I am indebted to Scott Altman, Richard Bronaugh, Marshall Cohen, Chaim Gans, Elizabeth Garrett, Alon Harel, and Wil Waluchow, for helpful comments on earlier drafts.

1. Allow me an apologetic note on bibliography: the literature on the critique of constitutionalism is vast, particularly in the US. Dozens of books and hundreds of articles cover many, if not all, of the issues I discuss here. I couldn’t possibly do justice to all this literature, and I am sure that most of the arguments I raise here have been alluded to by somebody somewhere. My purpose here is to present a systematic, but not an encyclopedic, treatment of the debate on the moral legitimacy of constitutionalism. I have tried to incorporate relevant references, but the notes are not meant to be comprehensive.

2. And I will confine myself to constitutions of countries, not sub-federal states or regions. Those raise very different moral-political issues.

3. The main reason why the existence of a constitutional document makes such a difference consists in the fact that without such a canonical document, courts would find it very difficult to exercise their power of judicial review. Typically, this power is granted to the courts by the constitutional document. But even if it is not, the document makes it much easier for the courts to hold the legislature under their review power.

4. The constitution’s normative supremacy should not be confused with the idea that all law derives its legal validity from the constitution. This latter thesis, famously propounded by Kelsen, is probably false in most legal systems. (see Kelsen, Hans, Introduction to the Problems of Legal Theory, trans. by Litschewski Paulson, Bonnie & Paulson, Stanley (Oxford: Clarendon Press, 2002) at sec. 31Google Scholar.

5. Typically, this would mean, de facto, that the highest court of appeal in the country is basically its constitutional court. Whether this is the case, and to what extent, mainly depends on how easy it is to appeal constitutional cases to the country’s highest court.

6. A very interesting and suggestive exception is section 33 of the Canadian Charter of Rights and Freedoms which allows the legislature to overrule constitutional decisions of the supreme court (both preemptively or ex post), as long as it is done so very explicitly and renewed every five years. More on this in the last section.

7. As I have argued elsewhere, the content of the constitution is bound to change according to its interpretation by the courts. See my ‘Constitutional Interpretation’ in Interpretation and Legal Theory, 2nd ed. (Oxford: Hart, 2005) at ch. 9. Some implications of this will be discussed below.

8. To be sure, I am not claiming that important moral content is unique to constitutions. A great deal of statutory law also regulates matters of great moral importance.

9. Once again, constitutions vary considerably in this respect as well. Many constitutions contain very specific provisions even in the realm of rights and principles.

10. Another aspect of the US constitutional regime that makes is relatively robust has to do with the fact that in the US there is no separate constitutional court. The highest court of appeal in the country is also the main constitutional court. Many countries have separated these two legal functions. There is something to be said in favor of such a separation, but I have no evidence to support my intuitions here.

11. Rawls, John, A Theory of Justice (Oxford: Oxford University Press, 1971) at 3 Google Scholar.

12. I am using purpose or rationale in singular only for the sake of simplicity. Constitutions may have several rationales.

13. Let me add two clarifications. First, there is another sense in which the educational value of a constitution is parasitic on its legitimacy: for something to have such value, it must be morally sound. There is no reason to celebrate and teach something that is actually wrong. But this is not the main point I want to make in the text. Second it may be suggested that if an institution is not quite, but almost legitimate, its additional values may tilt the balance, as it were, and then these values may turn something that would otherwise not be legitimate into a legitimate institution. Perhaps so. But this would be an odd chance, and I think we may dismiss it.

14. See Elster, Jon, Ulysses Unbound (Cambridge: Cambridge University Press, 2000) at ch. 2CrossRefGoogle Scholar.

15. And, of course, some of them are morally very disturbing (e.g., huge national debt, irreparable damage to the environment, etc.).

16. Waldron, Jeremy, Law & Disagreement (Oxford: Oxford University Press, 1999) at 243–49, 268CrossRefGoogle Scholar.

17. This is one of the well-known themes in Rawls’ Political Liberalism. I tend to agree with Waldron, however, that reasonable disagreement extends to conceptions of the rights as well, not just conceptions of the good as Rawls seems to have maintained.

18. I take it that this is Waldron’s view. See supra note 16.

19. I have defended this position in my ‘Authority, Equality, and Democracy’ (2005) 18 Ratio Juris 315.

20. England has had a pretty stable regime for the last few centuries without a written constitution. New Zealand does not seem to be in any danger of instability because it does not have a written constitutional regime. At the same time, we know that there are countless instances of political instability in countries that have admirable constitutions.

21. Perhaps one can point to the familiar idea of the role of constitutions as a ‘civic religion’; the idea is that constitutions tend to provide a focal point of civic identity and social cohesion. A trite saying has it that constitutionalism is our civic religion with the constitution as its holy scripture. The problem is not with the sociological insight here, which may well be more true and more interesting than it sounds, but with its normative significance; it is difficult to extract a moral-political argument from this piece of folk sociology. Perhaps we should stick to constitutional atheism: It is far from clear that healthy democracies ought to have a civic religion. (Nor is it clear that constitutions have a significant role to play in actually creating the conditions for its emergence.) In any case, to the extent that constitutions are conducive to the maintenance of some social cohesion and civic identity, that might be an added benefit of constitutionalism (akin to its potential educational value), but not a moral justification of its legitimacy.

22. See my Interpretation and Legal Theory, supra note 7 at 137-38, 146.

23. Raz, Joseph, ‘On the Authority and Interpretation of Constitutions’ in Alexander, Larry, ed., Constitutionalism: Philosophical Foundations (Cambridge: Cambridge University Press, 1998) 152 at 173Google Scholar.

24. For a much more detailed account of the nature of social conventions see my ‘On Convention’ (1996) 107 Synthese 349 and Positive Law & Objective Values (Oxford: Oxford University Press, 2001) at chs. 1 and 2.

25. Note that we are talking here about persistent and vulnerable minorities. Anyone can find himself in the minority on some issue or other, but this is not particularly problematic. Our moral concerns pertain to minorities that are particularly weak or vulnerable and tend to persist as minorities for a considerable period of time.

26. One clear example is the election system: Proportional representation tends to protect minorities much better than non-proportional representation. Other examples concern districting, the role and structure of political parties in the political landscape, etc. See, for example, Lijphart, A., Patterns of Democracy, Government Forms and Performance in 36 Countries (New Haven, CT: Yale University Press, 1999)Google Scholar, and Horowitz, D., Ethnic Groups in Conflict (Berkeley: University of California Press, 1985, 2nd ed. 2000)Google Scholar.

27. This formulation is admittedly too strong. Of course there are some constraints on judicial decision making in constitutional cases, mostly those that derive from precedents and constitutional tradition. But it should be kept in mind that those precedents and traditions are created by the judiciary, that is, by the same institution that is supposed to be constrained by it.

28. See, for example, Garrett, Elizabeth, ‘The Purposes of Framework Legislation’ (2005) 14 J. Contemp. Legal Issues 717 Google Scholar.

29. This is basically the main intuition, I think, that drives Ely, J.H.’s ‘procedural’ conception of judicial review. What he sees as legitimate in the US constitutional review is the protection of the democratic process, not ‘substantive’ rights. See his Democracy and Distrust (Cambridge, MA: Harvard University Press, 1980)Google Scholar. It is a difficult question which I need not address here, viz., whether Ely’s position is a feasible constitutional interpretative strategy. Many have raised doubts about this, and I suspect rightly so.

30. See, for example, Arneson, R., ‘Democracy is not Intrinsically Just’ in Dowding, K., Goodin, R.E. & Pateman, C., eds., Justice and Democracy (Cambridge: Cambridge University Press, 2004) 40 CrossRefGoogle Scholar.

31. See Estlund, D., ‘Beyond Fairness and Deliberation: the Epistemic Dimension of Democratic Authority’ in Bohman, & Rehg, , eds., Deliberative Democracy (Cambridge, MA: MIT Press, 1999) 173 Google Scholar.

32. A point often stressed by Waldron. See supra note 16.

33. As an example, consider this: J.K. Rowling’s decision whether to write another episode in her Harry Potter series affects the lives of millions. Surely, that is not the kind of decision that ought to be subject to democratic process.

34. Waluchow, Wil, A Common Law Theory of Judicial Review: The Living Tree (Cambridge, MA: Cambridge University Press) [forthcoming, 2006]CrossRefGoogle Scholar.

35. Those who know something about the workings of the US supreme court may raise their eyebrows here; at least in the US, a great deal of decision making in the supreme court is certainly based on subtle bargaining between the nine justices.

36. I do not wish to put any weight on the term ‘discovery’ here. Waluchow advocates something like a Rawlsian Reflective Equilibrium method, Dworkin relies on his theory of ‘constructive interpretation’, and yet others may have different ideas in mind. Whatever method one has in mind should not affect the arguments in the text. With one notable exception: the so called ‘orig-inalism’ in constitutional interpretation would not be compatible with Waluchow’s argument. But on this we are in complete agreement. On the question of why originalism makes no sense in constitutional interpretation I have elaborated in my ‘Constitutional Interpretation’.

37. Not without good reasons. Dworkin’s arguments in Law’s Empire (about the value of integrity and the importance of seeing political decisions as if the community speaks with one voice), certainly support Waluchow’s interpretation of Dworkin. More generally, their constitutional theories are very similar.

38. Notably, contemporary theories that espouse such an emphasis on the common good the so called ‘deliberative democracy’ theories, hold the opposite view: they rely on the value of broad inclusive, and egalitarian public deliberation as the kind of process that is likely to yield decisions that constitute, or are in line with, the common good. See, for example, Joshua Cohen, ‘Deliberation and Democratic Legitimacy’ in Bohman & Rehg, eds., supra note 31 at 67.

39. In my ‘Constitutional Interpretation’ (supra note 7) I have argued at some length that rights discourse is particularly deceptive in this context. Basically, the argument is that rights in pluralistic societies are such that it is relatively easy to agree on the rights we should have, but that this consensus is very deceptive. The underlying reasons for having rights and their appropriate ramifications are almost inevitably controversial.

40. See Raz, Joseph, Ethics in the Public Domain (Oxford: Oxford University Press, 1994) at ch. 12Google Scholar.

41. See my ‘Constitutional Interpretation’, supra note 7.

42. For the purposes of the example, we should ignore practical or institutional constraints that may be involved in such decisions.

43. I need to qualify this: I do not wish to claim that this argument applies without qualifications to federal systems, where the court’s decision amounts to allowing the states (or regions) to make the decision democratically within their jurisdiction. This is a much more complicated matter, involving difficult questions about the appropriate division of democratic processes between federal and local authorities.

44. The specific formulation of Article 33 raises interpretative issues, mostly about its scope, that we need not discuss here. I am not suggesting that the particular formulation of Article 33 is ideal. To this date, Article 33 has not been invoked by the Canadian legislature at the federal level. I would not find this particularly discouraging. Legal sanctions don’t always have to be applied in order to change the behavior of the relevant agents.