I should find this book more appealing. I mean, I am receptive to its basic thesis: that high inequality is bad for a society. I find it plausible to think that societies with a high level of status inequality and status competition do worse in a variety of ways than more equal societies, and in particular that large inequalities may induce bad health effects via elevated stress levels (as they apparently do in all hierarchical primate societies). (Indeed, I do not object to the kinds of comparisons between people and other primates that Wilkinson and Pickett use: they seem fruitful enough to me). But there is something that bothers me about it. Since I am supposed to say something about it from a political theory perspective in a “policy forum” here at VUW in a couple of weeks, consider the (extremely long) post below my attempt to make sense of why I find it unsatisfactory, or perhaps more accurately, my attempt to clear my mind about what it is that is problematic (or not) about inequality.
Before I get to the point, however, I should say that the book does not strike me as an example of good social science. To be sure, most claims in it are supported by an impressive array of footnotes to other research and scientific-looking scatterplots, but even for a lowly political theorist like me some problems were obvious: there is sample selection bias (why should places like South Korea, Israel, Hong Kong, Slovenia, Slovakia, the Czech Republic, and Taiwan be excluded from the analysis, even though they have per capita incomes comparable to if not higher than that of Portugal, which is included?); the graphs all represent static pictures, which fail to control for any factors which may interact with inequality, and often seem to be driven by one or two outliers; there is little discussion of time trends, and what little there is seems unduly focused on the UK and the USA and does not always support their argument; and there is a lot of dubious sociology (one thing that jumped at me is the claim that the Japanese level of income inequality means that Japan is a highly equal society; passing acquaintance with Japanese society indicates that though Japan may be a highly equal society in terms of income, it is also a highly hierarchical society, which would seem to conflict with the mechanism Wilkinson and Pickett identify as responsible for the bad effects of inequality).
I understand that some of these issues have to do with the availability of comparable data on income inequality, and that many of them would have been addressed in some of the studies they cite; but I find that these issues raised enough doubts in my mind to find critiques like Christopher Snowdon’s plausible (the title of Snowdon’s book tells you pretty much what he thinks of The Spirit Level: it’s called The Spirit Level Delusion). On issues that I know a little bit more about, though still not much – like the political science literature on the determinants of inequality, or the literature on happiness and income – I found their summary of other people’s research superficial and sometimes misleading. And obvious alternative mechanisms for the effects they see are simply not explored: for example, it may be that the key variable here is exposure to labor market competition; it is the fear of losing a job that generates stress and bad health effects, but this variable may be systematically associated with inequality. (I am not claiming that this is the case, but it seems like something worthy of investigation).
But I am not a specialist on the literature on health and inequality, and at any rate nothing hinges on any one study or picture. It is the cumulative evidence that matters, and I would not be surprised if something like the Wilkinson-Pickett thesis on inequality were reasonably close to the truth, though probably in a more qualified way. Inequality probably interacts with other factors, including “cultural” factors, in complicated ways; but I imagine that at least some of what they are saying reflects the broad trends of research on inequality (though I could be wrong). I do not, at any rate, have the time or inclination to fact-check the book and wade through the many studies they cite, so I shall take what they say as generally on target, even if subject to discounting.
Moreover, it would be quite surprising if high levels of income inequality did not have any bad effects on social life. Income inequality may be a highly imperfect proxy for the main hierarchies of status and power that structure a complex society like our own, but it probably is correlated with such hierarchies; and highly status-stratified societies are unlikely to be “well integrated.” (The idea that there are “happy hierarchies” – of caste or occupation or the like, where “each is set in his own rank,” to use the words of an ancient defender of hierarchy – is a figment of the ideological imagination). To be sure, the kinds of status anxieties and status competition that Wilkinson and Pickett believe mediate the effects of income inequality on bad social outcomes are typically centered on one’s neighbours, not on distant rich others, but this does not prevent such competition from having effects across the entire hierarchy. I am reminded of some of the things Don Herzog describes in his book Poisoning the Minds of the Lower Orders (discussed in a previous post): in the England of circa 1817, a highly unequal society in almost every respect, each social group sought to preserve its own status by despising those below them in the status hierarchy, a chain of contempt that was created by local effects yet had global implications. This was not merely a “psychological” reaction, but part of a system through which the “lower orders” were put in their place: inequality here was translated into domination and subordination.
But even if the specific mechanism that Wilkinson and Pickett favour to account for the effects of inequality (the induced stress resulting from higher levels of status competition in unequal societies) turns out to be wrong or highly exaggerated, the claim that wealth or income inequality (and not just poverty) is bad for societies has a long and distinguished pedigree. It is interesting to note that many defenders of hierarchy in the history of political thought – writers who did not think human beings are of equal moral worth in the sense that we think they are, and who were willing to defend slavery and other practices we would find abhorrent – thought that, among people who are presumptively equals (like free male citizens), large income inequalities were corrosive of political and social life and generally to be avoided. So there may be something to it.
Instead of debating the factual basis of Wilkinson and Pickett’s views, then, I want to ask about the reasons we should care about (income) inequality, and about how much we should care. Wilkinson and Pickett’s view for why we should care about income inequality is essentially consequentialist: we should care about inequality because having less of it would make us healthier and happier. But, even if we accept this consequentialist framing of the question (about which more in a minute), we would still have to ask how much less income inequality would we want, and at what cost to other values, including other forms of equality.
Consider a simple example. The communist societies of Eastern Europe typically had very low income inequalities, as Wilkinson and Pickett are well aware. And some of them even achieved relatively high levels of “development,” measured conventionally (e.g., East Germany, Czechoslovakia, and the Yugoslavia of Tito; Cuba is a non-European example today, with its high levels of “human development” – though the typical figures cited in this connection are more than a little dubious these days). Yet most of us would agree that the mechanism through which income equality was achieved in these cases violated rights that we would consider to be fundamental. Indeed, Wilkinson and Pickett say so. But more importantly for my purposes here, these societies were not really equal in an important sense of the term: though income differentials didn’t matter much for social life, inequalities of political access mattered enormously. Today North Korea serves as an extreme example of the sort of “caste” society that can emerge even in the almost complete absence of income differentials; where money has very little value, family background, regime loyalty, etc. can all become terribly important. (And even here, there are nuances and interesting complexities: since the famine in the 1990s, political controls have eroded, money and markets have become more important, and income inequality has increased. But this has been, on the whole, a good thing for a lot of people!).
Now, someone may object that this is obvious and irrelevant: nobody is arguing that we should implement a system of state socialism in order to diminish income inequality. Wilkinson and Pickett’s proposals tend more towards ideas for worker self-management within a democratic framework, though the political mechanisms through which they expect such a system to sustain itself remain woolly. (It is worth noting that however attractive such proposals may appear – and I am not immune to their charms – the most extensive and successful system of worker self-management, in Tito’s Yugoslavia, did not survive the partial democratization and collapse of the country at the political level, and indeed seems to have precipitated the enormous Yugoslav economic crisis of the early 1990s). At any rate, the levels of equality in Sweden or Japan – the countries that Wilkinson and Pickett single out as the most equal - can be achieved within the framework of a typical capitalist economy (something that Wilkinson and Pickett seem to forget in their last chapters, where they basically attack capitalism as such). No complete transformation of property relationships is needed to reduce income inequality substantially through taxes and transfers (within the limits given by the efficiency losses induced by high taxes, available technologies of tax evasion, the political resistance of the rich, and cultural ideas about what constitutes “excessive” taxation).
But the problem is that simple income equality is unlikely to be what we are really looking for, as Wilkinson and Pickett seem dimly aware of in their last chapters with their proposals on worker self-management. When money fails to serve as an indicator of distinction, other things may easily take its place. Consider the fact that Japanese CEOs get paid less in money than American CEOs (who many people consider to be obscenely overpaid relative to workers), but more in deference and perks. This may be good or bad, but it is not altogether obvious that one sort of inequality is preferable to the other, even if Japan is in fact a healthier society than the USA for any number of reasons. (Indeed, I rather prefer the money inequality to the status inequality in this case: at least the money inequality does not involve direct subordination. Marx himself considered the “cash nexus” of exploitative capitalism an advance over the sorts of paternalistic subordination typical of “feudal” relations). So if we are interested in equality, we need to think clearly about what, exactly, is bad about income inequality, and whether the means we could use to eliminate it simply introduce other perhaps even more invidious inequalities in their wake.
This is a point made more elegantly and generally by Michael Walzer in his book Spheres of Justice. In Walzer’s view, the different goods produced by social cooperation have different meanings that indicate the appropriate way to distribute them; principles of distribution make sense only in specific spheres. Grades should not be distributed according to wealth, or emergency healthcare according to ability to pay, and more generally, not everything should (or could!) be distributed according to money. But most goods can also “cross spheres,” money and political power being the most obvious cases. It is the convertibility of goods that raises problems; if money could never be converted into status, political power, personal domination, and so on, income inequality would be far less problematic. So an egalitarian society, in Walzer’s view, limits the convertibility of the various goods produced by social cooperation. In particular, an egalitarian society should limit political power from being easily transformed into wealth, and wealth from being easily transformed into political power and social status. But an egalitarian society would not, necessarily, limit inequalities in wealth except to the extent that such limitation arises endogenously from the attempt to prevent wealth from turning into political power, status, and domination over others (Walzer thinks, for example, that some forms of large-scale property are inherently political, and hence ought to be limited or regulated by “voice”). There are complications here, for political power and money are both special sorts of goods (they “seep” through all the other spheres of distribution, and sometimes what looks like one is really the other), and there appears to be no stable equilibrium that completely guarantees the nonconvertibility of coercive political power or dominating social status into personal income and vice-versa (indeed, there are some problems with the very coherence of Walzer’s proposals, but discussing them would take us too far afield).
But in general Walzer’s intuition is that when we say that we are interested in equality, what we (should) mean is that we are interested in nondomination, and this intuition seems reasonably sound. We are not actually interested in income equality or inequality (at least to the extent that we are not simply motivated by envy), but in not being dominated by those with money and in retaining our self-respect as equal citizens. A similar point was made in a very influential article by Elizabeth Anderson entitled “What is the Point of Equality?” Anderson was criticizing what she considered to be the pernicious position of “luck egalitarians” (a term she coined, but which has been embraced by proponents of this form of egalitarianism). Luck egalitarians argue that the point of justice is to neutralize those differences in welfare and resources for which an individual cannot be held responsible. In its most persuasive form, luck egalitarians start from the common intuition that luck (chance) neutralizes desert and responsibility: someone who inherited lots of wealth could not be said to “deserve” it, and someone who does badly because of factors that are outside his or her control cannot be said to be responsible for his or her misfortune. But, by the same token, we do not deserve our good fortune in the form of good parents, or natural talents, or place of birth, all of which are factors that greatly influence whether we make lots of money or not, and whether we live satisfying lives or not. Exactly what we “deserve” in this sense is difficult to say; some luck egalitarians like G. A. Cohen seem comfortable saying that we basically deserve nothing save perhaps a bit of our effort, though others are more generous. This already seems to miss the prospective elements of desert – maybe you eventually come to deserve those opportunities, as David Schmidtz argues – but the more controversial move that luck egalitarians make is to argue that justice requires that those who are fortunate in ways for which they cannot be held (ultimately) responsible compensate those who are less fortunate. Luck egalitarians thus argued for things like a universal “basic income” and for various sorts of subsidies to equalize the prospects of the fortunate and the unfortunate (an example, taken from Anderson’s article: subsidized or free plastic surgery for the ugly), though they mostly accepted the legitimacy of market societies (with some exceptions, like G. A. Cohen, who was a sort of Marxist), and suggested that once inequalities of fortune have been equalized, inequalities resulting from “bad choices” are to be tolerated (even if they turn out to be quite large).
Anderson argued that this position was open to all sorts of objections (which I won’t rehearse here), but her main point was that we should care about equality as a matter of justice (in contrast with charity or humanitarian concern) not because we are interested in fixing the great “cosmic injustice” of individual differences, but because we are interested in not being oppressed and dominated and treated contemptuously by others. Luck egalitarianism, in her view, fails to care about those things; it fixates on natural differences rather than on the ways in which societies transform human diversity into rungs of hierarchy, it privileges envy rather and contemptuous pity rather than self-respect as the key moral sentiments, and it promotes intrusive forms of paternalism in order to avoid excessively bad outcomes.
Extremely long story short, I agree with this; I take it that we should care about income inequality to the extent that income inequality undermines the possibility of a society of equals in the sense of people who cannot dominate one another, and hence can respect each other’s liberty. Now, there’s a great deal of dispute about exactly what is needed for such a life in a modern complex society (do we need a more equal distribution of the means of production? do we need a universalistic safety net or some kinds of public health? do we need a minimum wage?); accepting this view does not say much about which institutions best enable a society of equals. One can certainly imagine that, as income differentials increase, these are translated more easily into power and dominating social status, and there seems to be some evidence for this view. But the same can be true of interventions designed to decrease income inequality – they may also enhance the power or status of some over others in unjustified ways, especially if they take the form of opaque and bureaucratic administrative programs rather than simple transfers. At any rate, in the grand scheme of things, income inequalities seem to be less “coupled” with power and status inequalities in modern capitalist democracies than in many other kinds of society, even if they could do better.
Moreover, there are other considerations that should limit the reach of egalitarian measures. Considerations of ownership matter to our moral thinking, though they are not always decisive. The income that we propose to redistribute is always somebody’s income; can the redistribution be justified in ways that are reasonable? Just saying that the redistribution will probably make people healthier, or that the person “doesn’t really need” the income, is not obviously sufficient, though it may serve a purpose in the political debate over equality. Considerations of past injustice also matter. If Maori have lower income than others in New Zealand, or blacks in the US, the relevant consideration is not necessarily the income inequality, but the past injustice by which such lower income was produced, and which may need to be rectified.
So, does it matter very much if Wilkinson and Pickett are right and more (income-) equal societies are healthier? It would certainly be nicer to have healthier and happier societies, but I am wary of saying that we should engage in large-scale redistribution just because we might gain a little health and happiness (the effects Wilkinson and Pickett note seem to be quite small in absolute terms). The consequentialist frame is too thin to hang the argument for equality, and it is always at the mercy of other research showing that no, inequality is not really associated with worse health and happiness. To the extent, on the other hand, that there are specific inequalities that prevent some people from living the life of a self-respecting, equal citizens – or that enable some to transform political into economic power or vice-versa - then I am more receptive.
And if you, kind reader, have made it this far, you probably deserve a prize. But there isn’t one. Tough luck.