Recent Innovations in Making Constitutional Law
I turn now to recent innovations in constitution-making processes. The first is a sharp increase in practices of popular participation in constitution-drafting. As the twentieth century waned, popular participation in ratifying constitutions, typically through referenda, increased, though it never completely displaced older modes of ratification through parliamentary approval.
This century’s innovation is to move high levels of popular participation back to the stage of constitutional drafting. Among other things, that innovation reduces the effects of upstream constraints on the selection of constitution drafters, although, as we will see in connection with Iceland, the weakening of that constraint may have downstream effects.The second innovation is greater openness in constitution-making by constitutional courts. This openness includes televising hearings in constitutional courts and the use of press offices by such courts, to notify the media of important decisions in advance and to offer court-endorsed summaries and explanations of those decisions. Here I deal with the Brazilian practice of “public hearings,” a special form of hearing arguments in cases selected by the Supreme Federal Court for such hearings. Here the secrecy of deliberations is weakened.[247]
The remainder of this Essay presents brief case studies of the recent Icelandic constitutional revision process and the Brazilian public hearing process. I do not claim anything like comprehensiveness, but I hope that my sketches will bring out features of the innovations of which they are examples that deserve closer study, both in connection with Iceland and Brazil and in connection with constitutionmaking more generally.
8.3.1 Iceland's Failed Exercise in Crowd-Sourced Constitution Making
Iceland’s financial crisis in 2008 led to widespread discontent with the nation’s political class.[248] The coalition government that melded the Social Democratic Alliance and the Independence Party fell apart.
The Social Democratic Alliance formed a minority government, which continued to govern after an election in 2009. That government proposed to initiate a process of constitutional reform; the Independence Party opposed the idea. At first the idea was that an elected constituent assembly would propose a constitution that would become legally effective after approval in a referendum without further parliamentary participation. After being advised that that course would require the adoption of a prior constitutional amendment modifying the permissible methods of constitutional amendment, the proposal was reshaped. In 2010 the parliament adopted an Act stating that a Constitutional Assembly would convene in February 2011 and finish its work by April of that year. To gain the Independence Party’s support, the proposal was that a Constitutional Committee appointed by Parliament would do preparatory work for the Constitutional Assembly; the Committee’s seven members included academic experts in law, science, and literature. The thought was that this procedure would lead to a set of constitutional amendments channeled through the regular amendment process - simple majority votes with one taken before and the other taken after an election, though by convention more than a majority was required.The Constitutional Assembly described in the 2010 statute never convened. One difficulty was that the process was elite-dominated and, perhaps more important, was the creature of the very political parties that had been brought into disrepute by the financial crisis. To offset that elitism, the Constitutional Committee convened a National Forum of 950 citizens drawn at random from the census list. The Forum met for one day in November 2010 and discussed basic constitutional principles, such as “one person one vote” - an important issue in a small nation with a population concentrated in the capital city, which in a one-person-one-vote system might disregard the interests of rural voters - and rights to natural resources.
As to the latter issue, some Icelanders believed that private control of fishery resources had created a network of interest groups that in turn created a culture of financial corruption that led to the crisis (Gylfason 2014).The Constitutional Committee then organized an election for members of a constitutional assembly. Candidates were basically self-nominated but had to get signatures to place themselves on the ballot. Twenty-five members were to be elected, from a group of 522 candidates, through a single transferable vote system that allowed each voter to vote for up to the full complement of twenty-five. The Independence Party and the Progressive Party, two major political parties in the pre-crisis period, opposed the process as a whole, and did not vigorously participate in supporting candidates.
Iceland’s Supreme Court voided the elections for the Constitutional Assembly: it found that the process was flawed because the secrecy of the ballot was threatened by, for example, numbering ballot papers, although there was no evidence that anyone had actually tried to link a voter with his or her ballot after the event. Instead of taking over the constitution-amending process itself or running another set of elections (the latter option was ruled out in part because of the cost of running another election), Parliament designated the twenty-five winners of the election as members of a constituent body - called the Constitutional Council - whose authority derived, at least in theory, from parliament rather than from the people directly.
With an initial deadline two months away, later extended by another two months, the Constitutional Council then got to work. Its members were drawn from a number of professions, and had varying political affiliations, but none were important figures in or seen as representatives of the (discredited) political parties. They broke up into three working groups and, importantly, solicited comments and suggestions from every Icelander by establishing a web-site and social media accounts to which suggestions could be posted.
This practice was the basis for describing the drafting process as “crowd-sourced.” The Council received about 3600 comments as well as 370 “formal suggestions” - not a trivial number in a nation with a population of under 400,000.[249] [250] Although most of the posts were generic, a fair number suggested substantive provisions for inclusion in the draft constitution. One member of the Constitutional Council stated that the Council members deliberately refrained from setting up special meetings with “representatives of interest organizations.”11After public deliberations and by a unanimous vote, the Constitutional Council adopted the draft constitution, anticipating its initial ratification (or disapproval) in a popular referendum. With respect to six specific matters, including “one person one vote,” the possibility of national referenda upon citizen demand, a state religion, and ownership of natural resources, the Constitutional Council presented a “yes or no” option.
The referendum, though, would not be the final step even if the voters approved the new constitution. Instead, in light of the Constitutional Council’s origins in parliamentary action, parliament would have to approve the constitution after the referendum. In November 2012 voters did approve the draft constitution by a margin of 67-33 (and chose which of the options they wanted on each separately identified issue), but the 49 % turnout was smaller than many expected or hoped, and the margin of approval was similarly smaller than expected or hoped. After the referendum, parliamentary leaders invited the Venice Commission to provide comments on the draft constitution, nominally to guide the parliament’s decision on approving or disapproving the document; the comments found various technical deficiencies in the draft, suggesting that clearer language could have been used on some matters and asserting that some important issues had not been resolved. There is some reason to think that these technical problems resulted from a combination of the strict time limits under which the Constitutional Council worked, with the relative inexperience of the Council’s members in politics generally and in constitutional design in particular.[251] The draft was modified to deal with the Venice Commission’s comments, without objection from the members of the now-dissolved Constitutional Council.
Parliament never took up the proposed constitution, so it did not go into effect. Throughout, the established political parties held themselves at arms’ length from the process, in part because they had been discredited but in larger part because their leaders disagreed with the idea that a totally new constitution had to be adopted, rather than discrete amendments that could have been adopted by the ordinary process of parliamentary vote.
In no particular order, here are some features of the overall Icelandic process.
1. The financial crisis provided the impetus for the constitution-making process. That process took time, with the referendum held three years after the crisis’s peak and as Iceland was clearly on a path to recovery. The Supreme Court’s intervention required that the process be restructured and seems to have had something of a disruptive psychological effect on the Constitutional Council’s members, who nonetheless did complete their work within a four-month period. The passing of time, coupled with a degree of recovery from the financial crisis, reduced the felt urgency of constitutional reform.[252] The default was the existing constitution, which might have seemed “good enough” as reform pressure waned. Elster argues that constitutions are often made in times of crisis, which leads them to be less well-designed than would be the case were the drafting process to be extended over time. But, he observes, when the public does not see the polity as facing a crisis it may lose interest in constitutional design. That seems to have been true in Iceland.
2. The Constitutional Council did not have international advisers as such. The Venice Commission’s intervention after the draft constitution’s approval in the referndum did bring an outside and purportedly technical perspective to the prereferendum discussions. The Venice Commission’s views probably reduced enthusiasm for the document. To some degree, in its role as commenter on constitutional amendments and revisions the Venice Commission exists to identify problems.
Sometimes doing so leads to improvements in the documents upon which the Commission comments, but only if the domestic process can accommodate the comments by revising drafts and proposals. The document was revised, but the Venice Commission’s comments probably cast a modest adverse light on the process that produced the proposal it evaluated.I think it worth noting that some of the Venice Commission’s critical comments identified relatively minor flaws in the draft, such as areas where terminology was unclear or where obvious issues were left unresolved. Crowd-sourcing constitutional drafting, or using a true citizen (nonprofessional) constituent assembly, will almost certainly generate documents with these sorts of flaws, and perhaps outside advisers and commenters should modify their practices when constitutions are drafted with extremely high levels of popular participation.[253]
3. Crowd-sourcing, it might be thought, would be a process in which those who participate in “ordinary” politics at low rates would be enabled to increase their level of participation. Even in a nation as small as Iceland, going to parliament to testify, or even writing a letter to a member of parliament, is likely to be more difficult than going on-line to submit a comment or proposal. Whether crowdsourcing had that effect in Iceland is as yet unclear. One early study suggests that the constitution-drafting process was less demographically representative than the ordinary legislative process (Helgdottir. 2014). The study examined the relative participation of men and women in the two processes. It found that, relative to the ordinary parliamentary process, men were overrepresented in the social media comments in the crowd-sourcing process. The results are suggestive, but not nearly definitive. For one thing, representation in the ordinary legislative process is mediated through civil society organizations, whereas interventions on social media need not be. That is, it is possible (as the study’s authors acknowledge) that different subgroups of women and men are represented in the ordinary legislative process and the crowd-sourcing one.
4. The Icelandic process was as open as can be. According to Elster, then, it should not have presented opportunities for bargaining, and that appears to have been the case. Elster’s concern that participants in an open process will posture rather than deliberate, though, seems not to have been realized, at least on the accounts currently available in English. Perhaps the reason is that participants were true “one-shotters.”[254] That is, not only were they not politicians who might be concerned about appealing to external audiences for future support, but they were complete amateurs (there were eight academics, in fields that included economics, political science, and philosophy, but also mathematics and
theology - and no legal academics) with no continuing interest in implementing the constitution they drafted.[255]
Relying on the so-called “self-denying ordinance” of the French Constituent Assembly, which barred participants from office under the constitution they were to create, Elster points out that constitutions drafted by one-shotters may be defective because the drafters have no continuing responsibility for the actual operation of the government they are creating. They may adopt provisions that seem in principle desirable, but need not worry about whether the provisions will work well in practice. Because the Icelandic constitution did not go into effect, we cannot know whether Elster’s concern would have been realized under it. But, the “one-shotter” concern is related to another, as to which the Icelandic experience is instructive.
5. Continental constitutional theory may have mattered as well. A directly elected Constitutional Assembly could have been seen as a true constituent assembly, speaking for the people as whole without its actions being mediated through preexisting political institutions. As a constituent assembly its actions would be those of the people, and - given modern practices - the referendum endorsing the new constitution would have similarly been a direct act of the constituent power. The theoretical picture changed when the Supreme Court invalidated the election of members to the Constitutional Council. Probably out of a desire to keep the process moving and not for theoretical reasons, parliament appointed the winners to the Constitutional Council. But parliamentary appointment broke the direct connection between the people and the Constitutional Council’s members. They became the recipients of authority delegated to them by parliament, and, again as a matter of continental constitutional theory their principal - the Parliament - had to approve of what they did.
Using Elster’s terms, we can say had the Constitutional Council been a constituent assembly, the only downstream constraint would have been the need to obtain popular ratification of the draft constitution. The substitution of the parliament for the people as the source of the Constitutional Council’s authority meant that ratification by the parliament, and therefore support by a decent share of the nation’s political leadership, became a downstream constraint. But, the parliament’s designation of the winners in the voided election as the members of the Constitutional Council meant that satisfying the requirements of the nation’s political leaders had not operated as an upstream constraint on the body’s composition. Under the circumstances, perhaps, failure was quite predictable.
6. Iceland’s political parties did not participate in the drafting process, and indeed the random selection, self-nomination, and other processes for selecting members of the Constitutional Assembly/Council almost guaranteed their exclusion.[256] Then, at the final stage in the process, the parties in parliament defeated the constitution. Elster’s terminology is not exactly apt in this instance, but the intuition behind it, is. Party participation was in fact an upstream constraint on the drafting process, although it was not understood to be such at the outset. The outcome was failure attributable precisely to the fact that this upstream “constraint” was ignored.[257]
Suppose, though, that the parties had been included, to a degree, in the drafting process. Then the problems of posturing (by participants with long-term interests persisting after the constituent assembly ends) and of impediments to bargaining in an open process might have occurred. And, of course, their participation would have tempered those portions of the reform agenda aimed at the features of the precrisis political system that were thought (by some) to have produced the crisis. Accommodating the political parties upstream might then have reduced the probability that the resulting proposal would receive downstream endorsement in a referendum.
The general lesson of these observations, it seems to me, is that a program of increasing the level of public participation in constitution-making might have some attractions from a democratic point of view, and might seem achievable with modern technology, and yet implementing that program calls for quite careful thought and attention to the questions of bargaining, deliberation, and constraint to which Elster directs our attention. It probably was not inevitable that Iceland’s process would fail, but now that we have seen its failure, we might be able to identify some “red flags” that those who seek to implement similar processes should direct their attention to.
8.3.2 The Brazilian Public Hearings
Brazil’s Federal Supreme Court holds its deliberations in public, and there is a television station dedicated to broadcasting its proceedings. In addition, the Court is authorized to hold “public hearings,” which are different from ordinary oral arguments in ending cases. When cases arrive at the Court, one justice is assigned responsibility for the dossier. The rapporteur has a discretionary power to call for a public hearing in two circumstances. When the case is a “direct action” on constitutionality - a proceeding filed in the Federal Supreme Court in the first instance, without any lower court proceedings, challenging the constitutionality of a statute - the rapporteur may do so if the record is incomplete with respect to some important facts; when the case is a general claim of unconstitutionality, the rapporteur may do so apparently without restriction. Participants in the public hearing are defined by statute as those with “experience with and authority on” the question at issue.
The statute authorizing public hearings was enacted in 1999; the first such hearing was held in 2007, with a total of fifteen through mid-2014.[258] Among other topics, the constitutional challenges dealt with laws on stem-cell research, on the possibility of terminating pregnancies of anencephalic fetuses, on affirmative action in the form of strong quotas, and on banning the importation of used automobile tires. The number of participants has been reasonably large, ranging from 10 to more than 50. The rapporteur generally has divided the participants into two groups, those favoring a finding of constitutionality and those favoring one of unconstitutionality. The participants have been drawn widely from civil society, typically through ordinary civil society organizations.[259]
I begin the analytic portion of this section by distinguishing the public hearings from the U.S. amicus curiae practice. Public hearings do resemble the amicus curiae practice because they allow interested parties to present their views to the court. They differ, though, because in the amicus curiae practice the presentations are almost entirely in writing; rarely the Court will allow one amicus curiae to participate in the oral argument, and never more than one or two. In contrast, the Brazilian public hearings involve in-person presentations by a large number of interested participants.
A study of Brazilian right-to-health-care cases illustrates the public hearing process (Wang 2013, 75). The Supreme Federal Court had considered a large number of such cases prior to 2009, issuing decisions that began to lay out the contours of a constitutionally permissible program for allocating health care. In 2009 it convened a public hearing, which led to a set of decisions in 2010 setting out criteria for allocating health care in a manner consistent with the Constitution. According to Daniel Wang, “These decisions establish a comprehensive set of criteria that present a more refined and realistic interpretation of the right to health than was exhibited in previous BFSC case law” (Wang 2013, 82). The criteria, while sensible, do not establish a fully comprehensive approach to health care because they do not, at least directly, confront questions about rationing that result from mandatory allocations to one population in situations where other populations receive discretionary allocations. Still, the study suggests that there is reason to think that the public hearing improved the quality of the Court’s jurisprudence.
Again, some analytic points in no particular order:
1. The public hearings are quasi-legislative in character. The rapporteur who issued the first call for a public hearing relied on parliamentary by-laws dealing with legislative hearings for the procedures to be used in the public hearings (Hennig Leal 2014, 9). They are quasi-legislative in substance as well as procedure. The non-judges who participate present the entire range of arguments bearing on the constitutional question before the court. These include analysis of relevant policy considerations and the relation between sometimes contested facts and constitutional interpretation. One might analogize them to a hearing before a constitutionally responsible legislative committee devoted to legislative consideration of the constitutionality of a specific proposal.
In the United States, legislative hearings are often quite scripted, confirming Elster’s sense that public “deliberations” lead to posturing rather than true deliberation; research on whether this is true of the Brazilian public hearings would be valuable. Note, though, that even such hearings are different from the debates in a public constituent assembly. In the constituent assembly all participants are members of the body, and, after the assembly opens, they may quickly stake out positions and become aware of the positions firmly adhered to by others. The “in person” nature of the discussions then might not overcome the scripted posturing. In contrast, legislative hearings and the Brazilian public hearings involve “repeat players” on one side - the legislators or the judges - but, typically, “one- shotters” on the other. It may be that social norms dealing with respect in inperson conversations will induce a somewhat more genuine practice of deliberation in the legislative hearings and the Brazilian public hearings.
2. The quasi-legislative character of the public hearings can be taken to reflect a Kelsenian understanding of constitutional interpretation. I take Kelsen to argue that constitutional interpretation is a complex blend of law and politics. Such an understanding accounts for the structure of the Kelsenian constitutional court in a civil law tradition: those sitting on the constitutional court could not be exclusively drawn from the career judiciary, because career judges, while perhaps talented in doing law, would not have adequate experience in politics. The Kelsenian court is selected outside the ordinary civil-law processes of judicial selection so that some, perhaps all, members will have some facility in blending legal analysis with political sensitivity. That is what happens, at the argument stage, with the Brazilian public hearing as well: participants drawn from civil- society organizations bring something more than a perspective on law alone to the discussions.
3. More recently, scholars have begun to distinguish reasonably sharply between legal (better, “judicial”) and political constitutionalism. Legal or judicial constitutionalism lodges final and primary responsibility for constitutional interpretation in the courts; legislators and executive officials may, but need not, take constitutional considerations into account as they act, but the conclusions they reach about constitutional meaning can always be displaced by the judges’ contrary conclusions. Political constitutionalism, in contrast, gives legislators and executive officials a large and honored place in constitutional interpretation, and in some versions give them the final word. The Brazilian public hearings can be understood as blending political and judicial constitutionalism. The hearings are before a court, which has the final word on constitutional interpretation. But, the hearings can involve a large number of civil society organizations offering their views on constitutional interpretation, which can be understood as related to the general practice of political constitutionalism. In addition, ministers from the executive government both attend and participate in the public hearings, making them a locus where the executive government and the judiciary receive simultaneous input from the public on the constitutional issues being examined.
4. An interesting strand in recent U.S. constitutional scholarship examines the ways in which social movements - a subcategory of civil society - affect the development of constitutional law.[260] One mechanism by which they do is straight-forward: Once they gain sufficient force, social movements influence the composition of the courts. Politicians satisfy a movement’s supporters by appointing judges sympathetic to the movement’s constitutional views to the courts, and then, once on the courts, those judges interpret the constitution as incorporating those views. Sometimes, though, it seems that social movements affect constitutional interpretation without having influenced judicial selection. In the United States the most dramatic examples are recent: The Supreme Court adopted constitutional interpretations consistent with the views of the women’s rights and gay rights movements at times when those movements were socially significant but well after the justices deciding the cases had been appointed to the Court. The mechanism by which this occurs remains obscure. Public hearings in the Federal Supreme Court are formal mechanisms by which the views of contemporary civil society can be brought into the court’s deliberations.
Monica Clarissa Hennig Leal describes the public hearing as a mechanism that advances the “openness and democratization” of the judiciary in a constitutional system that gives the judiciary the final word on constitutionality. The mechanism, for her, is one way in which constitutional interpretation is itself democratized. The finality of judicial interpretations depends in part on the strength of the amendment rules in place. The Brazilian amendment rule relatively lenient: an amendment becomes effective if approved by three-fifths of both houses in two readings. Put another way, the Brazilian Constitution is already a reasonably open and participatory one. Public hearings in the Federal Supreme Court may reflect, but also enhance, that characteristic.
8.4