DELIBERATIVE CONSTITUTIONAL REFERENDUMS IN DEEPLY DIVIDED SOCIETIES.

• Author: O'Flynn, Ian
• Position: ARTICLES

INTRODUCTION

While all societies are divided to one extent or another, the prefix "deeply" typically denotes a divided society in which national, ethnic, religious, linguistic, or other divisions are severe enough to threaten the very nature or existence of the state. Clearly, this standard definition is extremely broad. Since it covers vastly different cases, it is normally refined for research purposes. In practice, scholars more or less consciously narrow the definition in ways that help them to identify and illuminate a specific class of cases. (1)

The most common approach is to focus on cases where the main divisions in society are ethnic in kind and where those divisions engender civil violence. (2) By contrast, our concern in this article is with societies that are seeking to manage their divisions by democratic means but where the consent of at least one major group or segment is still in question. While our approach also differs methodologically in that it is normative rather than comparative, cases that illustrate our core concerns include Colombia, Cyprus, Northern Ireland, and South Africa. Cases that do not involve a serious attempt to manage their divisions by democratic means and that therefore do not fall (at least for now) within our focus include Kashmir, Mindanao, Syria, and Yemen.

In theory, the links between democracy and consent are obvious enough; democracy is rule by the people. But since the sheer scale of most modern democracies is such that the people cannot rule themselves directly, there must be a set of fundamental rules or a constitutional settlement by which representatives govern on their behalf. The caveat is that a new constitutional settlement will be regarded as democratically legitimate only insofar as it has the consent of the people.

At a general level, we take it that there is nothing particularly controversial about this understanding of the relationship between democracy and consent. Yet, since democracy and consent can take very different forms in practice, scholars need to argue for their preferred conceptions. In this article, we contend that people should signal their consent for a new constitutional settlement through a plebiscite or referendum--that is, through a vote of the entire electorate intended to register the balance of support for the settlement. (3)

Granted, referendums in deeply divided societies have a mixed track record. One can certainly point to relatively successful cases: the 1992 referendum on ending apartheid in South Africa or the 1998 referendum on power-sharing in Northern Ireland. Yet one can also think of failures: the rejection of the UN-sponsored 2004 Annan Plan for reunifying Cyprus or, more recently, the rejection of a peace deal between the Colombian government and the Revolutionary Armed Forces of Columbia (FARC) guerrillas in 2016. (4) It is not always easy to say what makes the difference from case to case or, indeed, to predict what might happen next. However, the assumption on which this article rests is that the quality of the deliberation that occurs during the course of the referendum is likely to have an important bearing on the referendum's chances of success--that is, its prospects of generating democratic consent for a new constitutional settlement.

Deliberation is a feature of all conceptions of democracy, including competitive party models. (5) Ye t the conception of democracy that informs our thinking in this article is "deliberative democracy". (6) As the name suggests, deliberative democracy is a conception of democracy that reserves a central place for public reasoning about important matters of law and public policy. Importantly, the members of such a democracy do not seek to impose their competing views on one another through, for example, the strategic force of numbers or simple brute force. Rather, each side seeks to convince the other that the better arguments are on its side. They do so because, from a deliberative view, "outcomes are democratically legitimate if and only if they could be the object of a free and reasoned agreement among equals." (7)

There are many reasons to value deliberative democracy. (8) Yet in the sorts of deeply divided societies with which we are concerned in this article, one reason stands out as especially pressing: constitutional settlements usually require major compromises--generally speaking, each side will have to make painful concessions to the other's constitutional vision of the state. For example, the majority may have to accept that power will be shared, while, for its part, the minority may have to give up its claim to independent statehood. As a consequence, constitutional settlements usually require considerable forbearance--typically, each party to the compromise will have to patiently endure the criticisms others level at it for making concessions. More precisely, in deeply divided societies, the moderates who negotiate a settlement must endure the criticisms levelled at them from the extremes. Ye t insofar as moderates are able to explain to voters during the course of the referendum why they acted as they did--why moderation served their purposes better than extremism--the settlement that they reach may have a better chance of taking root and enduring.

Of course, moderation is a long-established theme in the literature on deeply divided societies. For example, Donald Horowitz argues that the alternative vote electoral system, and in particular the high threshold for election and preferential nature of its ballot, may encourage political parties from one group or segment to look for votes from another. Since lower-order preferences may make all the difference between winning and losing seats, parties have a real incentive to moderate their claims in the hope of broadening their appeal. (9) Yet while we agree that electoral systems can be designed to encourage moderation, what is novel or distinctive about our approach is its emphasis on deliberation. With specific reference to the referendum, we favour not just electoral engineering, but deliberative electoral engineering.

Admittedly, the idea of a deliberative referendum--a referendum purposefully designed with deliberative principles in mind--is not new. (10) Nor is the idea of applying deliberative democracy to divided societies. (11) Ye t to date, very little has been written on the application of the notion of a deliberative referendum to deeply divided societies. (12) Even less--if anything at all--has been written about the idea of using the deliberative referendum as a driver of moderation in the context of generating popular consent for a new constitutional settlement.

Before proceeding with this argument, one obvious practical objection needs to be confronted. On the face of it, deeply divided societies would seem to be infertile grounds for deliberation. In deliberation, as noted, each party seeks to convince the other that the better arguments are on its side. To that end, they invoke considerations and exchange reasons in an endeavour to arrive at an agreed judgment or a shared view. (13) Ye t in a deeply divided society, it may be difficult even to get people on opposing sides in the same room. And if they do meet, they may not really deliberate: they may fail to listen to one another with an open mind or reflect seriously on what others have to say. Since trust is likely to be low, they may simply regard one another's reasons as insincere cover for sectional interests. As Brendan O'Leary pointedly remarks, "those who embrace a politics of deliberative democracy as the prescription for conflict need reminding that deliberation takes place in languages, dialects, accents, and ethnically toned voices and that it is not possible to create 'ideal speech situations.'" (14) There may, for instance, be "enclave deliberation" in which most of the discussion occurs among people on the same side. (15) Enclave deliberation is likely only to deepen the lines of division, as Cass Sunstein's work on polarization suggests. (16)

For analysts such as O'Leary, democracy in deeply divided societies is best handled by political leaders or other elites (17) who are both willing to take tough decisions and strong enough to bring their supporters along with them in implementing those decisions. (18) Ye t if political leaders, who also speak "in languages, dialects, accents, and ethnically toned voices," (19) can negotiate within appropriately structured institutions, why can't ordinary people deliberate, also within appropriately structured institutions?

As we indicated above, the assumption on which this paper rests is that deliberation matters to a referendum's chances of success. More specifically, a peace referendum designed with deliberative principles in mind may stand a better chance than one that does not generate popular consent for a new constitutional settlement. Crucially, however, this implies that the quality of the deliberation that takes place during the course of a referendum campaign will itself be affected by the design of the referendum. Put another way, while deliberative democracy may guide the design of the referendum, the referendum should in turn make meaningful deliberation possible. It will provide an institutional framework within which deliberation can occur. As Joshua Cohen puts it, the members of a deliberative democracy share "a commitment to co-ordinating their activities within institutions that make deliberation possible and according to norms that they arrive at through their deliberation." (20) Theory and practice should be seen as mutually implicating.

To explore these considerations in greater detail, we proceed as follows. In the first section, we offer three interrelated arguments in support of the use of referendums in deeply divided societies. First, since constitutional settlements have far-reaching implications--implications that may extend to almost every corner of...