Estlund Reading Group Chapter 5

In the previous chapter, David showed that procedural fairness could not by itself justify democracy. According to “fair proceduralism”, a law that is the outcome of a democratic vote is legitimate because everyone had an equal role in determining the outcome, regardless of whether it is good or just by other standards. Yet since a coin flip also provides equal input, fair proceduralism must incorporate some substantive value, such as the principle that outcomes should be positively sensitive to voter preferences.

Chapter 5 confronts a possible response to this argument, a response based on a partial concession. Yes, we have to appeal to substantive values to justify democracy, but we can do so with respect to the intrinsic characteristics of decision-procedures, without making any claims about their tendency to yield good decisions or (incidentally) to generate other desirable outcomes. We thus avoid any substantive assessment of political decisions / outcomes (except for endorsing the preservation of democracy, which is an outcome, but the outcome that consists in the continued functioning of our intrinsically-valued procedure). David calls this position “intrinsic democratic proceduralism” (86). The chapter criticizes 3 versions of this doctrine: Habermas, Cohen, and Waldron.

According to Habermas, there are “no standards that loom over the political process, policing its decisions” (88), but only standards of procedural rationality implicit in democratic deliberation itself. So it looks at first as if Habermas avoids any substantive assessment of outcomes. However, it turns out that one of the norms that is implicit in deliberation is that outcomes are legitimate only if they could have been agreed to in the ideal speech situation. The standard is not that outcomes in fact result from a particular, intrinsically-valued procedure, but that they could result from a hypothetical procedure. Habermas too judges outcomes by a standard independent of their actual procedural source (89), and so is not, ultimately, an intrinsic democratic proceduralist.

Cohen too claims that democratic choice is not under any external authority, not even the authority of independent moral truths. Democratic deliberation is free in that participants regard themselves as bound only by the results of their deliberation and by the preconditions for that deliberation – deliberation between free and equal citizens. One of these preconditions, however, is that coercive political arrangements are morally illegitimate unless they can be accepted by citizens with a range of moral, religious, and philosophical views. In other words, a “qualified acceptability requirement” (QAR, as we have been saying) is part of the democratic ideal. Yet Cohen only reaches this conclusion, David claims, by expanding the boundaries of the concept of democracy. Citizens are not radically free, constrained only by the preconditions of deliberation and whatever laws they choose. They are constrained also by the principle that laws must be acceptable to all reasonable people.

Waldron claims that large assemblies are valuable because, whether or not they produce good decisions, they air the views of a wide range of citizens. Apart from any thought about the wisdom of crowds, fairness requires that everyone’s views have a hearing. However, fairness by itself would be compatible with an equal non-hearing of views. If some participants had a greater opportunity than others to develop and air their views, we could achieve equality by leveling down, if fairness were all that were at stake. Imagine that individuals have views of no better than random quality and that the deliberative decision-procedure is no better than random. Individuals would have no interest in participating equally in such a process, when equal treatment in decision-making could be achieved far more easily by lottery. Or, suppose that individuals varied in their persuasiveness, and could hope to attain results they desire, while the decision-making process as a whole is still no better than random in attaining justice / the common good (or whatever non-procedural standard for the assessment of outcomes one prefers). Why should persuasiveness be given special importance in social choice? “The idea of fairness to views is difficult to understand without assuming that the goal is a deliberation or outcome that is more responsive to the genuine balance of applicable reasons” (96).


1. I think David is right that Cohen stretches the idea of democracy too far, but I also think that there is something to Cohen’s argument. Cohen didn’t simply stipulate that “democracy” as he conceives of it includes qualified acceptability. He tried to offer an argument that a QAR was a precondition for democratic deliberation. His main example was religious freedom. Given the overriding importance people place on religious obligations, forcing people to pray or not to pray in a particular way denies them equal standing as citizens, or “full membership” (Cohen 1996, 103). Yes, they could continue to exercise the full rights and liberties of citizens if they were willing to respect the law, but this would mean abandoning their faith. If the majority cannot democratically kill or enslave the minority, perhaps it cannot democratically curtail religious liberty. The majority might impose such policies, but even if popular, enslavement and religious intolerance would still be inconsistent with the ideal of free and equal citizenship. This inconsistency isn’t as clear in the case of restrictions on religious liberty as it is in cases of killing or enslavement of the minority, because dissenters do have the option of converting, but this option isn’t much of a choice. In short, if the idea of democratic citizenship involves an anti-caste conception of equality, then religious liberty is plausibly a precondition of democracy.

I think this argument has some plausibility. However, I also think that it works only for certain core rights. It does not justify a blanket exclusion of reasonably contestable reasons, no matter the kind of policy in question. The mere fact that a reasonably contestable reason is necessary to justify a policy does not imply that someone’s basic rights have been violated. Someone who justifies more generous redistribution of wealth on grounds of Christian charity is not violating my religious liberty, nor excluding me from membership. Not every form of differential treatment can count as a denial of my equal standing as a citizen, even if justifiable only on reasonably contestable grounds. There might be democratic grounds for arguing that “constitutional essentials” must meet a QAR, but any more expansive requirement of qualified acceptability would need an independent (liberal) rationale. And even with respect to constitutional essentials, what matters, on the democratic argument, is not what ideas are invoked but whether there is or is not religious liberty. So I think there is more to Cohen’s argument than David allows, but at the end of the day David is right that Cohen is taking substantive liberal values and relabeling them as democratic.

That’s not much of a question, but I would be interested in knowing whether the modified Cohen argument, the more modest claim that religious liberty is a precondition of democratic deliberation, would count as intrinsic democratic proceduralism.

2. More generally, I guess my question would be why someone couldn’t dispense with all claims about justice or correctness in outcomes, but accept that individuals have interests they want to defend in the setting of policy, and so argue for democratic decision-making on the basis of a conception of equality. We might call this “fairness to interests”, in contrast with “fairness to views”, though we’re talking about fairness to interests in decision-making, not in decisions made. This would seem to be the simplest kind of intrinsic democratic proceduralism. The previous chapter ended with the consideration of this kind of a view, on p.83 – equal (positive) sensitivity to voter preferences. If Chapter 5 is supposed to have “cast serious doubt” on the ambition of normative theory to avoid appealing to procedure-independent standards, doesn’t “fairness to interests” deserve more attention? Of course, this view wouldn’t have much place for deliberation, and the point of the chapter was to criticize the deliberative flight from epistemic substance, but I’m wondering if I’ve missed a step in the argument. No plausible theory avoids all appeal to substantive values(Chapter 4), and the leading theories that claim to appeal to substantive values only in the justification of procedures in fact contain outcome-standards too (Chapter 5), but the minimal p.83 view still seems to be on the table as an example of intrinsic democratic proceduralism.

Be Sociable, Share!
This entry was posted in Posts, Reading Group. Bookmark the permalink.

6 Responses to Estlund Reading Group Chapter 5

  1. Hi Andrew,

    Thanks for the great post. I’d like to try and offer an extra spin on your second point. You say (and I agree) that it would be interesting to hear more from David as to why a ‘fairness to interests’ approach as you call it isn’t still on the table as a form of non-epistemic proceduralism (even though it isn’t fully procedural in the sense spelled out in chapter 4). I suppose what David might say is that if we just care about being fair to everyone’s interests, we could use a coin-toss or some other randomized procedure (king/queen for a day) which would do the job. But most democrats want something like deliberation followed by majority rule, and the brute appeal to treating interests fairly won’t justify deliberation or majority rule.

    But here’s a proposal (following on from what I said last week) that seems like it stands a chance of overcoming this problem. David’s coin-toss or randomized procedure objection appears to rely on the implicit premise that political power is mostly a positional good. What that means, roughly, is that what matters is how much of it you have relative to others. The value of positional goods will depend on where you stand relative to others in terms of your possession of it. If political power is a positional good, then randomized procedures don’t look to be any different than majority rule because they each give everyone the same share. But political power might only be a partly positional good. It might also be the kind of good that is valuable to possess in absolute terms, regardless of how much others have. This seems plausible to me since the more political power I have the more control I can exercise over my own life. This increase in control is valuable even if I still have less control than everyone else.

    If we suppose that the value of political power is partly non-positional, then it might make sense to see it as a kind of primary good, one which parties behind a veil of ignorance would always prefer to have more of. On this view it makes sense to talk about the parties caring directly about how much political power they have in absolute terms. And because the parties are behind the veil of ignorance they will agree (let’s suppose) to give each person the maximal amount possible compatible with a similar amount for all. This provides a rationale for majority rule as superior to a coin toss or other randomized procedures since those do not give each person the maximal share of political power possible, rather they give everyone no share, or they give everyone an equal chance at the maximal share. I think it’s plausible to suppose that both of these options would not be preferred in the original position (i.e. the parties do not give themselves an equal chance at gaining basic rights over everyone, rather they give themselves an equal share of rights and liberties).

    What about the value of deliberation? David says it will be hard to explain the value of deliberation without relying on epistemic considerations, but consider the following alternative strategy. Suppose we have selected majority rule for the reasons suggested in the previous paragraph. But suppose we also endorse an independent moral principle (this principle is borrowed from Christopher Eberle): respect for persons requires that we always try to find public justifications for the laws we support (note this is very different from Rawls’s duty of civility). The best way to try and find public justifications for the laws you support is to actually deliberate with others in public, to try and see if you can justify your favoured policy to others who disagree. You owe this effort to other citizens as a mark of respect to them as reasoners who will be affected by your use of political power. This explains why we ought to deliberate in a democracy. This argument is not epistemic since it does not assert that public deliberation is valuable because it is more likely to produce publicly justifiable decisions. The argument only holds that you ought to try and find public justifications as a mark of respect for others. If you are unable to find public justifications for your favoured policy then you can still go ahead and vote for it – it’s the sincere effort to engage with other people’s reasons, not any measure of success in doing so, that matters in terms of respect for persons. So even if democratic deliberation was no more likely to produce publicly justified decisions than a coin toss, it would still be preferable because the effort at deliberating with others manifests respect for them. I think this is different than Waldron’s position since it is not about allowing each person to air their views, but rather about the moral duty to try and justify yourself to others. I think (though David may disagree) that this means it is a more plausible non-epistemic story.

    In sum, majority rule is justified in a manner similar to the way the basic rights and liberties are justified in TOJ, and deliberation is justified via appeal to an independent moral principle based on respect for persons. Epistemic considerations don’t seem to enter the picture. I should perhaps add that I don’t endorse the argument I’ve proposed (in particular I reject Eberle’s principle), but it seems like a possible way of avoiding the difficulties David has laid out for non-epistemic theories.

  2. Yes, the phrase “fairness to interests” isn’t good, given the argument of chapter 4. I’m just looking for a moniker for the p.83 view, which involved equal positive sensitivity to expressed voter preferences. This seems like a view that avoids appealing to substantive values in the assessment of outcomes, unlike deep deliberative democracy, from Chapter 5. Your distinction between the positional and non-positional aspects of political power is useful to explain this view. If it’s good for people to have control over their lives, then it is good for the decision procedure to be positively sensitive to their preferences, unlike the coin flip. Start with one person; it’s usually better to make one’s choices for reasons rather than flipping a coin. Next three people: flipping a coin would not be unfair, but if two out of three prefer one choice, why not let them have their way – assuming that no one can reasonably reject the idea of having a common policy for the group.

    The non-epistemic defense of deliberation you develop is interesting, but I’m not clear why it’s respectful to look for a public justification of a policy if one doesn’t feel bound to vote for or support only those policies one thinks can be publicly justified. Also, are we looking for a justification any reasonable person could accept, based on some limited set of public considerations, or just a justification that the opponents of the law can accept, based on their own comprehensive set of views? I ask this because of your expression “the sincere effort to engage with other people’s reasons”, which to me suggests the latter, which would be looking for a justification from (part of) the opponent’s comprehensive point of view.

    Here’s a thought about what might be wrong with the “equal positive sensitivity to expressed preferences” view. This view is supposed to be entirely divorced from any claim about the justice of the decision-procedure’s outcome, or its tendency to realize the common good, or whatever objective, procedure-independent standard one prefers. Therefore, in the case in which A, B, and C agree that common policies P1 and P2 are better than not having a common policy, A and B are free to vote for P1 over P2 just because it is very beneficial for them, even if P1 is only slightly better than no policy for C. Why should that be legitimate? If no one is trying to vote for the common good, why should C be obligated by the results of the procedure, even when that result is better for C than would be the absence of a common policy? To rule themselves, people need to be participating in the making of laws. But to rule themselves, people need to be participating by aiming at justice and the common good, so that those in the minority can still see themselves as obeying themselves, as Rousseau would say. But I’m getting a bit ahead of the game, here: Chapter VI on “Epistemic Proceduralism” will deal with Rousseau.

  3. Ben Saunders says:

    Sorry for not commenting before now; it’s been a busy week. Most of my thoughts have already come up, either in the comments directly above or in previous weeks though.

    I assume we all agree that a coin-toss provides equal respect for all. I’d have thought an argument for majority rule could begin from offering equal and maximal concern, which reminds me of Christiano’s claim that we care about equality of things that are good, so the principle of equality itself prefers more to less and so resists levelling down. This seems to lead in the same direction as Jonathan was going, towards wanting equal slices but also bigger slices, because power is not merely a positional good.

    Of course, David may simply be unmoved, given his claim that fairness doesn’t really matter, but I’m still not really convinced of that. It seems that fairness on its own may not do much (as in coin-tossing), but perhaps the appropriate form of fairness is combined with other values, fairness remaining an important element.

    The contrast between deliberation and throwing darts, at the end of the chapter, is a small point (or small illustration of a larger point), but one of the parts that I found most interesting. Again, it seems we can say that either procedure is fair, so why prefer one to the other? David I think assumes deliberation is epistemically better. I think the reasons may depend on outcomes, but by-products rather than decisions – we’d rather develop people’s capacities for persuasion and argument than their darts.

    Anyway, sorry again that this is probably banging on about the same issues, but this chapter is essentially a continuation of the last, and I want to try to be clear about what’s rejected and why before we move to the positive proposal.

  4. Like Ben, I’ve had a busy couple of weeks, so I’ve been behind on the reading till now. It would also be far easier to do political philosophy if politics didn’t keep getting in the way. I’ve been tempted to put up a poll about who will win the nomination.

    One thing I am a bit confused about in this chapter is the response to Cohen. As I understand it, David’s view has 1) an epistemic component — democratic procedures will be valuable because they score fairly well in terms of getting the right answer, and 2) a qualified acceptability requirement — democratic procedures are the best epistemic procedures that also satisfy the QAR. As alternatives to democracy, epistocracy is ruled out by 2) and coin-tossing is defeated by 1). Now we know that the QAR has a kind of axiomatic status in the theory here — there is no direct argument for it, and I suppose the support it does have will be generated from coherentist considerations about the theory as a whole, and perhaps the weakness of direct attacks on the QAR. But in any case, the QAR and the epistemic component are distinct.

    Cohen seems to me to employ a version of 2) without 1). His theory does involve appeal to substantive values of course, such that some outcomes of even the best actual procedures are ruled out. But these substantive standards Cohen wants to say are not independent of democracy. Instead, we can get them by massaging our understanding of the democratic ideal so that it isn’t just about procedures. In addition to the basic idea of reasoned deliberation, we are also to offer reasons that are reasonably acceptable to all people. Hence, those outcomes of real procedures that could not also be justified in an ideal deliberative procedure by reasons acceptable to all will not be legitimate.

    David’s response to this, I take it, is that this is all well and good as far as it goes, but it goes further than the concept of democracy. The QAR, or a Cohen/Rawls principle of individualised justifiability, is a specifically liberal principle of justification (p. 92). So Cohen’s position is not a good representative of intrinsic democratic proceduralism.

    I think there could be a couple of things Cohen could say in response. First, Cohen’s commitment to a QAR for reasons in public deliberation is based, in part, on the idea that it is a “forceful rendering … of the fundamental democratic idea — the idea that decisions about the exercise of state power are collective” (PSDD, 415). So when David distinguishes between popular sovereignty and individualised justifiability, I take it Cohen would think that too quick. Sure, there can be forms of the latter which do not involve the former (a la Hobbes), but it doesn’t follow that popular sovereignty should not be grounded in some idea of a collective viewpoint, or a perspective common to all. This is just a basic inheritance from Rousseau, isn’t it? The majority don’t rule in their own name, but rather in the name of the People as a whole. So Rousseau will invoke the general will as the collective viewpoint, and Cohen will invoke an ideal deliberative procedure instead. This all seems to me to be operating within the idea of democratic popular sovereignty, rather than being an extraneous latter-day importation from liberal theory. Or at least, Cohen seems to think that it’s enough of the former to not involve any contradiction of the priority of democracy to other values (PSDD, 410).

    Well, we may now wish to prod Cohen on exactly why popular sovereignty should be understood in terms of the rule of reasons common to all. I don’t favour that sort of approach myself, but I’m not sure if he has done anything much different than what David has done by asserting the QAR as a basic premise.

    The second point is to say, well, it doesn’t really matter whether you think of individualised justifiability as internal or external to democratic popular sovereignty. What’s important is simply whether you accept it as some or other substantive constraint on decision-making procedures. But say we do accept it. Does this bring us any closer to accepting the epistemic element 1), i.e. the tendency of democracy to produce the right answer? Or are we still trying to sort out just how, where, and in what form to apply 2)? It seems to me that we’re still quite far away from the conclusion that the tendency to generate the right answer is crucial to democratic legitimacy. I have a feeling that the next chapter is going to be key in figuring out the route to that conclusion.

    I’m still thinking about what everyone else has written above. Following the equal chance/equal share distinction Jonathan draws, I think there are parallels that can be drawn here with equal of opportunity — we don’t usually toss coins to select people to occupy various positions, so an equal chance model is a poor model for equality of opportunity. I don’t think this is only because of the absence of a random procedure to select the right people for the job.

  5. I forgot to insert a comment here saying that I posted some responses to the comments here, through comment 3. I hadn’t seen Simon’s comment 4 at that point. I’ll say something about Simon’s comment if I can find time.

  6. I think we were writing simultaneously David.

Leave a Reply