Blog of the International Journal of Constitutional Law

Book Review/Response: Stephen Tierney and Peter Oliver on Constitutional Referendums

[Editor’s Note: In this installment of I•CONnect’s Book Review/Response Series, Peter Oliver reviews Stephen Tierney’s recent book Constitutional Referendums: The Theory and Practice of Republican Deliberation, just released in paperback. Stephen Tierney then responds to the review.]

Review by Peter Oliver

Peter Oliver, Vice Dean, Full Professor and Member of the Public Law Group, Faculty of Law, University of Ottawa, reviewing Stephen Tierney, Constitutional Referendums: The Theory and Practice of Republican Deliberation (Oxford 2012)

Constitutional commentators are divided on whether constitutional referendums are a good thing. With his characteristic clarity and thoroughness, Stephen Tierney helps us to determine when and why constitutional referendums are appropriate and how they should be conducted in the circumstances.

In this review, I would like to comment on three elements in the book: theoretical premises; practical concerns; and unexplored possibilities.

Stephen addresses theoretical premises in the opening chapter. He describes the book’s framework as one of “constitutional theory”,[1] a statement which is potentially confusing, in that it suggests a greater consensus regarding the meaning of that term than exists.  Could it ever be said that constitutional theory yields a single, agreed upon framework? Is it likely that a Dworkinian such as T.R.S. Allan would approach referendums in the same way as, say, Martin Loughlin, simply on the basis that they share the perspective of constitutional theory?  No one, including Stephen, would think so.  But this is because, rather than assuming that constitutional theory is univocal, Stephen adopts a particular version of constitutional theory that is set out early on[2] but is never really defended against rival constitutional theoretical positions.

Stephen sets out his own view most clearly towards the end of the book, when it become apparent that “constitutional theory” for him is not a description of a shared theoretical perspective, but an approach which employs elements of normative and functionalist branches of constitutional theory, with a clear tilt toward the latter:[3]

[M]y approach has been one of “functional normativity”; a model of constitutional ethics, rather than morality.  Normative presuppositions are inherent within any exercise of constitutionalist analysis, but at the same time the constitutional theorist’s role, while alive to this normative dimension, must remain grounded in functional possibility and the suitability of constitutional prescription to the political environment that the constitution is modeled to serve.  Constitutional theory must engage the real world or political practice, and in particular any normative turn, must, to be meaningful, address feasible improvements in actual constitutional institutions and processes.

This is an approach with which I have much sympathy.  However, given the existence of constitutional theory based on different premises, much of it still inspired by Ronald Dworkin’s writings, it would have been interesting to learn more about why Stephen rejects other theoretical frameworks, and why, for example, constitutional theory is “not a Platonic quest from original normative principles”.[4]

While the reader does not get the benefit of a more elaborate discussion of constitutional theory, which in Stephen’s hands would certainly have been particularly fascinating, the theoretical framework for the book is set out in basic terms early on.  The following paragraph provides an excellent summary:[5]

In other words, as an exercise in constitutional theory, this book looks immanently at constitutional practice, assessing it by its own light. But since this by necessity implies some form or normativity I will thicken this analytical account by deploying recent turns in both republican and deliberative theory – both of which pose their own difficult challenges for referendum democracy. What I propose is a hybrid model of assessment that adopts complementary elements of both. This allows me further to refine the key question this book will attempt to answer: can the referendum, from the perspective of civic republican deliberative democracy, be an appropriate mechanism with which to make democratic constitutional decisions, and if so when and how?

Given this framework in particular, it now seems appropriate to consider referendums as a matter of constitutional practice.

It is not possible in this review to set out the great wealth of information in this book regarding the world’s experience with constitutional referendums.  The eight-page Appendix on “Referendums from 1898-2011” is in itself a fabulous resource, but it is the quality of the preceding analysis that is the book’s most valuable contribution.

Stephen is nothing if not fair in setting out the pros and cons of referendums. Those who are skeptical about excessive or inappropriate use of referendums will find material, old and new, to fuel that skepticism. Stephen cites, for example, the abuse of the referendum by past centuries’ most infamous dictators: Napoleon’s referendums of 1800, 1802 and 1804 to approve him respectively as consul, consul for life, and Emperor;[6] Hitler’s four referendums between 1933 and 1938 to consolidate his rule and his plebiscite in Austria in 1938 after the Anschluss;[7] Mussolini’s two referendums, in 1929 (recognizing Vatican sovereignty) and in 1934 (authorizing fascist rule);[8] the sham referendums in Eastern Europe from 1947 to 1986, involving turnouts of nearly 100 per cent and approval ratings at a similar level;[9] and Duvalier’s similarly unbelievable referendums aimed at strengthening his grip on Haiti.[10]  Stephen notes that commentators with the reputation of de Tocqueville[11] (who criticized the Bonapartists) and Arendt (who saw the danger in the unbridled rule of public opinion) are among the group of referendum sceptics.[12]

He also cites abuses regarding the timing of referendums: choosing and then moving referendum dates according to polling data; holding referendums alongside elections (thereby running different issues together); and doubtful sequencing.[13]

The former socialist republic of Yugoslavia presented other referendum pathologies, where the “crude application of the referendum to settle territorial claims was incendiary”,[14] due to the presence of minorities of different nationalities across the republics. Stephen notes that “the use of a referendum in this type of situation is particularly dangerous not only because with an in-built majority there can only be one outcome”, but also because “this clear act of hegemony is presented as being democratic, offering a spurious constitutional legitimacy for political power play”.[15] Avigail Eisenberg has noted the unfairness in emphasizing “undifferentiated equality” in such circumstances.[16]

Stephen also sets out the French tradition of executive-led constitutional referendums acting as a sort of “unsettling political trump card”[17] in relation to the entrenched constitution. French constitutional culture has been favourable to this process, based in radical republican tradition, though sceptics have been present since Napoleon’s time, as already noted.

Stephen’s purpose, of course, is not to dwell on abuses. The book is committed to the value of civic republican deliberation and democratic decision making, and devoted to ensuring that defects such as elite manipulation, deliberation deficits and well-known majoritarian dangers are dealt with in appropriate fashion.  However, the impressive list of referendum abuses gives rise to a question of principle and practice: can all referendum pathologies be cured using Stephen’s prescriptions?

Stephen’s answer would appear to be that most (all?) objections of principle are in fact problems of process and design and therefore remediable.[18]  If there are referendum problems and scenarios that are beyond remedy, for reasons of principle and/or practice, then it would be useful to see Stephen set these out more clearly.  For example, at the opposite end of the spectrum of referendum process and design that Stephen views as optimal are there highly problematic elements which, in certain circumstances and at certain times, have proved resistant to even the most well-thought-out cures?  As Stephen himself states,[19] his claims must be tested against the facts, and if some problem facts are intractable then we are as well to know. And even if a cure is available, in what circumstances is it preferable to employ the more traditional methods of representative democracy?

This in turn relates to the third topic identified at the outset: unexplored possibilities. This book is written in such a careful, intelligent and thorough manner that it seems churlish to point out things that have not been explored. But the format of this review/response gives me the opportunity to make special requests for further exploration and elaboration.  I have already noted the interest in providing more detail regarding the sort of constitutional theory that Stephen adopts (and that he rejects). I have also suggested that, in addition to the seemingly foolproof recommendations for referendum best-practice, it would have been useful to see a more elaborate description of circumstances in which even Stephen’s well-researched and well-thought-out medicine is less than effective.  One thinks for example of “agonists” or “radical difference” theorists who are skeptical about the very idea of deliberation in divided societies.[20] One thinks also of “deliberative democrats”[21] who question the purported virtue of referendums – their finality – because this closes debate prematurely and irrevocably, thereby undermining the very promise of deliberation which Stephen and others value so greatly.

I would also add to the list of under-explored topics a consideration of the effect of technology in general and social media in particular on the referendum in the twenty-first century. Furthermore, it would have been interesting to learn more about the point set out in a footnote[22] regarding the importance of keeping in mind non-rational forms of communication, for example, emotional expression and story-telling.

According to Stephen, “[t]he constitutional referendum, it seems, is here to stay, and the undertaking with which we are faced is how to make the best of that reality”.[23]  Even if the referendum is part of our twenty-first constitutional reality, and even if undoubtedly those referendums can be arranged more intelligently using Stephen’s recommendations, it remains the case that whether or not to use a referendum is a question of judgement: a question of political statecraft, certainly, but also a question of principled and practical judgement according to the terms of constitutional theory.  Deliberation and democratic participation are powerful, resonant terms in contemporary constitutional theory.  However, if there are real limits to citizens’ willingness or ability to consider certain types questions in certain circumstances, then perhaps we should be repairing representative democracy rather than expanding direct democracy in those instances. Stephen is quite right to say that referendums should not pay a price for a general malaise regarding contemporary democracy.[24]  Nor should they be treated as a panacea. It is important then to determine the issues of principle and practice that should inform our judgments as to when the referendum is an appropriate response to the demands of constitutional design and when it is not.


[1] S. Tierney, Constitutional Referendums: The Theory and Practice of Republican Deliberation (Oxford University Press, 2012), p. 2.

[2] Ibid., pp. 2-3.

[3] Ibid., p. 286.

[4] Ibid., p. 2.

[5] Ibid, pp. 4-5.

[6] Ibid., p. 100.

[7] Ibid., p. 101.

[8] Ibid., p. 101n.

[9] Ibid., p. 101.

[10] Ibid., p. 102.

[11] Ibid.

[12] Ibid.

[13] Ibid., pp. 111 et seq.

[14] Ibid., p. 71.

[15] Ibid.

[16] A. Eisenberg, “The Medium is the Message: How Referenda Lead Us to Understand Equality for Minorities” in M. Mendelsohn & A. Parkin, eds, Referendum Democracy: Citizens, Elites, and Deliberation in Referendum Campaigns (Palgrave, 2001), p. 149, discussed in Tierney, ibid.

[17] Tierney, ibid., p. 131.

[18] Ibid., p. 18:

I will … use case studies to help illustrate some of the problems that referendums present for contemporary constitutional democracy and also to show how, by careful planning and the efforts of political actors and ordinary citizens, process design can be shaped which will help overcome the democratic problems that have at times beset the application of the referendum.

[19] Ibid.

[20] See, e.g., C. Mouffe, “Deliberative Democracy or Agonstic Pluralism” (1999) 66 Social Research 745; C. Mouffe, The Democratic Paradox (Verso, 2000); J. Drysek, “Deliberative Democracy in Divided Societies: Alternatives to Agonism and Analgesia” (2005) 33 Political Theory 218; discussed in Tierney, ibid., pp. 243 et seq.

[21] See, e.g., A. Gutmann & D. Thompson, Why Deliberative Democracy? (Princeton University Press, 2004);  A. Gutmann & D. Thompson, Democracy and Disagreement (Harvard University Press, 1996); M. Saward, “Enacting Democracy” (2003) 51 Political Studies 161; J. Mansbridge et al., “The Place of Self-Interest and the Role of Power in Deliberative Democracy” (2010) 18 J. Pol. Phil. 64; discussed in Tierney, ibid., p. 263.

[22] Tierney, ibid., p. 198n.

[23] Ibid., p. 18.

[24] Ibid., p. 302.


A Reply to Peter Oliver

Stephen Tierney, Professor of Constitutional Theory, University of Edinburgh and Director of the Edinburgh Centre for Constitutional Law

I am extremely grateful for Peter Oliver’s generous and insightful review of Constitutional Referendums. Peter focuses on three elements in the book – theoretical premises, practical concerns and unexplored possibilities – and his comments on each are thought-provoking. I will try to engage with each of the questions Peter raises.

Beginning with the theoretical premises of the book, Peter asks if there is consensus on the meaning of ‘constitutional theory’ or a ‘single, agreed upon framework’ in relation to constitutional theory. In particular he queries whether a Dworkinian would take the same view of referendums as would, for example, Martin Loughlin. In answer to this question I think we need to distinguish between the discipline of constitutional theory as a method of inquiry on the one hand, and particular normative theories concerning the good and the bad within constitutions on the other. This is not to say that constitutional theorists do not often bring their own normative world view to the table when seeking to explain the nature of constitutions and constitutionalism, but merely that there is a discrete discipline of constitutional inquiry, which we can call constitutional theory (or constitutionalism), which is concerned with analyzing and evaluating a particular area of social activity: the framing, founding, practice, and changing of constitutions, operational through internationally comparable institutions and rules.

My intention is to use the term ‘functional normativity’ in part to distinguish constitutional theory from political theory. Whereas the latter is avowedly, and indeed primarily, a normative exercise in assessing the merits of political arrangements, often from first moral principles, constitutional theory is an analysis of the particularities of constitutional institutions, processes and systems. In this sense the normative element involved may be limited to assessing how well these institutions etc. work by their own lights. In this way it becomes more of a functionalist account, albeit imbued with a sense of successful v. unsuccessful institutional arrangements – and hence some idea of good v. not so good institutions. And even when the theoretical analysis is more substantively normative, assessing constitutions as good v. bad or better v. worse in light of externally-generated moral values, this is still, to remain ‘constitutional’ rather than ‘political’ theory, structured within a functionalist context of the good or bad constitution. A particular constitution is evaluated, or constitutionalism as a broader theory is presented, in the context of the functions which a constitution is intended to serve. It is in this sense then an immanent field of enquiry, one that assesses the normative credentials of a constitution against principles of good constitutionalism rather than merely goodness in the abstract moral sense; or it may be more narrowly concerned still, assessing a particular constitution in light of the particular principles or values which that constitution has famed in its text.

In this sense the normative assessment of the constitution emerges more as an exercise in ethics than morality. And it is this which primarily demarcates constitutional theory as a field of inquiry distinct from political theory which is, in my view, more of a Platonic quest for the ideal system, drawing exclusively upon externally-generated models of the good.

On this basis I would link Dworkin, TRS Allan and Loughlin as scholars who each apply institutionally-informed moral reasoning, assessing the good and the bad against the institutional values that are either inherent within the particular practice of constitutionalism or are to be found in the text or values of a particular constitution. The moral (ethical?) values involved are specialist, drawn from the nature of constitutional activity as distinct from political activity: the rule of law, separation of powers, rights (positively defined), responsibility, accountability, proportionality, reason-giving, subsidiarity, public participation etc. Some of these overlap with broader moral reasoning to be found in particular political ideologies, for example in liberal theory – this is not surprising since the dominant ideology underpinning modern constitutions is of course liberal democracy – but that is not to say that the exercise in constitutional theory is simply a mirror for political theory or that constitutional values are merely synonyms for free-standing moral values.

I think therefore that Peter’s point about TRS Allan and Loughlin having different normative views of referendums is separate from the question of whether there is a discrete discipline called constitutional theory which they both practice. Peter is surely right that a liberal constitutionalist would take a very different view of referendums than would a more political or republican constitutionalist. This is something I discuss in the book. Indeed, having argued for the existence of a distinct discipline of constitutional theory built around the idea of functional normativity, I go on to describe how one’s attitude, in normative terms, to referendums will depend upon the values one brings to assessing the practice of referendums: ‘[o]ne’s preferred model of democracy will in many ways shape one’s attitude to referendums.’ (p.21)

In short, one key endeavor in the book was to find space for constitutional theory as a discipline that seeks to understand, and by turns to evaluate, constitutional practice as a discrete form of human behavior, just as others have sought to find space for the wider category of legal theory as a sub-category of practical philosophy. In this sense I do believe that, despite the impossibility of cocooning constitutional institutions from the political world whence they derive and in which they develop, it is important that we do not allow our discipline to be subsumed within political theory more broadly.

The second element Peter focuses upon is constitutional practice. His comments here flow from the first point. He rightly alludes to a key passage in the book where I set out my core task: to assess whether a referendum can, from the perspective of civic republican deliberative democracy, be an appropriate mechanism with which to make democratic constitutional decisions. Here I adopt the normative position of civic republicanism, combining this with the recent turn in deliberative democratic theory (‘civic republican deliberative democracy’) as a benchmark. A liberal constitutionalist may well reach different conclusions about the merits of referendums. Having said that, I would be interested to know to what extent that is in fact the case since my defense of the limited use of referendums seeks to take full account of individual rights and sets out the protection of minorities as a core concern – so further engagement with the book by liberal constitutionalists would be very welcome!

Peter is very complimentary concerning my empirical analysis and assessment of referendums and what he takes to be even-handedness in my approach. I do endeavor to explore fully the three main failings from which referendums have suffered: what I call the elite control syndrome, the deliberation deficit and the majoritarian danger. It is my view that these objections, when valid as descriptions of particular referendums, are largely problems of practice not principle, and secondly that the recent turn in deliberative democratic theory provides practical ways of overcoming these problems. I argue that the principles of popular participation, public reasoning, and equality and parity of esteem can be used to help surmount these practical problems. Nonetheless, Peter does raise the very important question: ‘can all referendum pathologies be cured using Stephen’s prescriptions?’

In short, I don’t think so. This is one of the reasons why the defense I present of referendums in the book is a very narrow one. I argue that referendums are defensible for higher order, or constitutional, decision-making. My view is that otherwise, for the vast majority of political decisions in a democracy, representative institutions are better tailored. Why then are referendums potentially suitable for highest order constitutional decision-making? I have discussed this at length elsewhere. Briefly, given the importance of constitutional decisions, and the impact these can have upon citizens’ lives, in some cases serving even to help shape citizen identities over time, there is, from a civic republican perspective, a plausible argument for engaging citizens directly in exercises of constitutional authorship.

I also discuss the deep-seated problems of attempting to stage a referendum in deeply divided societies where the referendum ‘can serve to expose and indeed inflame what is often a dormant disjuncture between the boundaries of territorial government and the nature of the demos/demoi within that territory’ (p75). A key point is that since the referendum calls upon ‘the people’ to speak, the boundary question (as political scientists call the demos issue) must either be uncontroversial or must be fully addressed before a referendum in a divided society is possible. I go on to contrast, for example, the failure of the referendum in Northern Ireland in 1973, which was used in place of a settlement of the core issue, with that in 1998, where the referendum came at the end of a carefully-crafted peace agreement, offering the people on both sides of the dispute a direct say on the Belfast Agreement; a referendum which, as an exercise in popular endorsement of a cross-community agreement, has been a considerable success (pp 248-59).

The third element which Peter discusses is unexplored possibilities. I am very grateful for his ideas here which will certainly inform on-going work which I am doing. In response to some of Peter’s specific questions I would say it is not just agonists who are skeptical about the prospects of success for deliberation in divided societies. As well as the Northern Ireland case, I also discuss the referendum in Bosnia Herzegovina in 1992 which served only to inflame the political problems there (pp72-74). Another situation where the referendum has offered little is in New Caledonia, where it is clear that direct democracy alone cannot overcome the hegemony of a colonial power and the dominance of a society by a colonial elite (pp 84-97). Where I, and others, would part company with agonism is in the tendency both to over-exaggerate the salience of differences, particularly in societies which are not divided as such, and to under-estimate the potential for deliberative democracy to bring people together to the point of agreeing, as Richard Bellamy puts it, to a particular outcome, if not with that outcome.

Another important question is the seemingly inherent tension between referendums as vote-based decision-making devices on the one hand, and deliberation, which emphasizes the role of discussion and reflection, on the other. In the book I address this issue, discussing how deliberation should be, in Simone Chambers’ terms, ‘talk-centric’ not ‘vote-centric’. But I also note that as constitutional theorists, situated in the reality of constitutional practice, we must be concerned with deliberative democracy, and not just deliberation. A constitutional theorist, even one promoting the advantages of deliberation to decision-making, must remain alive to the hard fact of constitutional life that decisions must be made at the end of that process. Applying deliberative democratic theory to the referendum or other constitutional processes is an attempt to arrive at better and more plurally acceptable decision-making, not perfect decision-making. A meaningfully deliberative process can enhance the legitimacy of a constitutional decision but that decision must still be reached, whether by referendum or by some other means, and this outcome will not satisfy everyone.

Finally, Peter reminds us also of the effect of technology and social media as an important consideration for the use of direct democracy more generally. The focus of my book is on major constitutional referendums in which the use of technology has hitherto not been a major issue. But this is a rich seam for new research, and there does seem to be ample opportunities to connect different literatures: that on the use of social media in deliberative discussion; that on the use of new technology and innovative ways to engage citizens in constitution-building, looking at the recent experiments in Iceland for example; and that on innovations in technology that allow more flexible approaches both to deliberation and to vote-taking.

To conclude, the referendum is a controversial tool, but I hope I have presented a convincing argument that many of the democratic problems associated with it, from whatever ideological position a constitutional theorist approaches the issue, can be viewed as weaknesses of practice not principle. I also discuss how many of the most egregious abuses of the referendum process take place in states where representative democracy itself does not function properly; Peter is surely right that defects in representative democracy must also themselves be corrected. If the book has served further to illuminate some of these problems then another aim will have been satisfied.

But beyond this I do make a case for the further direct engagement of citizens in political life as a constitutional good. If we accept that there is value in public participation then I would submit that we should take the same approach to practical defects in referendum processes that we bring to those which afflict representative democracy: attempted remedies within, rather than amputation from, the body politic. It is in this context that I offer a limited defence of constitutional referendums. I explore how referendums offer citizens a direct say in the major constitutional decisions which shape the world they live in, and in doing so can even play a role in shaping their public identities as citizens. In an age where the connection between citizens and the world of politics is becoming ever more strained, this is surely of some prima facie value. Even here my approach is one of defense not prescription; I believe it is for states, with their own constitutional traditions and constitutional values, to determine if the referendum is a model of decision-making they wish to utilize when taking major constitutional decisions. My view is that if the process is properly constructed, meeting the deliberative principles I set out in the book, and thereby overcoming the democratic objections to which referendums have been subjected, then the referendum can be a legitimate tool of constitutional decision-making. Referendums are not a panacea for the failings of representative democracy and should not be treated as such, but they are a means of connecting citizens directly with their own constitutional identities. If this is indeed the age of direct democracy, as the proliferation of direct democracy in constitutional practice seems to suggest, then I hope that this book, in offering a road map with which to build a better connection between citizens and constitutional outcomes, will be seen as an attempt to make some kind of virtue out of this reality.

Suggested Citation: Peter Oliver and Stephen Tierney, Book Review/Response, Constitutional Referendums: The Theory and Practice of Republican Deliberation, Int’l J. Const. L. Blog, February 25, 2014, available at:


Leave a Reply

Your email address will not be published. Required fields are marked *