I. Introduction

It is clear what Cristina Lafont is for and what she is against in Democracy Without Shortcuts (2019).1 She is for the ‘full endorsement’ of all laws and policies by all those subject to them, which she thinks is best achieved by full deliberative participation across the entire community (Lafont 2019: 4). She is against all sorts of shortcuts that in one way or another demand ‘blind deference’ to the judgment of others (Lafont 2019: 8).

That, however, is to propose an impossible ideal on the one hand, and to offer an impoverished set of alternatives to it on the other. If ‘full endorsement’ means anything like ‘complete consensus’, then that is an utterly unrealistic goal in any actual polity. Lafont must be prepared to accept something short of that, in the real world. On the other side, ‘blind deference’ is simply the limiting case of the great many ways we can, and do, take the inputs of others into account when arriving at our own decisions.

The aim of this article is to populate the ‘missing middle’ between Lafont’s two polar extremes. I begin by identifying various types of ‘non-blind deference’—a category I mean to include all the ways in which we might reflectively take the input of others into account when coming to our own decisions. I then consider to what extent the standards that Lafont herself accepts as adequate approximations to ‘full endorsement’ in the real world mirror the standards for morally acceptable ‘non-blind deference’ of other sorts.

II. Backing Off Blind Deference

The essence of moral autonomy is to ‘give laws to yourself’. Its antithesis is to do or believe what someone else says, just because she says to (Hanrahan & Antony 2005). Therein lies the classic tension between autonomy and authority at the individual level (Wolff 1970: 12–19). Applying an analogous principle at the collective level, Lafont (2019: 3 and passim) offers ‘blind deference’ as the antithesis of the ideal of democratic self-government.

Blind deference’—doing or believing something just because someone else says to—is very much a limiting case, however. When on a sinking ship, I do as the captain commands not just because he says so but also because of the importance of an orderly evacuation for everyone’s safety. When I do what my doctor or lawyer tells me to do, it is not just because they say so but also because they are probably right in what they say.2 In a raft of quite common cases, we do what someone else says but not just because they say so. Insofar as we defer to them at all, we do so only partially and certainly not blindly.

Does that still count as deference? Maybe not. Maybe to be deference at all, it must necessarily be complete deference. One definition of ‘to defer’ is ‘to submit (a matter to a person, etc.) for determination or judgement’—to ‘refer’ the matter to that other person for decision, in a way that you yield entirely to that other person on that matter.3 The Oxford English Dictionary says that that usage is now ‘obsolete’. But its principal current definition of ‘deference’ nonetheless retains a crucial allusion to ‘submission’.4 Similarly, political and legal philosophers typically analyze ‘authority’ in terms of notions of ‘exclusionary reasons’ (Raz 1985 [emphasis in original) and ‘surrender of judgment’ (Flathman 1980, emphasis added)—phrases that similarly suggest that, in deferring to the authority of another, one necessarily does so completely.

The point nonetheless remains that people can, and often do, do what others tell them to do in part (but only in part) because they tell them to do so. For convenience, I shall continue to refer to cases like that as ‘partial deference’. But nothing turns on the terminology: substitute for that any other form of words you prefer.

The larger issue is the moral one. Does ‘autonomy’ necessarily preclude any amount of ‘deference’ whatsoever? Maybe. But even if any deference whatsoever disqualifies you from being fully autonomous, there may nonetheless be gradations within the category of non-autonomous action. We deem you to be more or less autonomous, depending on the extent to which you ‘defer’ to others.5 We deem your failure to be autonomous as more or less morally bad, depending on those considerations.

In any such comparative moral assessments of cases of partial deference, there is one consideration in particular that inevitably, and rightly, looms large. That is the extent to which you exercised your own judgement in deciding to defer to the judgement of others.

Go back and reexamine my earlier examples of partial deference in this light. I defer to orders of the captain in evacuating the sinking ship, because in my judgement we need to coordinate our escape and in my judgement the captain is the most salient person whose commands are most likely to provide effective coordination. I defer to the orders of my doctor and lawyer because I judge them to have expert knowledge that I lack and that I judge to be relevant to the medical and legal decisions before me. In all those cases, I have decided to defer to those others, and I based that decision in some (perhaps large) part on my own judgement concerning my values and interests and how they are likely to be best furthered.

Now transpose those familiar points about personal autonomy to the social and political realm. The great lesson of social epistemology is that the bulk of our knowledge is second-hand knowledge, acquired from others (Goldman 1999). And that is inevitably so for boundedly rational individuals, limited as we invariably are in our time and attention (Simon 1982). What we believe and what we value, and what we decide to do on the basis of those beliefs and values, is powerfully shaped by inputs from others.

How can individuals who are limited in these ways, as all of us are, nonetheless make rational decisions for themselves? The standard answer among political scientists and psychologists is that they take ‘cues’ from other sources that they trust (Lupia & McCubbins 1998). Those might be trusted individuals, or political parties, or trade unions or trusted media outlets. If I trust them, and they tell me something, then I can rationally do or believe what they suggest without checking for myself the underlying grounds for their recommendation. That is how ‘low information rationality’ works (Goodin & Spiekermann 2018: ch. 12). But note well: ‘low information’ is not ‘no information’. It is only rational for people to take cues from those others, and to defer to them in that way, insofar as they have independent grounds for trusting that their recommendations will (at least typically) represent their own values and deeper beliefs.6

This is not only how boundedly rational individuals come to their own judgements. It is also how boundedly rational individuals come to collective judgements. Tanasoca (2020) shows how deliberation in the real world essentially involves serial interaction among interlocutors a few at a time; cumulated across the entire social network, that comes to constitute ‘public opinion’.7 In such processes, people acquire information from others that they critically assess according to their own lights and incorporate into their own belief sets as appropriate.

In one sense, when taking account of the views of trusted others in this way, people are failing to act as full independent, autonomous agents. Their independence and autonomy are compromised by the facts that they take cues from one another; they read the same newspapers and blog posts as one another; they are bombarded by messages from the same would-be opinion leaders and so on. But as long as the opinion leaders and so forth do not lead them by the nose—as long as individuals preserve their capacity to exercise independent critical judgement in deciding whether or not to accept what they have been told—they are epistemically ‘independent enough’ for political purposes.8

While Lafont is clearly dead set against ‘blind deference’, she also clearly does not regard all deference as necessarily being objectionably blind. But from her few scattered remarks on the subject it is difficult to piece together exactly what (certainly what all) she thinks might make deference unobjectionable, or anyway significantly less objectionable. At various points it seems as if the crucial consideration, for her, is whether citizens have the opportunity to contest the decisions of those to whom they defer.9 As Lafont (2019: 8) says in what seems to be her most general, official pronouncement on the matter, ‘to the extent that citizens maintain some control over these actors [to whom they defer], they are not doing so blindly. By contrast, deference is blind if there is no such capacity for control.’10

Now, I fully appreciate the democratic importance of the ruled having capacity for some sort of control over the rulers. Having that capacity is arguably the defining feature of a democracy. But I baulk at treating having that capacity, alone, as the defining feature for distinguishing ‘non-blind’ from ‘blind’ deference.11 To be sure, if people do successfully exercise control over their rulers, then they will not be blindly following rules made by others but instead following rules made, to some extent, by themselves.12 But that is merely to say that ‘democratic control of rulers by the ruled’ requires both having ‘capacity for control’ and ‘non-blind deference’ in exercising those capacities. While both are required, it is clearly wrong to conflate the one with the other. They are separate requirements.

A far better way of distinguishing non-blind from blind deference, to my mind, would be to say that the former involves individuals deferring to others on the basis of their own critical assessment of the concordance between their own values and beliefs and those of the people (or embedded in the institutions) to which they defer.13

That has various highly desirable effects, as Lafont rightly observes.14 But what really matters, purely from the perspective of ‘blind deference’, is that deference of this sort works crucially through the critical judgement of the person doing the deferring. In deferring to others in this way, you are in some important sense ‘making their judgement your own’. That is what is involved in ‘deferring non-blindly’ to some other person or process. And as I have said, in real-world politics we do that all the time.

Now let me note that these sorts of considerations may well vindicate some of the ‘democratic shortcuts’ that Lafont is keen to dismiss. I hasten to add that they will do so only insofar as citizens themselves internalise the reasoning I am about to sketch—there is no justification, in anything I am about to say, for demanding deference to the shortcuts from people who do not themselves feel the force of the following arguments and themselves accede to them.15

Take first the case of deferring to the opinion of your better-informed counterpart in a Deliberative Poll (Fishkin 1991: 81).16 You may quite reasonably judge it to be a good idea to do what your better-informed self would do, and to take the pattern of preference changes during a Deliberative Poll as representative of that.17 Consider, for example, the Deliberative Poll in the run up to the 1999 referendum on Australia’s becoming a republic. A huge proportion of people who went into the event opposed to a president being elected by two-thirds of Parliament (calling that a ‘politician’s republic’) changed to supporting that over direct election of the president (on the grounds that only political parties would be able to mount a nationwide campaign of that sort, making political hack being elected president more likely than in a two-thirds vote of Parliament, where no party has ever enjoyed such a supermajority) (Fishkin 2018: part IV, sec. 7). Once non-participants learned that outcome from the Deliberative Poll, from watching the Sixty Minutes TV programme reporting on the event, for example, it would be perfectly reasonable for them to update their own views accordingly.18

Second, take the case of deferring to the will of the democratic majority. Suppose that, after due study and reflection, you are convinced of the validity of Taylor’s (1969) and Rae’s (1969) proof that, under certain conditions, majority rule is the decision rule that uniquely maximises the frequency of your getting your way (i.e., having your beliefs and values being enacted into law).19 Suppose you believe those conditions are satisfied in your electorate, and you therefore decide to adopt the rule of abiding by the majority verdict (in hopes that others do likewise).20 You are then engaged in the project of democratic self-governance on two levels: you (together with others) decided the community’s decision rule, and you (together with others) provide the inputs that produce community decisions under that rule.

Finally, suppose you have read Estlund’s (2008) book and been persuaded by his defence of ‘epistemic proceduralism’, and you come to believe that majority voting is the procedure that is most likely to yield epistemically correct conclusions. Suppose further that you think that ‘being correct’ (at least on matters of fact that are centrally relevant) is the preeminent virtue in social decisions. And suppose you resolve to adhere to the majority’s verdicts accordingly. Once again, you are then engaged in the project of democratic self-governance on those same two levels as above.

None of those count as cases of ‘blind deference’ in the terms I have been defending. Instead, you exercise your own independent judgement in deciding to defer in those ways. And you defer, insofar as you do, on fundamentally substantive grounds, which is to say, those procedures in your view are the most likely to yield outcomes you have good reason to believe to be correct (variously understood). Thus, your deference is not then ‘blind’ at all, in the way that matters.

III. Backing Off Full Endorsement

I opened by complaining that Democracy Without Shortcuts offers an impossible ideal (full endorsement) combined with an impoverished set of alternatives to it (blind deference). So far I have shown there are many alternative ways of deferring that are not ‘blind’. Next I shall show that even Lafont herself is willing to accept something less than full endorsement, in effect endorsing some of the alternatives that I just discussed.

Cristina Lafont thinks that the ‘full endorsement’ of laws and policies that the democratic ideal requires can be secured only through the direct participatory engagement of all members of the community. But it is not altogether clear whether that standard of democratic legitimacy is supposed to be a moral or merely a sociological one. When Lafont (2019: 4 and on many subsequent occasions) talks, as she often does, of ‘identification with’ and ‘alienation from’ the laws and policies of the community, it seems as if the standard is a sociological or psychological one—a matter of people’s feelings towards those laws or policies.

Clearly, such subjective perceptions morally matter in all sorts of ways. Clearly it is better for people to conform to laws and policies voluntarily than it for them to be coerced into doing so (4 and on many subsequent occasions). We want our interlocutor ‘to endorse the policy as reasonable upon reflection, so that he can identify it as his own and comply with it on its own accord’ (167). But, equally, surely it is morally better for people to voluntarily comply for good reasons, rather than because they have been hoodwinked into doing so by some cleverly co-optive feature of institutional design, or whatever (Saward 1992).21 Here, therefore, I shall be focusing on the quality of those reasons rather than the sheer sociology or psychology of the phenomenon.

Now, there are multiple modes of reason-giving and multiple standards of what counts as ‘good enough’ along those lines (Goodin 2018). For certain purposes (e.g., a ‘rationality review’ in administrative law), we might be satisfied with a demonstration that ‘there exist reasons’, whether those reasons were the actual ones that motivated the actor or whether those reasons would be accepted as good reasons by the enquirer. For other purposes, we are crucially concerned with the real reasons that actually motivated someone’s actions. It is clear that the sorts of reasons Lafont is looking for deliberation to provide are (ideally) reasons that the people being given the reason could endorse as ‘their own’—or anyway (minimally) as reasons that those people could see to be ‘reasonable’, even if they do not themselves fully endorse them.22

A. Beyond Full Endorsement in Ordinary Politics

While securing everyone’s ‘full endorsement’ of the laws and policies being proposed is clearly ideal, and Lafont spends the great bulk of her book reiterating that ideal, at crucial points she also clearly recognises that that is an unattainable ideal in any real-world polity. Accordingly, Lafont (2019: 12–13) advocates ‘an institutional approach’ according to which ‘democratic legitimacy does not require every single person to agree on the reasonableness of each coercive law to which they are subject at any given time’.

What is crucial for Lafont is for:

institutions to be in place such that citizens can contest any laws and policies … by asking that either proper reasons be offered for them or that they be changed. To the extent that such institutions are available to all citizens, even to those who happen to find themselves in the minority, they can see themselves as equal members of a collective political project of self-government. (Lafont 2019: 12)

What is crucial, in other words, is ‘the existence of effective rights to political and legal contestation that allow them to trigger a process of public justification for the reasonableness of any policies that they find unacceptable’. (Lafont 2019: 13)

That does not mean that everyone in the end actually come to endorse the policies. Of course, it would be ideal if they did.23 But all that is minimally required is for them to comes to see the policies as ‘reasonable enough’ that they not ‘actively oppos[e] them’ (Lafont 2019: 174).24

In the end, therefore, it is not actually the substance of the policy proposals that people need to endorse, after all. Lafont would ideally like the full community to fully endorse any proposal that is enacted. But she accepts that, at least as regards some enactments, she will need to make do with the less-than-full endorsement of at least some parts of the community. Not only will she sometimes have to accept weak rather than strong endorsement (‘it’s reasonable’ rather than ‘it’s right’) but sometimes she will have to accept people’s simply saying ‘it’s not unreasonable’ and ‘I won’t actively oppose it’. And that hardly sounds like any sort of endorsement of the outcomes at all.

Rather, it turns out to be the process by which proposals come to be adopted that people need to endorse. So long as that process contains institutional guarantees that people unpersuaded by the policies can express their continuing dissent, and those advocating them make a good-faith effort at assuaging their concerns, Lafont’s hope-cum-expectation is that people will be prepared to see the policies as ‘reasonable enough’—and thus to ‘acquiesce’ to them, if not to fully endorse them.25

In the first instance, that once again seems to be an empirical sociological–psychological proposition. But why should people go along in that way with policies to which they still object, just so long as their objections have been heard and overruled? I am unsure of Lafont’s answer, but here is mine.

People should see such procedures as being most likely, over the long haul, to enable them to get what they want given their own interests and values. That is mechanically so with majority rule, as Taylor’s and Rae’s proof has demonstrated. And it is not too much of a stretch to imagine that the same might be discursively true as well, for much the same reasons. Insofar as people actually do decide to be bound by the outcomes of fair deliberative procedures in that way, the procedures could be said to have their ‘full endorsement’ even if each and every substantive outcome might not.26

Note well what would have happened on that account. Lafont’s deliberative institutions will have come to satisfy the ‘democratic ideal of self-government’ just because (and just insofar as) people exercise their own critical judgement in placing their trust in those institutions (and, we might add, they have good reasons for doing so). But that is not then importantly different from the warrant that can be claimed by various other forms of ‘partial deference’ in which people exercise their own critical judgement to place their trust (with good reason) in other agents and agencies.

B. Are the Politics of ‘Judgements of Justice’ any Different?

That is how Lafont seems to see things working for ‘ordinary politics’, where all that is at stake are people’s ‘conceptions of the good’. She would impose different standards, however, when people’s ‘sense of justice’ is at stake. She variously refers to those as cases involving what people see as involving ‘fundamental rights and freedoms’, or what Rawls dubbed ‘constitutional essentials’ (Lafont 2019: 20–21). In ordinary sorts of cases, Lafont accepts that ‘no society can affirm all values and ways of life simultaneously’; and she is prepared in those cases to accept something less than full endorsement of every citizen to every law and policy. But, she writes, ‘it is a different situation when the laws and policies to which citizens are subject fail to conform to their judgments about justice’ (21).

If the claim here were merely that people’s fundamental constitutional rights should not be up to a majority vote, then that would be a relatively unexceptionable claim. That is the interpretation suggested by Lafont’s talk of ‘fundamental rights and freedoms’ and indeed ‘constitutional essentials’. That reading is reinforced by Lafont’s recurring references to ‘judicial review’—appeal to the counter-majoritarian institution of the courts—as an important part of the solution in such cases.

It seems that that is not all that Lafont means, however. In the passage quoted above, Lafont identifies the target class of cases as being ones in which ‘the laws and policies to which citizens are subject fail to conform to their judgments about justice’ (21; emphasis added). That seems to suggest that what count as ‘fundamental rights and freedoms’ is, for Lafont, not objectively fixed (in the constitution, for example) or even socially fixed (by widespread agreement across the community). Instead, it seems to be determined by each individual’s own judgements about justice. ‘Fundamental rights and freedoms’ are, apparently, whatever anyone thinks they are.27

There are all sorts of problems, both principled and pragmatic, with that approach. Not least of them is that it constitutes a licence to ‘print political money’, entering all sorts of trumped-up claims by simply dialling up the ‘fundamentality’ that you say you attach to those claims.28

Just how much of a problem that will be depends purely upon just what special powers and privileges are conferred upon people claiming that their ‘sense of justice’ is being violated. At one extreme, pressing a ‘fundamental rights’ claim of that sort in a court might literally invalidate legislation. In that case, the ‘printing political money’ objection has real bite. More unexceptionably, it might amount to a claim for there to be periodic elections where the outcome of the last election can be recontested at the next. Most innocuously, it might amount to a mere request to ‘say more’ in defence of the outcome than the majority endorsed. In the latter case, the ‘printing political money’ objection is cancelled, but only by making the procedure almost purely symbolic.

Lafont (2019: 22) writes that ‘only a democratic political system in which citizens can participate in shaping the laws and policies to which they are subject, can ensure that these laws and policies conform to their judgments about justice’. But where citizens have incompatible views as to what justice requires, they cannot ensure any such thing. Among people with incompatible views about justice, the social decision will favour some people’s views over others. Or else it will refrain from imposing any view—which is, of course, to go against everyone’s (differing) views of what should be done (Goodin & List 2006).

Even as regard matters of ‘justice’—of ‘fundamental rights and freedoms’—Lafont is thus once again forced to accept something less than full positive endorsement of the laws and policies from the full society. She requires only that citizens ‘endorse the laws and policies they are bound to obey as just or at least as reasonable’, once again (Lafont 2019: 21; emphasis added). In a passage adjacent to (and hence presumably continuous with) her discussion of matters of ‘justice’, Lafont adds that:

a perfect alignment [between laws and policies and everyone’s sense of justice] could [n]ever be reached. Thus, the ideal can be both feasible and action-guiding [only] if it is understood to require democratic institutions and practices to provide citizens with as many (effective) opportunities as possible in order to prevent a permanent disconnect between the policies to which citizens are subject and their considered opinions and will. (Lafont 2019: 23)

Hence it seems that, even as regards matters of ‘justice’ and of ‘fundamental rights and freedoms’, Lafont is willing to accept something far short of full endorsement of all the laws and policies by the full community. She exhorts us to keep talking with those who disagree with the existing laws and policies, particularly when they tell us that those offend their ‘sense of justice’. But apparently talking is enough. We do not need to wait until they have been talked around in order to legitimately enact and enforce those laws and policies.

IV. Conclusion

The headline claims in Democracy Without Shortcuts are that we should require ‘full endorsement’ of laws and policies by everyone subject to them, and we should eschew all shortcuts that demand citizens’ ‘blind deference’ to the judgement of others in lieu of that. But full endorsement is an impossible ideal, and blind deference is not the only alternative. Reading below the fold, it turns out that even Lafont agrees.


  1. I am grateful to Cristina Lafont for much discussion of these issues. I am also grateful for feedback from Ana Tanasoca and two anonymous referees.
  2. These examples come from Wolff (1970: 15–16).
  3. OED 1989, qv. ‘defer, v.2’, def. 3.
  4. ‘Submission to the acknowledged superior claims, skill, judgement, or other qualities, of another’ (OED 1989: qv. ‘deference, n.’, def. 2).
  5. That might depend on how heavily you weigh reasons emanating from others’ say-so in your decision calculus. Or it might depend on how often or on how wide a range of decisions you let yourself be influenced by such reasons (Wolff 1970: 15).
  6. Warren (2020) makes much the same point. Perhaps this is what Lafont (2019: 9, n. 24) means to telegraph when saying, ‘when I defend a “democracy without shortcuts” I mean specifically a democracy “without shortcuts that bypass the citizenry by requiring citizens to blindly defer to the decisions of others.” I do not oppose citizens using heuristics as shortcuts in general (e.g. deferring to representatives, political parties, organizations, and so on).’ As she says later in the book (126), people ‘might cast their votes … relying on the recommendations of groups whose political views they share or on experts whose judgments they have reasons to trust. However imperfect, this type of deference is at least not blind …’ [original italics]. See further MacKenzie and Warren (2012).
  7. Lafont (2019: 165–166) posits something very much like that in her ideal of social deliberation, modeled on her arguments with her teenager over texting while driving.
  8. Understood as ‘independent enough for their aggregated opinions to be more likely to be correct than their individual opinions’: Dietrich and Spiekermann (2013), Goodin and Spiekermann (2018: 60–62, 67–82) and Hawthorne (2001: sec. 3.3). Bächtiger and Goldberg (2020) argue similarly for citizens to critically assess their trusted information sources.
  9. Speaking of representative democracy, for example, she says: ‘citizens can defer a lot of political decision-making to their representatives so long as they are not required to do so blindly’; and she goes on to say that that criterion is satisfied ‘so long as there are effective and ongoing possibilities for citizens to shape the political process as well as to prevent and contest significant misalignments between the policies they are bound to obey and their interests, ideas, and policy objectives …’ (Lafont 2019: 23). Or again, speaking of judicial review, Lafont (13) says: ‘By securing citizens’ right to legal contestation judicial review (whether national or transnational) offers citizens a way to avoid having to blindly defer to the decisions of their fellow citizens. This is the case insofar as it sets up an institutional venue where they can call their fellow citizens to account by effectively requesting that proper reasons be offered in public debate in order to justify the laws and policies to which they all are subject.
  10. Lafont (2019: 127) subsequently reiterates the point: ‘It is an essential feature of political representation that those represented defer to their representatives. However, they are not supposed to do so blindly. Some level of control over the representatives must be kept by those they represent in order for the latter to count the former as their representatives at all’ (original italics).
  11. That is better seen as a test for whether deference is democratically acceptable, not of whether it is ‘blind’ (Warren 2020).
  12. Of course if successful control is exercised pre-emptively in selecting a representative who embodies my own values and beliefs, then subsequently I can defer to her blindly and I need no subsequent opportunity to exercise control over (e.g., to sanction) her. That is because, ‘in following her own convictions, she also reflects my own’ (Lafont 2019: 128; see similarly Brennan 1996; Mansbridge 2009; Miller & Stokes 1963).
  13. Both in deferring to other individuals and in deferring to institutions and procedures, we are doing so on the grounds that they are likely to produce outcomes in line with our own beliefs and values. For that purpose, it makes no difference whether this concordance is produced through human agency (the like-minded other individual acting on her own similar beliefs and values) or more mechanically through the workings of institutional or procedural mechanisms.
  14. One is helping to ensure that, as Lafont (2019: 8) says, ‘the political decisions endorsed by the agent to whom one is deferring are those that one would have endorsed if one had thought through the issue with access to the relevant information’. Another is helping to avoid ‘a permanent misalignment between the beliefs and attitudes of the citizenry and the laws and policies to which they are subject …’
  15. Lafont might say that they should not, on the grounds that to do so would be to violate the ‘ideal of democratic to self-government’. But I cannot see how, for example, adopting a belief that public policy should track the truth is in any way inimical to that ideal—any more than my deferring to my oncologist’s therapeutic recommendations is inimical to the corresponding ideal of personal autonomy.
  16. Note, however, that ‘your better informed counterpart’ is not any particular individual taking part in the minipublic but, rather, what you can infer from the deliberations of the minipublic how your own views would have changed had you been a member of it.
  17. Lafont (2019: 123, 128) is absolutely right that that does not vindicate literally deferring to the decision of the minipublic, the majority of which may well have been composed of people who were in no sense your counterparts. But if a large proportion of the people who shared your view going into the event came out of it with a different view, you may reasonably suppose that you may well have done likewise.
  18. That is a weak sense of ‘deferring to’ the Deliberative Poll; it may be more aptly described simply as ‘being informed by’ it. Still, that is all Fishkin proposes. Bächtiger and Goldberg (2020) usefully elaborate on this suggestion.
  19. Even if you have no reason to suppose that the majority to whom you defer on any given occasion ‘shares your deeper beliefs and fundamental values’, this proof shows that everyone ‘deferring to the will of the majority’ maximises the probability that the ‘deeper beliefs and fundamental values’ of each will be reflected in the laws that are enacted.
  20. ‘Adhering for now’, anyway: that is not to say you will not try to overturn the majority verdict the next time the matter comes up for a vote.
  21. Wolff (1970: 18–19) likewise notices this psychological sense of ‘ownership’ of the laws, writing: ‘When I take a vacation in Great Britain I obey its laws, both because of prudential self-interest and because of obvious moral considerations concerning the value of order, the general good consequences of preserving a system of property, and so forth. On my return to the United States, I have a sense of re-entering my country, and if I think about the matter at all, I imagine myself to stand in a different and more intimate relation to American laws. They have been promulgated by my government, and I have a special obligation to obey them’ (original italics). But, as Wolff goes on to admit, ‘my feeling is purely sentimental and has no objective moral basis’.
  22. The latter formulation is suggested, for example, in Lafont’s (2019: 174; emphasis altered) remarks that ‘the project of self-government’ involves advancing ‘considered judgments … that others who will be subject to them can also find acceptable or at least reasonable enough to act accordingly instead of actively opposing them’. Elsewhere she speaks of citizens ‘not being forced to blindly defer to political decisions made by others that they cannot reflectively endorse as at least reasonable’ (177; emphasis altered; see similarly 193) and says that ‘the democratic ideal of treating each other as free and equal depends upon being committed to convincing one another of the reasonableness of political decisions to which we are all subject’ (3–4; emphasis added).
  23. And if they endorsed the policies for the same reasons: Lafont (2019: 214, n. 45) clearly regards ‘convergence’ models in which people endorse the same policies but for differing reasons as not good enough. Lafont’s (13) ideal is thus for the ‘processes of opinion- and will-formation to be structured in such a way that disagreements can be reasonably overcome among citizens with very different views, interests, attitudes, and so on’.
  24. As she writes elsewhere, ‘The ideal that one should not be subject to laws that one cannot see oneself as an author of … seeks to avoid being coerced into obeying laws that one cannot endorse as at least reasonable upon reflection. Avoiding sheer coercion does not require that one literally be an author of the laws, but it does require that one can obey them based upon insights into their reasonableness’ (Lafont 2019: 18; emphasis added; see similarly 21).
  25. Lafont (2019: 174) writes, ‘no one should unilaterally impose her views on others without first trying to persuade them of their reasonableness by offering reasons and considerations that they too can reasonably accept … The fact that some consideration matters to citizens, even if they are in the minority, means that such considerations must be engaged in public deliberation and properly responded to with counterarguments by those who reject it on their merits, instead of simply being ignored by those who happen to be in the decisional majority … This in turn implies that public deliberation must focus on actual views, interests, and policy objectives of democratic citizens, however wrongheaded they may seem to those who disagree with them.’
  26. It is an interesting question, not addressed in Lafont’s book, what the ‘closure’ rule should be. Of course, formally, any one sovereign legislature should be able to overturn the decision of any previous sovereign legislature, and electorates likewise. But there are clear costs to ‘keeping everything up in the air’, and correspondingly good pragmatic arguments for ‘getting closure on some questions, at least for a time’ (Goodin 2012; Holmes 1988).
  27. At least for the purposes of legitimately initiating the process of contesting a previous majoritarian decision. Lafont may want to go on to say that ‘of course litigants do not have a right to the outcome that they want’. But that is to give with one hand and take away with the other: yes, they can get their case heard; but the ruling may still go against them. ‘Thanks for nothing’ would be a natural response to that. And it would be little consolation to say, as Lafont would, ‘but you can keep trying, doing it all over and over again’.
  28. Barry (1965: 245–249) offers similar arguments about the vulnerability of unanimity rules to people deploying an ‘offensive veto’.

Competing Interests

The author has no competing interests to declare.


1 Bächtiger, A., & Goldberg, S. (2020). Towards a more robust, but limited and cogent defense of the political uses of deliberative minipublics. Journal of Deliberative Democracy, 16(2), 33–42. DOI:  http://doi.org/10.16997/jdd.390

2 Barry, B. (1965). Political argument. London: Routledge & Kegan Paul.

3 Brennan, G. (1996). Selection and the currency of reward. In R. E. Goodin (Ed.), The theory of institutional design (pp. 256–275). Cambridge: Cambridge University Press. DOI:  http://doi.org/10.1017/CBO9780511558320.011

4 Dietrich, F., & Spiekermann, K. (2013). Independent opinions? On the causal foundations of belief formation and jury theorems. Mind, 122(487), 655–685. DOI:  http://doi.org/10.1093/mind/fzt074

5 Estlund, D. (2008). Democratic authority. Princeton, NJ: Princeton University Press.

6 Fishkin, J. S. (1991). Democracy & deliberation. New Haven, CT: Yale University Press.

7 Fishkin, J. S. (2018). Democracy when the people are thinking. Oxford: Oxford University Press. DOI:  http://doi.org/10.1093/oso/9780198820291.001.0001

8 Flathman, R. E. (1980). The practice of political authority. Chicago, IL: University of Chicago Press.

9 Goldman, A. I. (1999). Knowledge in a social world. Oxford: Clarendon Press. DOI:  http://doi.org/10.1093/0198238207.001.0001

10 Goodin, R. E. (2012). On settling. Princeton, NJ: Princeton University Press. DOI:  http://doi.org/10.1515/9781400845316

11 Goodin, R. E. (2018). Reasons for reason-giving: The Obamacare debates. Journal of Health Politics, Policy & Law, 43(6), 901–1,012. DOI:  http://doi.org/10.1215/03616878-7104405

12 Goodin, R. E., & List, C. (2006). Special majorities rationalized. British Journal of Political Science, 36(2), 213–242. DOI:  http://doi.org/10.1017/S0007123406000135

13 Goodin, R. E., & Spiekermann, K. (2018). An epistemic theory of democracy. Oxford: Oxford University Press. DOI:  http://doi.org/10.1093/oso/9780198823452.001.0001

14 Hanrahan, R., & Antony, L. (2005). Because I said so: Toward a feminist theory of authority. Hypatia, 20(4), 59–79. DOI:  http://doi.org/10.1111/j.1527-2001.2005.tb00536.x

15 Hawthorne, J. (2001). Voting in search of the public good: The probabilistic logic of majority judgments. Retrieved from https://www.researchgate.net/profile/James_Hawthorne2/publication/228917887_Voting_in_Search_of_the_Public_Good_The_Probabilistic_Logic_of_Majority_Judgements/links/00b7d51bb34fb364a5000000.pdf (accessed 17 July 2020).

16 Holmes, S. (1988). Gag rules, or the politics of omission. In J. Elster & R. Slagstad (Eds.), Constitutionalism and democracy (pp. 19–58). Cambridge: Cambridge University Press. DOI:  http://doi.org/10.1017/CBO9781139173629.002

17 Lafont, C. (2020). Democracy without shortcuts: A participatory conception of deliberative democracy. Oxford: Oxford University Press. DOI:  http://doi.org/10.1093/oso/9780198848189.001.0001

18 Lupia, A. W., & McCubbins, M. D. (1998). The democratic dilemma: Can citizens learn what they need to know? Cambridge: Cambridge University Press.

19 MacKenzie, M. K., & Warren, M. E. (2012). Two trust-based uses of minipublics in democratic systems. In J. Parkinson & J. Mansbridge (Eds.), Deliberative systems (pp. 95–124). Cambridge: Cambridge University Press. DOI:  http://doi.org/10.1017/CBO9781139178914.006

20 Mansbridge, J. (2009). A ‘selection’ model of political representation. Journal of Political Philosophy, 17(4), 369–398. DOI:  http://doi.org/10.1111/j.1467-9760.2009.00337.x

21 Miller, W. E., & Stokes, D. E. (1963). Constituency influence in Congress. American Political Science Review, 57(1), 45–56. DOI:  http://doi.org/10.2307/1952717

22 OED. 1989. Oxford English Dictionary. Oxford: Oxford University Press.

23 Rae, D. W. (1969). Decision-rules and individual values in constitutional choice. American Political Science Review, 63(1), 40–56. DOI:  http://doi.org/10.1017/S0003055400261467

24 Raz, J. (1985). Authority and justification. Philosophy & Public Affairs, 14(1), 3–29. DOI: https://www.jstor.org/stable/2265235

25 Saward, M. (1992). Co-optive politics and state legitimacy. Aldershot: Dartmouth.

26 Simon, H. A. (1982). Models of bounded rationality. Cambridge, MA: MIT Press.

27 Tanasoca, A. (2020). Deliberation naturalized: Improving real-existing deliberative democracy. Oxford: Oxford University Press.

28 Taylor, M. (1969). Proof of a theorem on majority rule. Behavioral Science, 14(3), 228–231. DOI:  http://doi.org/10.1002/bs.3830140307

29 Warren, M. E. (2020). Participatory deliberative democracy in complex mass societies. Journal of Deliberative Democracy, 16(2), 81–88. DOI:  http://doi.org/10.16997/jdd.395

30 Wolff, R. P. (1970). In defense of anarchism. New York: Harper & Row.