Wednesday 4 August 2010

Does The Spirit Level refute Rawls?

The Spirit Level, by Richard Wilkinson and Kate Pickett, has been getting a lot of discussion of late. Most of the discussion has focused on the empirical claims it advances. But over at OurKingdom, Rupert Read has a post exploring what he sees as the philosophical implications of the book (responding to Gerry Hassan's review ).

According to Read The Spirit Level empirically refutes John Rawls's political philosophy, specifically Rawls's 'difference principle.' He tells us that he's talked to Richard Wilkinson about it, and Wilkinson agrees that the findings in The Spirit Level do indeed provide an empirical refutation of the 'difference principle'.

Read says:

'For me as a philosopher, the thing about The Spirit Level that is most exciting is that as a study of the pervasive harms of inequality it strongly suggests that John Rawls's 'difference principle', which says that inequalities are OK provided that they materially benefit the worst off, a principle that has dominated political philosophy for 40 years, is simply wrong. Empirically wrong.

Which means that put into action ‘the difference principle’ will create a worse society, across a whole index of measures. Perhaps surprisingly, it will make virtually everybody, and certainly the worst off, worse off. (Or at least: worse off than they could be if an alternative way of ‘organising inequalities’ – a more egalitarian way - were settled upon.) Even if they have more money or more things (are ‘materially better-off’), this will not translate into an improved quality of life: on the contrary.

In sum: it is now possible for the first time to show that the difference principle (and, by extension, liberal political philosophy whether or not of the ‘trickle-down’ variety) makes the worst-off on balance worse off, and this can I think reasonably be taken to constitute an empirical refutation of the claim that it could possibly be just.'

My first thought is that there is something intrinsically odd in the idea that the difference principle can be 'empirically wrong'.

Rawls's difference principle is a normative principle. As Read says, the principle holds that inequality is justified only if it works to improve the position of those who are worst-off under the ensuing inequality. Inequalities aren't justifiable merely because they increase overall GNP or overall welfare, or because they raise the average level of income or welfare (or whatever else one thinks matters fundamentally from a distributive point of view): they must improve the position of the worst-off.

However, the difference principle does not say that inequalities are justified because they work to the benefit of the worst-off. It does not assert the truth of any 'trickle-down' theory.

The principle is entirely agnostic, as a normative principle, on the empirical question of how far economic inequalities do or don't, in fact, work to the benefit of the worst-off. It just says that, if they are to be justified, then they must work out this way.

It is mysterious, then, as to how the principle could be empirically refuted. To be refuted in this sense, it would have to have make, or rely, on a specific empirical claim about how inequality actually improves the position of the worst-off. But, as just explained, the principle is entirely agnostic on such empirical matters.

As Read points out, the evidence marshalled in the The Spirit Level suggests that high levels of economic inequality make the position of the worst-off even worse. If this is the case, then the difference principle supports an unequivocal condemnation of these inequalities...because, according to the evidence, they worsen the position of the worst-off.

But have I perhaps missed Read's key point?

In the second and third paragraphs of the above quote, Read rightly notes the distinction between looking at the position of the worst-off in narrowly materialist terms (income and wealth) versus looking at it in a broader way. And in his first paragraph he specifically defines the difference principle as requring that inequalities 'materially benefit the worst off'. So perhaps the argument he has in mind goes something like this:

(1) the difference principle calls for inequality in income and wealth to be set at whatever level works to maximise the income and wealth holdings of the group which has least income and wealth;

(2) this principle would call for a quite high level of income/wealth inequality;

(3) The Spirit Level shows that societies with this level of income/wealth inequality - the level of inequality called for in (1) above - leave the worst-off worse off in an overall sense: although they have more income and wealth than in more materially equal societies, this is cancelled out by their having much worse psychological well-being, e.g., diminished self-respect;

(4) therefore, The Spirit Level gives us reason to reject the difference principle.

However, as a basis for rejecting the difference principle, this has two problems.

First, even if we accept as accurate the narrowly materialistic characterisation of the difference principle in (1), it is not obvious that this principle would in fact mandate such a high level of income/wealth inequality as to have the negative effects charted in The Spirit Level. Once again, the principle as such is agnostic on the empirical question of how much inequality actually is needed to satisfy the principle. It could be that the principle, even as understood in (1) above, is satisfied only in societies with very high levels of equality in income/wealth - societies which escape the negative effects of inequality discussed in The Spirit Level.

But more importantly, the difference principle is not simply about income and wealth. The principle, as defined by Rawls, is based on a pluralistic understanding of citizens' interests which certainly includes income and wealth but which also includes what Rawls call the 'social bases of self-respect.'

The principle calls on us to assess the position of the worst-off - to conceptualise who is 'worst-off' - not only in terms of material goods but in terms of how social institutions affect self-respect. So even if an unequal distribution of income and wealth does improve the position of the worst-off as measured simply in material terms, this would not necessarily be justified, under the difference principle, if the material inequality also damages the self-respect of the worst-off. The kind of psychological harms from inequality that the authors of The Spirit Level are concerned to highlight are amongst the things that the difference principle calls on us to be alert for in judging how material inequalities affect the position of the worst-off.

In short, The Spirit Level does not refute Rawls's difference principle.

Rather, it provides evidence that societies have to be pretty egalitarian in economic terms if they are to satisfy this principle.

A final point. There is a danger for egalitarians in overinvesting in The Spirit Level. It makes a strong case that economic inequalities have negative effects on the well-being of a society (and not just its materially poorest members). But reservations about its claims are by no means confined to the right; see, for example, Claude Fischer's essay in Boston Review. And what if, in fact, further research showed that all of the book's claims are false? Would this imply that existing levels of income and wealth inequality are fine?

I certainly don't think so. That's because I think a great deal of existing material inequality is intrinsically unfair, independent of whether or not it lowers overall or average or majority well-being. And perhaps some of the reasons why such inequality is unfair are those set out in the work of John Rawls?


Captain Cabernet said...

A couple of quibbles. First re your exposition in terms of "the worst off", given that the DP is actually about the (lifetime expectations) of "the least advantaged", since LA are clearly not the same as WO (for option luck reasons). The other point to make is surely that Rawls's version of the DP is broadly resourcist whereas TSL favours something closer to a capability metric. To show that TSL refutes the DP, Read would also need to argue that Rawls's resourcism is mistaken for the purposes of *justice* (as opposed to other reasons we might buy into a measurement of well-being).

Stuart White said...
This comment has been removed by the author.
James Graham (Quaequam Blog!) said...

I have been having a similar argument with Rupert for a while now. I hope that Wilkinson and Pluckett haven't really set themselves 'against' the difference principle as that would be very silly.

At low levels of GDP, the Spirit Level appears to suggest that growth is a more significant factor than reducing inequality. Given that, I can't even begin to imagine how it can be said to refute Rawls.

Stuart White said...

Chris: thanks for the quibbles.

On your first point: I think I agree.

On your second point: I also agree. However, I think one can reject Rupert Read's criticism of the DP even without engaging the debate between resourcists and capability theorists. This is because Rupert mischaracterises Rawls's resourcism, so far as the DP is concerned, as resting on a narrowly materialist conception of citizens' interests. But Rawls is clear that things like the social bases of self-respect are relevant to assessing how advantaged people are, not just income and wealth. And I think this makes Rawls, even as a resourcist, sensitive to at least some of the concerns about things like psychological well-being that Wilkinson and Pickett focus on.

Anonymous said...

Great post, Stuart.

Anonymous said...

whoops, forgot to say, you may also be interested in David Runciman's review of TSL in the LRB a few months ago (fair, but quite critical of some of the statistical methodology):

Rupert said...

Hi Stuart!
Thanks for this. I agree with a fair bit of what you say here, in your long post responding to my short post (which was itself just a short response to a long piece that I had problems with my Gerry Hassan). I appreciate your taking the time to correctly reconstruct at some length a lot of what I was arguing in extreme concision. I have a forthcoming article where, at great length, I explain why I do not agree however with your conclusion.
So: sorry: this comment is really nothing more than a promissory note. I have stuff that I will send you offline that I believe counters some of your points here (the ones I disagree with); but I can't really publish it online here, because that would get in the way of my (hopefully-forthcoming peer-reviewed) piece...
I'll just say this:
Your post does not counter what was my central point: that the difference principle is designed primarily to deal with the problem of absolute poverty, but what W&P (and lots and lots of others - The Spirit Level is iconic, but it hardly stands alone! In fact, it is a kind of summary of loads of research by loads of people) show is that the problem of relative poverty is in fact the one that is paramount in Western democracies (the places that Rawls was writing for). (Yes, Rawls talks about the social basis for self-respect; but virtually every single one of his actual invocations of the DP and of those of his followers concern dollars.) If the difference principle introduces inequality in order to eliminate 'deprivation', then it will, W&P have in effect shown, worsen the societies in question, and _especially_ the lot of the worst off.
That is what I mean by using the (cutesy, deliberately-provocative) formulation that we might even say that Rawls's DP has been empirically refuted by W&P. That aiming for lifting up the base, rather than aiming for equality, worsens virtually everyone's lot, and certainly that of the base.

Adam Humphreys said...

Stuart - good post, I agree. TSL suggests that, at least in more developed societies, greater inequality is unlikely to make the worst off better off. This is entirely consistent with the DP, which is a normative principle stating an "only if" proposition.

Stuart White said...

Rupert: thanks for your thoughtful reply.

I look forward to reading the paper, but my initial response is that it is not correct to say that Rawls' concern, the concern behind the difference principle, is 'dealing with absolute poverty'.

Rawls starts from the idea that equality (in relevant primary goods) is the morally relevant baseline. The onus is justifying departures from this baseline. The difference principle is his answer to the question of how far departures are just. So the underlying concern is about when inequality is justified not preventing 'absolute poverty'.

And even if Rawls were concerned only with absolute poverty, there is the point that he would conceptualise poverty in a way that takes account of at least some of the psychological effects - specifically, damage to self-respect - of economic inequality that Wilkinson and Pickett and yourself are rightly concerned with.

Stuart White said...

Also: its not right to say that Rawls and his 'followers' only ever refer to the difference principle in terms of 'dollars'.

It is true that as an expositional device, particularly when first introducing people to the idea, some of us sometimes present the difference principle in that way. But this is an expositional simplification and one that any good writer about Rawls will flag up as such (as I do in my book on equality).

Rawls himself has many discussions of the 'index of primary goods' which concern the difference principle in which he talks about income, wealth, leisure time, and the social bases of self-respect. And there is something of a cottage industry in the Rawls literature exploring the implications of giving different weights to these different primary goods.

Rupert said...

Chris; the point is that W&P demonstrate that resourcism has disastrous capabilities effects. If Rawlsians want, they can still hold onto resourcism; but it is hardly plausible that doing so will leave them in (reflective) equilibrium... Resourcism is unattractive, once one realises that what really matters in inequality.

Rupert said...

S; it is in my view more than just an expositional device. In many key quotes in THEORY OF JUSTICE, it is all that gets discussed.
The 'social basis of self-respect' feels like an afterthought, by comparison.
I talk about this in my paper.

Rupert said...

you say 'inequality', I say 'absolute poverty'. I also say 'inequality'. It doesn't make any difference. This is I think a merely semantic point.

Anonymous said...

"you say 'inequality', I say 'absolute poverty'. I also say 'inequality'. It doesn't make any difference. This is I think a merely semantic point."

No. It is not "merely" a semantic point. The difference between "inequality" and "absolute poverty" is extremely important, and is (I have to say) rather a basic one. For a start, we could have extensive inequality in a society within which absolute poverty has been eradicated - they're quite simply two different concepts.

I struggle not to be rude here, but Stuart has highlighted - both in the OP and in the comment replies - that you're quite extensively confused about the issues you are purporting to deal with. This is not a case of you saying one thing, Stuart saying another, and you both having a reasonable disagreement. It's a case of you getting things wrong, Stuart highlighting that you are getting things wrong, and you not realising/accepting that your positions have been demonstrated to be unsustainable and/or false.

OK that came out pretty rude. I (sort of) apologise. But sometimes the truth is best told mercilessly.

Rupert said...

'Thanks', badconscience.
I think we [Stuart and me, and you] may have been subject to a linguistic misunderstanding here, that has erroneously given you the impression that I don't know what I'm talking about...
My point was that Rawls's concern with _preventing_ 'absolute poverty' (in the [limited] sense of reducing the level in material terms of people who are in my sense absolutely poor: see below) and with _justifying_ inequality - _these_ two projects are what I meant to be identifying, in my (overly) compressed comment to which you have taken such umbrage, badconscience.
But when I use the term 'absolute poverty', to be clear, I don't mean to be referring to an level of indigence that can be cross-culturally identified at some fixed amount. I mean simply to be referring to what Rawls is talking about. In other words: by 'absolute poverty' I simply mean the level of the worst-off. If one assumes that there are some poor in every society, this I think is an acceptable way of using the term.
So now, my comment to which you [badconscience] responded negatively may I now realise have been very unclear, in more than one way; I apologise for that. I was OF COURSE not saying that there was no difference between absolute poverty and degree of inequality [which would be a dreadful confusion]: I was saying there is no difference, if you use the terms the way I do, between a concern with alleviating the condition of the absolutely poor (the worst off) and justifying inequality.
I hope that has cleared things up somewhat. If it has, then I might deserve to not be a victim of rudeness, in future.
Anyway: this will teach me to write overly concisified comments; I have had to write all this, to try to clear things up!

Chris Brooke said...

*** The 'social basis of self-respect' feels like an afterthought, by comparison. I talk about this in my paper. ***

Of all the things you've said in this discussion, Rupert, this claim seems to me to be the most outlandish -- so if you could give any indication about what you say in the longer paper to flesh out this observation, I'd be quite grateful.

(The way I read Rawls, it wouldn't be too much of an exaggeration to say that pretty much the whole enquiry is motivated by the question, "what would have to be the case in order to secure the social bases of self-respect?" So if you're right, then my whole approach to reading Rawls is hopelessly wrong, and I'd be interested to know what the reasons are for thinking that it might be (beyond the appeal to your "feelings" when you read "Theory").

TJ said...


"But when I use the term 'absolute poverty', to be clear, I don't mean to be referring to an level of indigence that can be cross-culturally identified at some fixed amount
by 'absolute poverty' I simply mean the level of the worst-off. If one assumes that there are some poor in every society, this I think is an acceptable way of using the term."

Not really no. If the phrase "absolute poverty" is to mean anything it must refer to some *absolute* standard of poverty. Otherwise it wouldn't be an absolute measure.

What you are describing when you talk of "the level of the worst-off" is relative poverty, i.e. poverty relative to others in a particular society.

Rupert said...

No, TJ, that's certainly not right.
I obviously still haven't made myself clear. Sorry. Let me put it really simply and directly:
The question is whether you focus on the absolute level of income/whatever of the poorest (as Rawls does) or on that level relative to the level that others are at (which I and W&P do).

Rupert said...

Another point on 'the social basis of self-respect', in reply to Chris Brooke:
Basically, Rawls deals with this primarily through the first principle. This, in my view, is a very serious blunder: what W&P show particularly clearly is that the social basis of self-respect depends upon degree of suffered inequality, not on 'bourgeois freedoms'.
When it comes to the 2nd principle, Rawls spends very little time on the social basis of self-respect, relegating some discussion of it to later in the book. This is why I speak of an 'emprical refutation' of Rawls: W&P SHOW this to be a bad strategy.

Anonymous said...

"The question is whether you focus on the absolute level of income/whatever of the poorest (as Rawls does) or on that level relative to the level that others are at (which I and W&P do)."

Not only is that a very bad and inaccurate description of Rawls' complex project, but you've just admitted that what you are interested in RELATIVE poverty - whilst bizarrely trying to insist that really this is absolute poverty. Apparently just because you say it is. Which all has the ironic result of proving the exact point TJ was making against you. Something which you appear to be making a habit of in this thread...

Anonymous said...

As for the idea that Rawls' deals with the bases of social respect via the first principle of justice, that's absolutely ridiculous.

The first principle pertains to the distribution of rights to equal compatible individual liberties. The bases of social respect, which receive extended treatment in Part III of TofJ, are classed by Rawls as primary goods: "For simplicity, assume that the chief primary goods at the disposition of society are rights, liberties, and opportunities, and income and wealth. (Later in Part Three the primary good of self-respect has a central place)." p.54, Revised Edition of TofJ

The two principles of justice taken together are intended to order the basic structure of a just society. The first principle is lexically prior to the second, and guarantees that rights to liberty cannot be curtailed so as to secure material (or whatever) improvements even for those having their rights curtailed. The second principle, operating within the constraints set by the first, covers the distribution of primary goods - of which the bases of social respect are clearly stated to be very important examples.

Thus the claim that Rawls treats social respect as an "afterthought", and only in relation to the first principle, is simply wrong.

Which leads me to be rude once more: if you're going to write about Rawls, you ought really to raed his book first.

Anonymous said...

Damn, sloppy comment:

"As for the idea that Rawls' deals with the bases of social respect via the first principle of justice, that's absolutely ridiculous."

Should read

"As for the idea that Rawls deals with the bases of social respect PRIMARILY THROUGH THE FIRST PRINCIPLE of justice, that's absolutely ridiculous."

Chris Brooke said...

Hello Rupert,

Thanks for your reply. I'm afraid I don't have my copy of "Theory" with me at the moment, so I can't cite chapter and verse. But -- like badconscience -- I'm not yet seeing why anyone should be persuaded by your approach to reading Rawls.

First, I don't understand why the fact that Rawls discusses something later rather than earlier in his book means that it's not important. If you're going to argue that the material in part III of "Theory" isn't important, you need an argument about why we can dismiss it, not an assertion that it "feels" like an afterthought or anything like that. I know that for a long time the literature on Rawls neglected great chunks of the book, and concentrated quite narrowly on the more methodological bits at the beginning, but now that we're past that unhappy period in Rawlsiana, I don't see why we should repeat its errors, or deny that they were errors.

More generally, though, I'm still wildly unpersuaded by the general view of Rawls your discussion suggests. You say that "what W&P show particularly clearly is that the social basis of self-respect depends upon degree of suffered inequality, not on 'bourgeois freedoms". OK. Let's say that they show this. (I haven't read a great deal of TSL, but I'm prepared to take your word for it.) What's the reason for thinking that Rawls wasn't sympathetic to this line of thought?

Consider: he spent his entire professional career lecturing on Hobbes and Rousseau, and when he lectured on Hobbes and Rousseau we know that he would discuss Hobbes on "glory" and Rousseau on "amour-propre" at some length, which concern (among other things) the ways in which individuals take delight in dominating other individuals. Both were concerned with the ways in which "glory" and "amour-propre" created problems for the creation and maintenance of a society of equals; and Rousseau in particular was interested in how, to use your exact words, "the social basis of self-respect depends upon the degree of suffered inequality" rather than on, say, bourgeois freedoms.

Rawls did a lot to popularise the term, "moral psychology". And when people say that Rawls was interested in "moral psychology", what they are getting at is precisely that he was terribly interested in these early modern arguments (which is why he spent about forty years or so lecturing on them), and reflecting on the implications that they had for contemporary political theory.

You might object that teaching and research don't have much to do with one another: that there's no reason to think that just because Rawls used to lecture on Hobbes's psychological account, that it played any role in his thinking about justice. But that would be obviously false: just to read the material that prefaces the Lectures on the History of Political Philosophy and to think about where the material in, e.g., Justice as Fairness comes from is to see how Rawls always saw his own project as a continuation of and critical response to the kinds of writers he considered in the courses he taught to undergraduates.

I'm not a Rawlsian -- I work on seventeenth and eighteenth century political thought, mostly. But the reason I'm impressed by Rawls is that he was the 20th century political philosopher who seems to me to have reflected longest and hardest on the legacy of early modern psychological and political theory when building his argument.

But if you're right, of course, I'm wrong, and I've been wasting my time thinking about Rawls, because what I'm so impressed by is just an afterthought, or something that doesn't matter in the slightest.

Rupert said...

badconscience's comments display an unfortunate lack of charity / ability to read: I have made it perfectly clear that my concern is relative poverty - that's the whole point of my piece to which Stuart's piece here is a response!)
Anyway: The key point that he and Chris Brooke dwell on concerns how self-respect is secured. There is lots to say on this, but my starting point is p.545 of the original edition of ToJ, where Rawls says quite plainly that "In a well-ordered society...self-respect is secured by the public affirmation of the status of equal citizenship for all; the distribution of material means is left to take care of itself in accordance with the idea of pure procedural justice." That makes my point nicely: self-respect is primarily 'secured' through the basic liberties. A desperately mistaken argument, in my view (and, implicitly, in W&P's.).

Rupert said...

What I confess to finding fascinating is the keenness of some on the Left, as well as many on the Right, to find fault with W&P; and the keenness of many Fabians etc. to find little fault with Rawls. I confess further to reading that keenness in the light of Stokeley Carmichael's great remark that the liberal is all for social change, provided that it doesn't affect his own privileged social position. And in the light of Raymond Geuss's brilliant critique - far more savage than mine, which is in many ways quite modest - of Rawls's entire enterprise, as a 'reconciliation' project, like Hegel's. 'Reconciling' us to pervasive inequality, as allegedly just. I commend Geuss's work (e.g. his PHILOSOPHY AND REAL POLITICS) to readers of this thread.

Peter said...


"I confess further to reading that keenness in the light of Stokeley Carmichael's great remark that the liberal is all for social change, provided that it doesn't affect his own privileged social position"

- This seems a strange remark. Do you seriously think that America satisfied the Difference Principle when Rawls was alive? Presumably not. In which case, as a relatively wealthy individual, the implementation of the policies he favoured would have affected his own privileged social position (at least monetarily).

As for Geuss, I haven't read him, but presumably this is the same Geuss who made elementary confusions about Rawls. (Sam Rickless in the comments here:

points one out). Hmmm ...

Rupert said...

Peter; I don't buy that. Given that the difference principle is compatible with indefinitely large inequalities, Rawls was performing for himself and fellow liberals the brilliant rhetorical trick of appearing 'egalitarian', appearing to have the interests of the worst-off at heart - while promoting a principle that posed no threat to his own class position.

As for the Rickless comment - I can't see in it any substantive criticism of Geuss. What I can see in it is a defence of inequality...

Peter said...


Thanks for the reply. It is true that the DP is (theoretically) compatible with indefinitely large inequalities. But how large the inequalities it in fact permits is an empirical question. Now, I admit I've never been to the USA, so maybe it is a Rawlsian utopia that satisfies the DP, but I find that unlikely. Indeed, I find it blindingly obvious that the UK does not satisfy the DP, and by all accounts the USA is a much more unequal society than the UK. If someone were to claim, with a straight face, that the UK satisfies the DP, then I'd think that I was talking to someone who was an incompetent witness to how the world is. Same thoughts apply to the USA. So I can't really take seriously the claim that the DP poses no threat to a relatively rich person like Rawls' economic position. (though I've got no data and I'm no social scientist, so I think we might just have to agree to disagree here)

Geuss claims that he can't see why the deliberators in the Original Position would agree. Rickless explains why this is thoroughly confused. That's a substantive criticism of Geuss, surely?

Rob Jubb said...


a central part of Rawls' case for his two principles over utilitarianism is that they, unlike it, are an agreement not to benefit unless all can benefit. That, he claims, is why they are capable of meeting the demands of publicity and stability: a conception which was not an agreement to tie together all's prospects in this could not be stable or public, because by failing to treat all those living under it as ends in themselves (or something similar), it would fail to give all sufficient (access to) self-respect. Clearly, a central way in which the two principles are agreements to benefit only if all can benefit is the difference principle. This is why, for example, Rawls describes the difference principle as a principle of fraternity and compares its norm of distribution to that in a family (pg. 105, ToJ, 1971). Now, that means that a central part of Rawls' argument for the difference principle turns on precisely the sort of claim you say he is systematically ignoring: a point about the social bases of self-respect. Rawls' argument here for the difference principle is that any other principle distributing the primary social goods apart from rights and liberties would give the poorly-off inadequate holdings of one of those primary social goods, the social bases of self-respect. Whether you think this argument works or not is a different question, but it clearly shows that Rawls is centrally concerned, as Chris Brooke says, with the social bases of self-respect. This is of course on top of the idiocy of thinking that an empirical claim (higher levels of inequality in income and wealth produce bad effects other than those particular inequalities, particularly but not only for those at the worse end of them) could refute a normative principle (inequalities are permitted to the extent that they improve the holdings of the worst off across a metric of goods, not limited to income and wealth). Presumably if you're invoking Guess though, the red mist has descended and you're impervious to correction.

Chris Brooke said...

Only intermittent access to the web right now, and, as before, no copy of TJ with me, but just dropping in to say that the quote you offered from p.545 doesn't make the point you imply it makes, and doesn't cut against anything I've said on this thread. It's not a point about the different roles of the two principles of justice (both of which are relevant to an ideal of equal citizenship) but a point about how to think about pure procedural justice and the distributive outcomes that such a scheme might generate.

Rupert said...

Thanks Chris. I appreciate your engagement, and the fact that unlike others on this string you haven't descended into name-calling against me (It is as I say quite fascinating to see how angry some liberals get when one dares to challenge their shibboleths). But I don't agree with your reading of the p.545 passage. I think that the wording of it speaks for itself, really, in relation to the point I am making. Maybe we will just have to agree to differ on that.

Rob: I am well aware that Rawls wants to say and be committed to all the things you say here (so kindly refrain from your unpleasant hectoring tone). But just because someone wants to support liberty, equality and fraternity does not imply that they are succeeding in doing so. My argument in fact is that there is an inevitable clash between liberty on the one hand and fraternity and equality on the other. Fraternity and equality build community (which as a left- and eco-communitarian is important to me); liberty undermines it. Rawls is a liberal, who (literally) prioritises liberty; hardly surprising then that he and his followers end up being apologists for inequality and thus for a regime in which there are, as W&P show, crushing burdens undermining self-respect.

Rupert said...

Thanks Peter. The question is whether there is really deliberation and discussion etc. in the original position or not. Geuss is only confused if one thinks it unproblematic to interpret the OP as actually only having one person in it (much like Utilitarianism, ironically). I quite understand that that is Rawls's 'official' position; but can it be squared with the many passages in Rawls that make it sound as if there are actually discussions going on between different people, behind a veil of ignorance about themselves, in the OP?
So I think Guess innocent until proven guilty. Rawls's text seems to me wide open to multivalent interpretations on this point.

Rupert said...

Responding to Rob: As I have already made quite clear, I am well aware that it is 'queer' to speak of a fact (an empirical claims) refuting a normative claim/principle. if Rawlsians want to stay in favour of the difference principle even after it has been shown to undermine itself / to result in grave 'unexpected' harms unless it is extensionally equivalent to equality, then they have a right to do so. The question is whether it makes much sense to hold onto a normative principle that has such results. Or, to put it another way: Is it reasonable to think that one would be 'in reflective equilibrium' who held onto the difference principle as a justification for inequality / for relative poverty, even after it had been shown that societies which did so were less happy, less healthy, etc etc etc.?

Anonymous said...

First of all, I'd like to just make public that I am Paul Sagar (this can easily be discovered by clicking-through on my link, but I just want this out in the open as have no desire to appear to be hiding in anonymity).
Secondly, regarding your earlier reply to me:
“badconscience's comments display an unfortunate lack of charity / ability to read: I have made it perfectly clear that my concern is relative poverty - that's the whole point of my piece to which Stuart's piece here is a response!)”
I must say I find this quite astounding, as my complaint against you was always and only that in this thread you were earlier attempting to claim that absolute poverty is relative poverty. Now I find you proudly declaring that you were only ever interested in relative poverty. Which is fine – except that before you tried to tell us all this was the same thing as absolute poverty. Which it isn't.
Thirdly, I want to take particular issue with your (now repeated) citation of a short sentence from p.545 of the original edition of A Theory of Justice. Chris Brooke is, of course, completely right that as presented that sentence does not offer a reply to his points above. Nor, I might add, does it reply to mine – which I'd also like to say have been reiterated with much more power and accuracy by Rob Jubb, whom again you have singularly failed to reply to.
But I'm afraid things get worse. I decided to head to the library and look up what Rawls actually says on p.545 of the original edition (having only the revised edition at home). So let's do this properly.
You quote Rawls (correctly) as saying: “In a well-ordered society...self-respect is secured by the public affirmation of the status of equal citizenship for all; the distribution of material means is left to take care of itself in accordance with the idea of pure procedural justice.”
Which is all very well and good. Except that there’s an ellipsis in your quotation, which serves to omit the word “then”. Which indicates that the sentence you offer up is located midway through a larger argument. And indeed, this is exactly the case.
But what might that larger argument be about? Well, the relevant section is entitled [in the Revised Edition at home; I forgot to check in the library before I got kicked-out at closing time] “The Grounds for the Priority of Liberty”. Interesting.
Furthermore, the opening sentences of the paragraph preceding the one you quote from read: “Of course, it does not follow that in a just society everyone is unconcered with matters of status. The account of self-respect as perhaps the main primary good has stressed the great significance of how we think others value us.”

[Cont. next comment]

Unknown said...

Which, if taken alone, might be read as offering something like support for your thesis that Rawls only seeks to secure the bases of self-respect via a prioritising of liberty. Unfortunately, if we read on we see that this is clearly not the case. For Rawls in this paragraph goes on to affirm precisely that equal rights and liberties are the basis of a society securing self respect for its citizenry. What Rawls is doing here is arguing that a society which cares about the self-respect of its citizens is going to need to start by affirming the first principle of justice, which also has priority, when ordering the basic structure. (The clue was in the subtitle of the section, after all).
But of course, Rawls doesn’t end things there. Because he give us precisely the sentence you quote . Which let’s recall, states that “In a well-ordered society then self-respect is secured by the public affirmation of the status of equal citizenship for all; the distribution of material means is left to take care of itself in accordance with the idea of pure procedural justice.” And what does procedural justice pertain to? The two principles of justice – including the difference principle – which as Rob has demonstrated above are intimately concerned with the bases of self respect understood over and above mere income and wealth, and as indeed established by the previous 544 pages of the work under discussion.
What you’ve done, Rupert, is pick one single quote from a section of TofJ which is specifically concerned with explaining why Rawls thinks that the two principles have indeed to be two principles, and why the first one indeed comes first. To present this as some sort of proof that Rawls is only concerned with the bases of self respect as some sort of “afterthought”, or that his discussion of the bases of self respect is predominantly conditioned through the first principle alone, is disingenuous to say the least. I believe in science they call such activity “cherry picking”. And I’m fairly sure academic political theory journals take as dim a view of it as scientific ones.

Rob Jubb said...


I'm sorry about the tone, which was a bit hectoring and uncalled-for. However, what you say fails to respond to what I said, which shows, fairly unequivocally, that the social bases of self-respect are a major part of Rawls' work, and in fact a crucial part of the argument for the principle you want to reject on the basis of some empirical work about them (and various related ideas). Whether or not liberty, equality and fraternity are in fact compatible is an obviously different question from whether Rawls thought they were, which was not anyway the claim I was responding to. If you have an actual argument that they are not compatible in the way that Rawls suggests they are, that is, you can point to one of the basic liberties - which do not, for example, include the right to private property in capital - that you have a plausible case ought to be restricted in order to avoid the kinds of harms that the Spirit Level blames on material inequality, then there would be a point of actual disagreement. As for questions about whether our judgments about the harms the Spirit Level blames material inequality for can be sustained in reflective equilibrium with affirmation of the difference principle, as I think several other people have said, you would have to convince me that the two principles mandated societies similar to those which the Spirit Level studied. I am intensely sceptical about that, not least because none of them guarantee either the fair value of the political liberties or fair equality of opportunity, whatever their (manifest) failings with regard to the difference principle.

Unknown said...
This comment has been removed by the author.
Unknown said...

I agree with much that has been said above, but rather than rehearse what has already been noted, I'll try to say something different.

1. The difference principle is the second part of the second principle. (From now on, let's do away with this Rawlsian terminology and call the two principles what they are - three principles). It is the third principle, lexically inferior (as has been noted) to the prior two.

The thesis which this post refers to claims that empirical findings of the effects of inequalities in actual real societies will come to pass in the (ideal) Rawlsian society, and so we should reject the difference principle.

The point about the difference principle being the third of three is important for two reasons. Firstly, the inequalities allowed by the difference principle will (it seems plausible to suggest) be massively constrained by securing both fair equality of opportunity and the fair value of the political liberties.

Secondly, how do we know the ills presented in the empirical works referred to are the result of difference principle -type inequalities, and NOT inequalities in the social bases of self-respect, opportunity (I would wager a lot is down to this*), or political liberties?

(*It might be objected here: ‘yes, some of the problems are down to opportunity inequality, but you can’t get equal opportunity without a (broadly) equal schedule of rewards. So it’s still about equality of resources (or whatever), which goes against the dp’. Fair enough, but the dp wouldn’t license any inequality that allowed inequality of opportunity to result, because it’s lexically inferior to the FEO principle).

For both these reasons, in order to properly assess this, we'd need to find a society which had fully realised all three principles, and a society that had realised the first two but gone for a more egalitarian principle instead of the difference principle and compare how well they were doing with respect to the kinds of problems flagged in this empirical literature. (Additionally, we'd have to be sure that they were problems of *justice*, a point made-with specific regard to metrics, but more generally here-by Chris Bertram above). The basic point is it’s hard to see how an ideal theory gets refuted by empirically observing problems in societies that are, by the lights of that ideal theory, deeply non-ideal (and not only with reference to the specific principle being ‘refuted’)

Continued Below...

Unknown said...

2. With regard to 'Rawls puts too much emphasis on basic liberties when dealing with self-respect'. I'm inclined to stick up (to an extent) for Rupert Reed here. I think a large part of Rawls’s argument that the (three) principles of justice underwrite the social bases of self-respect focuses on the first principle, and this was pointed out by Russell Keat and David Miller as early as 1973. Rawls says (TJ Rev, p 477): ‘The basis for self-respect in a just society is not then one’s income share but the publicly affirmed distribution of fundamental rights and liberties’. But the argument is not, as Reed seems to suggest, ‘in current societies, liberties are what matter for people’s sense of self-respect’. That claim could (presumably) be refuted by the empirical evidence on show. But rather, it seems to me, the argument is: ‘in a fully just society, we would do better to have a public/institutional culture in which our judgments of self-worth are focussed on equal citizenship and not on things like wealth and income’. (P478: ‘Having chosen a conception of justice that *seeks* to *eliminate* the significance of relative economic and social advantages as supports for men’s self-confidence’. This quotation accepts that people *do* base self-respect judgments on wealth and income, and claims that the problems that stem from this – the problems pointed out in TLS etc – would be eliminated by a different public way of making such judgments).

So showing that people *in current societies* care deeply (whether consciously or not) about income disparities doesn’t really seem to refute that point. The point is, in a just society (hopefully) they won’t – that’s one of the things for a just society to eliminate.

Now, you might think that it is a silly empirical assumption to think that people could *ever*, when judging self-worth, forget about disparities in wealth and income. Whether or not that matters depends on how ideal ideal theory should be. But Rawls is happy to accept that it might be too much to expect this:

‘Now it is quite possible that this idea [that self-respect be tied to equal liberty] cannot be carried through completely. To some extent men’s sense of their own worth may hinge upon their institutional position and their income share. If, however, the account of social envy and jealousy is sound, then with the appropriate background arrangements, these inclinations should not be excessive. But theoretically we can if necessary include self-respect in the primary goods, the index of which defines expectations. Then in applications of the difference principle, this index can allow for the effects of excusable envy; the expectations of the less advantaged are lower the more severe these effects. Whether some adjustment for self-respect has to be made is best decided from the standpoint of the legislative stage where the parties have more information and the principle of political determination applies’. (p478-9)

So, at most, this argument seems to say ‘Rawls made a silly empirical assumption, which he recognised might be a bit silly, and so this favours the second interpretation of the difference principle, rather than a purely resourcist one’.

I’m not personally a fan of the difference principle. I’m not a fan because it justifies *theoretically* unlimited inequality, and I’m not happy to say that *theoretically* unlimited inequality is through and through just. But that’s quite different from saying that, in operation, it will produce massive disparities in wealth, especially taking into account the prior two principles. Indeed, Rawls regularly (including in the account of ‘social envy and jealousy’ referred to above) assumes it won’t produce such large disparities (and provided alternative ideas if that proved not to be the case). And since the complaint is ‘IN REALITY inequalities are bad, therefore the difference principle is bad’ what we should care about (for this particular argument) is whether dp would produce these inequalities IN REALITY.

Rob Jubb said...

Although I think, unlike Paul and perhaps Chris, that Patrick is right and that Rawls thinks the social bases of self-respect are in a just society secured by the provision of the most extensive set of basic liberties compatible with like liberties for all, it's worth bearing in mind what he thinks goes wrong with the social bases of self-respect in unjust societies. For example, we can look at what he says about the superiority of the two(/three) principles to other principles around which a political order might be organised. Here, the fact that he thinks that utilitarianism would struggle as a public doctrine and could not be accepted by some members of society because of the damage it would do to their sense of self is obviously important, particularly since that turns on the reciprocity of the two(/three) principles, which is particularly embodied in the difference principle. We don't just need to know what Rawls thought was good about the societies he thought the two(/three) principles mandated, but what he thought was bad about other societies. That tells us that for him, a political order which did not require its citizens to treat each other with reciprocity in the distribution of income and wealth (amongst other things) would tend to undermine their self-respect. So whilst equal liberties are what self-respect is based on in Rawls' just society, he holds that material inequalities threaten that basis of self-affirmation. None of this, of course, is to defend Rawls on the substantative details of his arguments. I, although not really for the same sorts of reasons as Patrick, am sceptical about the success of Rawls' argument for the difference principle. Instead, it's to point out that Rawls has rather more resources to respond to the concern you raise that you initially seemed prepared to acknowledge.

Stuart White said...

As the comment thread here seems now to have run its course, I thought I would thank everyone for their comments.

I've certainly learned a few things about Rawls's theory in the course of the discussion.

Rupert at one point suggested that some people who defend Rawls might be motivated to do so by class interest - that his theory helps to justify inequalities that relatively rich liberal academics benefit from.

However, while Rawls's views are sometimes criticised for being insufficiently egalitarian (where Rupert is coming from), my view is that they are more usually set aside in contemporary debate outside the academy because they are inconveniently pro-egalitarian. They challenge widespread notions of 'desert' and 'entitlement' which many people now accept (some of them are even in the Labour party!).

So part of my motivation for defending Rawls from unfounded criticism is to try to prevent the marginalization of a thinker who, in my judgment, is one of the great demystifiers of what Tawney called 'the religion of inequality.'