An Empirical Refutation of Pareto-optimality?

Click to increase image sizeClick to decrease image size

Notes

¹This is not to imply that I am impressed with many of the criticisms of The Spirit Level. On the contrary, I find nearly all of them almost wilfully bad, more attempts at flak than at real debate. Wilkinson responds to the criticisms here: http://www.newstatesman.com/society/2010/11/inequality-social-health-essay. I find his and Pickett's published responses to the criticisms that have been made of them more than adequate to the task.

²This is demonstrated in the section of The Spirit Level helpfully entitled “Everyone Benefits,” found on p.180f. The graph on p.185 is particularly striking and helpful. See also the graph on p.109.

³See the early chapters of The Spirit Level for detail.

⁴This point is, naturally, likely to be of great interest to readers of CNS. Wilkinson and Pickett (2014) argue that the harmful effects of status-competition intensified by inequality are a major driver of environmental damage.

⁵This point of mine was partly anticipated by Patrick Shaw on p.360 of his “The Pareto Argument and Inequality” (1999).

⁶And thus my argument challenges what might be termed “the standard view” of Pareto-optimality's incompatibility with egalitarianism, as found for instance in the following pieces, pieces that have various differences from one another; but none of them considers the irony that it might be that the only way of saving Pareto-optimality would be via egalitarianism: Tungodden (2003), Brown (2003), Fleurbaey and Trannoy (2003) and Temkin (2003). Of course, none of these authors had the benefit of reading The Spirit Level.

⁷Not exclusively, of course—one has to assume norms such as that being healthier is a good thing. But the norms that have to be assumed are pretty unexceptionable!

⁸Again, Shaw (1999) partly anticipates my argument here, in arguing that the apparent self-evidence of the Pareto principle is put into question by serious consideration of the nature of inegalitarian societies. See p. 361 thereof.

⁹To understand the empirical evidence on this point, see especially note 2 above.

¹⁰On this point, see Scanlon (2003). Rawls (see below for discussion of his theory of justice) also seeks to make an argument on the basis of status against large-scale inequality. See Read (2011a) for a response to this.

¹¹Rawls prefers the term “efficiency” to “optimality,” for reasonable reasons that, however, need not concern us here. See p.66 of his (1971).

¹²For details as to what this claim amounts to, see Haksar's classic (1972) Analysis article.

¹³There is, of course, also a debate about whether Rawls really succeeds in giving a principle of justice at all, considering his dependence on Pareto. See especially p. 30 n.7 and 317–318 of Cohen (2008). In the present piece, I am seeking to be charitable to Rawls on this point, and arguing that he still takes a mighty hit from the argument that I have been considering and purveying, even when one is charitable in this way.

¹⁴An argument along these lines is made nicely by Wolff (2001). Some changes can be Pareto-efficient on some currencies of better/worse off but those very same changes can also be Pareto-inefficient on other currencies of better/worse off.

¹⁵Again, Wolff's (2001) piece is helpful here; see especially p .5 thereof.

¹⁶In future work I intend to explore these more fully. For, ultimately, mine is a critique of the kind of political or economic thinking that is exclusively concerned with making ‘optimal’ choices, where optimality is expressed in something like individualisable pecuniary terms.

In this context the Pareto principle is just an optimality requirement, in the sense that it is necessary in order to have a stable optimal choice. (The Pareto principle is, moreover, not only specifiable with respect to allocation problems but also to choice problems in general; thus, the potential impact of the line of thought I am pursuing in the present paper is wide indeed). An alternative is Pareto-optimal if no other alternative is unanimously preferred to it. Pareto optimality is reasonable under the presupposition that a formally optimal choice is what is being sought, but this is the framework that I am questioning. So the obvious apparently-completely-compelling force of the Pareto principle only emerges under presuppositions that I question. Unless, that is, we methodologically reinterpret the Pareto principle along the lines that I outline above: in which case the traditional mode of its application—to choice problems presented purely to individuals—no longer works, no longer applies. We have, rather, to choose as a society.

¹⁷This thought is close to Jerry Cohen's critique of Rawls. See e.g. his (1995), as well as n.13 above. On how to take that critique within a broadly eco-socialist context, see Read (2011c).

¹⁸This conclusion as to what is implied by Rawls's difference-principle argument is argued for in Read (2011a). Cf. also Read (2011b).