall settle all your grievances after the election.”Footnote 27 Taken broadly, Seshan's project to make political campaigns the location for rational public deliberation about competing policy agendas was never realistic, as the cartoon and the politician's speech show.
We might take Seshan's project more narrowly, though. Read against a background political tradition in which politics has “a carnival-like atmosphere”,Footnote 28 the project might be understood as an effort to change the balance in campaigns to make them somewhat more deliberative. Even on this understanding the project was probably unrealistic as well. Constraining the government-in-power's use of the levers of governance to further the party's electoral prospects would always be quite difficult. The apparent re-emergence of dominant party government in the 2010s might undermine the Commission's independence as well.
This case study might give us some insights into Election Commissions as IPDs. Seshan's vision of political campaigns is regularly characterized as bureaucratic and middle-class; a recent discussion refers to “a kind of technocratic efficiency by the ECI that appeals to the Indian middle class.”Footnote 29 The vision Seshan offered, whatever its specific roots in Indian history and society, derived in part from the largely technical work associated with election administration, a task performed by election commissions around the world. Seeing compiling voter rolls and accurately identifying voters as largely technical tasks, election commissioners may come to see the other matters in their jurisdiction, including the regulation of political campaigns, as similarly subject to technical constraint. And the source of the constraint is obvious: Counting votes is an objective activity; so (on the technocratic view) should political campaigns be objective.
But, as Seshan's failed attempt to distinguish between political promises and voter bribery indicates, the technocratic vision of political campaigns is itself a political one. In unpublished work Michael Pal suggests that the technocratic vision is part of a political program of nation-building. And the politics of expertise can be political in a narrower sense, as the use of the term “middle-class” to describe Seshan's vision suggests. Though that term might not be entirely accurate, it does suggest that just as economic classes support political agendas depending upon their class position, so too do technocrats.
South Korea
Following the end of the Japanese occupation in 1945, South Korea was first administered by the United States army. Accepting the post-1945 division of the peninsula, the Republic of Korea adopted its first constitution in 1948. From that time until 1987 South Korea had an authoritarian government, sometimes civilian, sometimes military.Footnote 30 Widespread protests in the 1980s ultimately led to the creation of a new democratic government. Since then Korean politics has been dominated by what one scholar calls a “two plus two” system: “two major parties (the conservative and the center-left) and two minor parties (another conservative and the progressive),” with the conservative party holding a majority in the National Assembly for most of the time since 1987.Footnote 31 The major parties, though, regularly split and recombine into new ones, primarily when factional leaders find themselves frustrated at developments within the parties with which they have temporarily affiliated themselves. Parallel to electoral politics, and important to the story that follows, there have been recurrent upsurges in politics “in the streets”, organized by social movements without formal affiliations to the existing parties.
Several Constitutional Court decisions reflect the view that politics should be relatively orderly. In discussing a statute limiting people from publishing advertisements supporting or opposing political candidates – basically, spending on campaigns independent of the parties – three justices observed that independent expenditures could result in “personal attacks or slandering the opposing candidates by spreading false information”, particularly when done anonymously on-line. This would, they said, disturb “the tranquility and fairness of the election.”Footnote 32
The Constitutional Court has decided only a handful cases challenging the NEC's regulatory decisions, but two of them are quite dramatic. Both involved President Roh Moo-Hyun. Roh, the candidate of the liberal Millennium Democratic Party (MDP), won the presidency in December 2002.Footnote 33 Though reasonably popular himself, Roh faced large political problems. The conservative Grand National Party (GNP) had a majority in the National Assembly, and the MDP itself was internally divided with Roh representing a younger insurgency within the party. Roh's supporters left the party in September 2003 to form the Uri Party, and the old guard formed an informal anti-Roh coalition with the GNP. On the policy level, Roh faced a weak economy and suspicion from the United States that he was likely to be an unreliable ally.
Elections for the National Assembly were scheduled for April 2004, and Roh hoped that its outcome would strengthen his political position by installing a substantial number of his supporters in the new Uri Party in the Assembly. Following up on an earlier speech urging people to support his party, in February 2004 Roh gave a speech in which he urged voters to support the Uri Party, saying that he “would like to do anything that is legal if it may lead to votes for the Uri Party.” Four days later the MDP filed a complaint with the NEC alleging that Roh's speech violated a statute requiring that the President be impartial in elections. Within a week the NEC sent a letter to Roh “requesting” that he remain neutral. A week after that Roh held a press conference in which he refused to apologize for making the speech and said that he disagreed with the NEC's substantive conclusion about his duties as President: “I would like to make it clear that the decision of the National Election Commission at this time is not convincing”, and was a continuation of pre-1987 practices in which the government “mobilized… the state institutions” to produce election results it favored.
His opponents then impeached Roh, citing his defiance of the NEC as one of many grounds; the precise formulation was that in refusing to do what the NEC asked Roh had failed to protect the constitutional order. Roh sought review in the Constitutional Court, which had the power, express in the Constitution, to review impeachments. Roh was suspended from office while the case was pending, but the April elections went forward. The Uri Party won a massive victory, gaining 105 seats, far more than the 47 it had held in the prior Assembly, amounting to an absolute majority; the MDP, Roh's former party, went from 63 to 9 seats.
In a classic “megapolitical” decision the Constitutional Court held that Roh had indeed violated his constitutional obligations, but that the violations were not serious enough to warrant his removal from office.Footnote 34 The opinion discussed all the grounds the National Assembly had cited in impeaching Roh, but here I focus only on the charge related to the NEC's letter because that charge involved a decision the NEC made, not the Constitutional Court. The Court's analysis began with a discussion of the duty of public officials to be neutral in elections, which, it said, followed from the “principle of free election … that the voters should be able to make their own judgment and decisions in a free and open process….” The President was a public official covered by this principle because he was “in a position to threaten” the free-election principle; in this he was different from legislators, “from whom political neutrality concerning elections cannot be requested” because they are “active figures at the electoral campaign.” Note that this suggests that Roh's mistake was to express an opinion about who should win the upcoming election for the National Assembly (and note as well that no President would ever be in a position to ‘influence” his or her own election campaign because Korean presidents serve a single nonrenewable six-year term).
The Court acknowledged the President was a member – indeed, the leading member – of a political party. And, the Court agreed, the President could continue to act within the party, for example by participating in a party's convention. But, though elected as a party member, once in office the President doesn't “implement the policies of the ruling party”, but “is obligated to serve and realize the public interest”, which presumably must be defined as distinct from the party platform on which the President ran. “The President is obligated to unify the social community by serving the entire population beyond that segment of the population supporting him or her.” Turning to the “feverous competition” among parties in legislative campaigns, the Court held that a presidential statement “unilaterally supporting a particular political party” necessarily “distorts the process of the independent formation of the public's opinions based on a just evaluation of the political parties and the candidates.” That competition “is significantly perverted by one-sided intervention of the President supporting a particular party.”
The Court agreed with the National Assembly that Roh's response to the NEC request violated his duty to protect and defend the Constitution. Of course Roh could criticize the underlying law on which the NEC relied, and could seek its repeal or amendment. But “questioning the constitutionality of [the] statute itself in front of the national public constitutes a violation of the President's obligation to protect the Constitution.” Such statements “might have significantly negative influence on the realization of a government by the rule of law, … by lowering the public's awareness to abide by the law.” Note here that this is an extremely strong rejection of the view that each branch of government is entitled to decide for itself on a statute's constitutionality, subject only to an ultimate duty to comply with judicial holdings. And this anti-departmentalist stance implicitly assimilates the NEC to the judicial branch, because the NEC's request plainly rested upon a contestable judgment about the constitutionality of the “political neutrality” statute's applicability to the presidency.
But, after all this, the Court allowed Roh to remain in office because his constitutional violations were not “grave” enough to justify removal. His statements were “unaggressive, passive, and incidental, during the course of expressing the president's political belief or policy design in the form of a response to the question posed by the reporters at a press conference.” And, returning to a theme it had briefly addressed, the Court noted the “blurred” boundary between permissible presidential actions within his party-political role and impermissible statements violating the duty of political neutrality.
Returned to office, Roh was not done with conflicts with the NEC. Roughly six months before his term ended in 2008, he again got into trouble with the NEC. This time a presidential election was in prospect (held in December 2007). Roh made a largely ceremonial speech to a public forum, in the course of which he said that “it will be a problem if foreign newspapers comment that the Korean leader is the daughter of a dictator [referring to Park Geun-hye, who narrowly lost the GNP primary election in several weeks later].” He also said that the GNP “is an irresponsible party.” A few days later Roh received an honorary degree from a Korean university and delivered an address nominally on democracy. In the speech he criticized proposals put forth by the GNP's presidential candidate. Two days later he offered “congratulatory remarks” at a celebration of the anniversary of one of the 1987 protests that led to the establishment of a democratic government in Korea. Among his comments were criticisms of his opponents, “those who were in power in the past allied with conservative press.” Within days of each speech the NEC met and reviewed Roh's statements, finding in each case that he had violated the duty of political neutrality. It sent him “notifications” of its conclusion that he had “defamed the opposition party and its potential presidential candidate,” and “advised” him “to abstain from making any speech which may influence elections.” Roh sought review in the Constitutional Court, arguing that the NEC's notices violated his right to free expression.
Again the Court found against Roh.Footnote 35 After disposing of the objection that, as mere “notices,” the NEC's letters were not exercises of public power that could be unconstitutional, the Court turned to the merits. In a section headed, “President as a politician,” the five-justice majority opinion devoted somewhat more space here than the Court had in 2004 to the real problem posed by the NEC's position. The discussion opened with the sensible observation, “Modern democracy has changed from representative democracy to party politics democracy.” Parties “shape political ideas and influence state policies.”Footnote 36 And, having been nominated by a party and winning as a party candidate, the President could continue to participate in party affairs. All this simply restated the analysis from 2004. But now the majority acknowledged openly that the President “is likely to be closely related to the policy and interests of a certain political group”, which created the possibility of a conflict between the President's “freedom of political activity” and his duty of impartiality in elections. The President “not only executes his political party's policy but also owes the duty to promote public goods” and must “serve all the people.” But, when a conflict arose, the duty of impartiality prevails.
The difficulty with this analysis of course is that, from the viewpoint of the President considered as party leader, the party's policies are the “public goods” that he or she must promote in his or her capacity as President. The opinion hints at a solution by suggesting that officials charged with the on-the-ground administration of elections might be influenced by, for example, Roh's statements criticizing the GNP: the statements might matter “because public officials tend to consider the political orientation of the President who supervises the personnel management although their employment is guaranteed under the law.” Exactly what this means is obscure, although it might be hinting at the possibility that civil servants will exercise their discretion to tilt their services in favor of the President's position and against the opposition's. Yet, when we combine civil service protections with the fact that a president can serve only one term and the fact that presidential statements directed at elections have to be made in some temporal proximity to an election, this risk seems quite small.Footnote 37
Taken on its face, the Constitutional Court's image of politics blends two ideas. One is express: that voters should choose their representatives based upon an unbiased evaluation of the candidates’ position – phrases like “free and open process” and “truly free decisions” recur in the opinions. The other is implicit but clearly present: that voters will be “unduly” deferential to statements made by the President, who – again implicitly but clearly – they believe represents the nation as a whole. And the latter is true even though voters of course know that the President was elected in a competitive party-based election. The decisions discussed here seem to be seeking to impose a degree of orderliness on the messiness of politics in such a world.
Of course the two decisions on which I have focused dealt with a single political figure, who perhaps fell outside the boundaries of what the legal and electoral specialists on the two institutions believed to be permissible (roughly centrist) politics. Notably, the membership of the NEC when it acted against Roh consisted almost entirely of appointees of officials in place before Roh became president (appointees of the prior president and the chief justice). And Roh's governing style of appealing directly the people was in tension with the dominant elitist style of governance that seeks to make politics “orderly.” It is not that the NEC membership was unfamiliar with politics, but rather that most of its members had a specific view of how politics should be conducted, a view that conflicted with Roh's governing style.Footnote 38
That said, neither the NEC nor the Constitutional Court seem to have a sensible account of a president's duties (or, more generally, of the idea of electoral fairness) in a party-political world. Recall the conceptual and functional cases for IPDs: They are needed because Madisonian mechanisms are inadequate to ensure the stability of democratic institutions in a party-political world, and they do so by deploying relatively less political forms of expertise. The institutions’ difficulties in coming up with a plausible account of the President's duty of electoral fairness suggests that IPDs might not be an adequate supplement to Madisonian mechanisms.
Conclusion: the ambiguities in evaluating IPDs’ success
This Article has described the conceptual and functional logics of fourth branch institutions. Those logics can help us understand what the institutions might do if they function well. But the logics have only a loose relation to the actual performance of their functions. It has also offered two brief case studies illustrating the ways in which an independent electoral management body can seek to impose its vision of what a healthy political order looks like. That suggests that IPDs might not be “above” politics, but rather might have their own form of politics, which deserves careful analysis. Additional studies of electoral management bodies and other IPDs might probe that suggestion more fully.
