On the site of distributive justice: Reflections on Cohen and Murphy

Philosophy and Public Affairs

Volumn 29, Issue 2, 2000, Pages 137-169

  Pogge, Thomas W ^a  

^a NONE

Author keywords

[No Author keywords available]

Indexed keywords

EID: 33747064435     PISSN: 00483915     EISSN: 10884963     Source Type: Journal    
DOI: 10.1111/j.1088-4963.2000.00137.x     Document Type: Article

References (63)

1. "Where the Action Is: On the Site of Distributive Justice," Philosophy & Public Affairs 26, no. 1 (Winter 1997): 3-30, henceforth SDJ.
2. Cf. also G. A. Cohen, "Incentives, Inequality, and Community," in Grethe Peterson, ed., The Tanner Lectures on Human Values, Vol. 13 (Salt Lake City: University of Utah Press, 1992), pp. 263-329, henceforth IIC;
3. and G. A. Cohen, "The Pareto Argument for Inequality," Social Philosophy and Policy 12 (Winter 1995): 160-85, henceforth PAI.
4. Two powerful earlier responses to Cohen are David Estlund, "Liberalism, Equality, and Fraternity in Cohen's Critique of Rawls," Journal of Political Philosophy 6 (1998): 99-112, henceforth Estlund;
5. and Andrew Williams, "Incentives, Inequality, and Publicity," Philosophy & Public Affairs 27, no. 3 (Summer 1998): 225-47, henceforth Williams.
6. An important defense and extension of Cohen's view is offered in Liam Murphy, "Institutions and the Demands of Justice," Philosophy & Public Affairs 27, no. 4 (Fall 1999): 251-91, henceforth Murphy.
7. Following Parfit, Cohen and Murphy like to express moral views in terms of goals or aims rather than, say, constraints or duties. I will state my points in their terms, but do not find the choice of language innocent. It creates a drift toward consequentialist reasoning, to which both authors are already sympathetic.
8. "If we care about social justice, we have to look at four things: the coercive structure, other structures, the social ethos, and the choices of individuals" (SDJ, p. 26).
9. John Rawls, Political Liberalism (New York: Columbia University Press, 1993), p. 6, henceforth PL.
10. Cohen uses rather dated figures from Lawrence Mishel and David M. Frankel, The State of Working America, 1990-1991 (Armonk, NY: M. E. Sharpe, 1991).
11. Inequality in the U.S. has greatly increased during the 1990s, as is amply documented in the latest report, Lawrence Mishel et al., The State of Working America, 1998-1999 (Ithaca: Cornell University Press, 1999).
12. The Economist (August 7, 1999, p. 18) states that "the average American chief executive now takes home 419 times the wage of the average factory worker."
13. Such options "accounted for a record 53.3% of the compensation given by America's top 100 companies in 1998 to their chief executives. This compares with 26% in 1994, and a mere 2% in the mid-1980s." Amazingly, "the total of shares and share options still 'live' in incentive schemes at the end of 1998 amounted to 13.2% of corporate equity or around $1.1 trillion" (The Economist, August 7, 1999, p. 18).
14. John Rawls, A Theory of Justice (Cambridge, MA: Harvard University Press, 1999), pp. 99, 293-94, henceforth TJ. Rawls does not say enough about what duties beneficiaries of an unjust economic order have toward its victims when they cannot promote the institutional reforms justice requires.
15. Following Rawls and Cohen, I take for granted that persons must not be legally coerced into particular careers.
16. My example focuses on choices among jobs rather than, as Cohen typically does, on choices about how productively to work within one's job, because the former description is more general. Choices about one's productivity can be regarded as choices among jobs more finely distinguished. But choices among quite different kinds of work are not captured by Cohen's favored description; and he therefore often overlooks the very important effects that net income differentials have through their impact on persons' education and career decisions.
17. My talk of "talented" persons follows Cohen and thus means no more than that "they are so positioned that, happily, for them, they do command a high salary and they can vary their productivity according to exactly how high it is" (SDJ, pp. 6-7, emphasis in original).
18. We may factor a tax regime into two components: One component raises revenues for government operations through taxes strictly proportional to income. The other component adjusts relative shares through positive and negative delta taxes, which typically reduce the higher incomes and enhance the lower ones. For simplicity, my focus here is entirely on delta taxes. My use of the Greek letter delta alludes to the fact that delta taxes specifically target income differentials. Rawls's difference principle, which specifically assesses differentials in socioeconomic position, contains the same allusion, but is not, of course, the only possible criterion for ranking alternative feasible delta tax regimes. The subscript i is a reminder that each delta tax regime would result in a different triplet of values for m, l, and n.
19. I have been sliding into a simplified version of the difference principle in order to . keep the discussion perspicuous. The simplifications do not materially affect the arguments to follow.
20. i. For all talented persons, the higher the prevailing n, the more attractive they find managerial relative to laboring work (cf. note 18).
21. I use "reward multiple" as a shorter and more suggestive synonym for "net pay rate ratio."
22. The managers' net contribution is the value of the labor they supply (reflected in gross pay) minus the value of the income they receive (net pay).
23. Cohen is of course aware of this fact. We have corresponded extensively about its relevance to his argument against Rawls in 1991, and he briefly mentions it in one of the essays here under examination (PAI, p. 170n34: "make-up of the person"; cf. also note 20 below).
24. Or for purposes of increasing corporate profits or the social product - the point applies to any market-driven system of job allocation.
25. Unlike a reservation wage (the conventional notion), a respio can conveniently be assumed to remain constant through changes in the net pay rate for the alternative job. It offers then an easy way of expressing a person's disposition with respect to two jobs: to do laboring work if the reward multiple is below her respio and to do managerial work otherwise. A change in the going reward multiple will not change this disposition, though it may well change which job she chooses on its basis.
26. What if their greater willingness could be selectively exploited? With mind reading a little more advanced, will the difference principle require a regime of individualized income taxes through which the economic-rent portions of special rewards are taxed away as much as possible? If so, the ideal tax regime for our sample world would induce all talented persons with respios under twelve to work as managers, each taxed at her own individualized rate determined by her personal respio. Cohen would like this idea. Rawls, I suspect, would reject it - though it is unclear what reasons he could give within his theory.
27. When persons receive equal pay for equal work, the economic-rent component of their pay may still differ on account of their different specific respios. In a discussion of David Gauthier's work, Cohen suggests that one should here rather speak of "producer surplus" and define "economic rent" independently of the supplier's preferences. Cf. G. A. Cohen, Self-Ownership, Freedom, and Equality (Cambridge: Cambridge University Press, 1995), p. 218.
28. do not follow his suggestion for three reasons: First, I do not understand Cohen's proposal for how the economic-rent component of a wage should be measured without reference to the labor supplier's preferences. Second, economists define "producers' surplus," synonymous with "differential rent," as the price a factor fetches in its present employment minus the price it would fetch in its second most lucrative employment. Cf. John Eatwell et al., ed., The New Palgrave: A Dictionary of Economics (London: Macmillan, 1987), Volume 3, pp. 142-43.
29. This definition, applied to labor, is clearly respio-invariant. Third, economists do include the phenomenon here at issue under the heading of economic rent. See, for example, P. A. Samuelson: Economics, Eighth Edition (New York: McGraw-Hill, 1948/1970), p. 555: "Most of the high earnings of outstanding individuals can probably be classified as 'pure economic rent.' Babe Ruth earned $100,000 a year playing baseball, something he liked to do anyway. Outside the field of sports it is doubtful if he could have counted on earning more than say, $5,000 a year." It is because of Babe Ruth's disposition (respio), then, that most of his income counts as economic rent; it matters that he liked playing baseball anyway and would have supplied (nearly) as much baseball effort even at much lower pay.
30. q and this manager's respio.
31. Analogous problems arise when one conceives economic rent, as economists do (note 20), in terms of a person's second most preferred job/pay package. I am downplaying this conventional conception of economic rent here, because it does not fit Cohen's program: It entails that a top earner who could make as much money in some other (no more burdensome) job receives no economic rent at all.
32. For further discussion of economic-rent-based taxation and its difficulties, see Barbara Fried, The Progressive Assault on Laissez Faire: Robert Hale and the First Law and Economics Movement (Cambridge, MA: Harvard University Press, 1998), pp. 134-59, 203-4, 213.
33. Some have suggested, and Rawls has considered, adding leisure time to this list. The purpose of doing this may already be largely achieved, however, by defining income not as an aggregate (lifetime income, say), but as a net pay rate (pay per hour of work, say). This way, two persons count as being equally well off economically, if they work (in equally good jobs) at the same net pay rate, even if one chooses to work more hours than the other. The extra income had by one is implicitly considered offset by the extra leisure had by the other. To be sure, this way of incorporating leisure is plausible only when persons have some freedom regarding their working hours, a freedom most people lack in the world as it is.
34. This is the "good case" of pay rate inequality that offsets (interpersonally invariant) special burdens. In order to skip unilluminating curlecues, I will, in my simple model, assume that managing and laboring are by these objective criteria equally good jobs. This assumption is, if anything, favorable to Cohen. For, if we made the more natural assumption that managing is an objectively better job than laboring, then his task would become even harder as he would then be challenging an even larger part of managerial net pay.
35. Williams shows quite effectively that Cohen had better take this line (Williams, pp. 235-37). But then he also writes that "the textual evidence is inconclusive" on whether he does (p. 235). On this textual point I differ from Williams: I find much textual evidence for and none against the view that Cohen specifies his expansive understanding of the difference principle so that it forbids Clara's retreat to laboring.
36. PAI, pp. 171-75. In our example, this equal share would be $21 per hour.
37. Again, I am leaving aside jobs imposing objective special burdens. According to both Rawls and Cohen, socioeconomic inequality is erased, rather than created, if such burdens are compensated by an appropriate amount of extra pay.
38. More talented persons being so disposed entails: (a) fewer talented persons disposed as parametric maximizers whom special rewards might attract to a managerial career, hence (b) a lower optimal reward multiple; and thus (c) a higher optimal tax rate on managerial work. It should be kept in mind that Sections III to VII discuss Cohen's stance on a society whose basic structure is fully just by Rawlsian lights. In such a society, the optimal tax regime is actually implemented, and any lasting change in conduct dispositions that affects optimal tax rates thereby affects actual tax rates as well.
39. This informational problem is different from those Williams has invoked to defend Rawls against Cohen. It would remain even if there were no self-deception and all persons were eager to do without special rewards whatever they would be doing if Rawlsian incentives were in place.
40. One might modify this duty, permitting people to choose less productive work provided they make up for the social loss by reducing their income even further (below unit pay).
41. $128,400 weekly product divided by 4,260 weekly work hours. Of course, unit pay could rise further still, if managers agreed to work even longer hours.
42. If laborers worked only 30 hours per week on average, and managers 60, then unit pay could reach $35.31 per hour: $119,700 weekly product divided by 3,390 weekly work hours.
43. Analogous-goal versions of monism may be worth their own investigation. And so may mixed versions, which combine goals of both kinds. Such a mixed version might demand, for instance, that any person or basic structure ought to minimize disrespect of human rights exactly insofar as this is compatible with respecting human rights him-/her-/itself.
44. Understanding this, some utilitarians have maintained that their own doctrine should be effaced from human consciousness, if this were requisite to realizing its content more fully. See Henry Sidgwick, The Methods of Ethics (London: Macmillan, 1907), pp. 489-90,
45. and Derek Parfit, Reasons and Persons (Oxford: Oxford University Press, 1984), §9 and §17.
46. PL, pp. 9-10, 66-71.
47. PL, p. 170; for the model consensus more generally: PL, pp. 145-46, 169-71.
48. His abstinence here instantiates a general effort to keep his theory compatible, as far as possible, with a plurality of more comprehensive doctrines by avoiding denial of any views they set forth as well as expressions of skepticism and indifference toward them (PL, pp. 150-54). Still, this method of avoidance can be taken only so far. Rawls is implicitly rejecting any candidate mastergoal that is clearly opposed to his criterion for basic structure design.
49. As Cohen parenthetically concedes with a reference to competitive sports, calling it a "false generalization ... that the point of the rules governing an activity must be aimed at when agents pursue that activity in good faith" (SDJ, p. 10). More significantly, the generalization also fails for markets (whose participants need not aim at an efficient allocation of resources), court systems (where defense attorneys need not aim at punishment of the guilty), and competitive political systems (where politicians need not aim for content-neutrality and a fair airing of all views).
50. He also seeks to downplay such divergences as unlikely, declaring that, under current conditions, "difference principle equality" cannot be achieved with an ethos that is anything but equality-inspired (SDJ, p. 14). To this one might reply that D cannot be fully achieved at all and that it is not unlikely that more can be done toward raising the lowest socioeconomic position in the U.S., say, by working to promote an ethos of piety and family values than by working to promote an equality-inspired ethos. As concerns his own favored goal, equality of access to advantage, Cohen seems reluctant wholly to let go of either of the two monisms. He distinguishes now between a just distribution and a just society (SDJ, p. 14). About the former, he writes: "My concern is distributive justice, by which I uneccentrically mean justice (and its lack) in the distribution of benefits and burdens to individuals. . . . [D]ifferences of advantage are a function of the structure and of people's choices within it, so I am concerned, secondarily, with both of these" (SDJ, p. 12). About the latter: "A just society, here, is one whose citizens affirm and act upon the correct principles of justice" (SDJ, p. 14). The big question facing him now is: Which justice matters? What is the relative weight of these two competing goals?
51. I am leaving aside here Cohen's appeals to some of Rawls's more uplifting formulations - invoking "the ideals of dignity, fraternity, and the full realization of people's moral natures" - because Cohen himself accepts that such formulations could be toned down or withdrawn and thus are "not decisive against Rawls" (SDJ, p. 17; cf. Williams, p. 229).
52. This problem for consequentialist personal moralities has been extensively discussed. See Bernard Williams, "A Critique of Utilitarianism," in J.J.C. Smart and Bernard Williams, eds., Utilitarianism: For and Against (Cambridge: Cambridge University Press, 1973);
53. Samuel Scheffler, The Rejection of Consequentialism (Oxford: Oxford University Press, 1982);
54. Michael Slote, Common-Sense Morality and Consequentialism (London: Routledge, 1985);
55. and Seana Shiffrin, "Moral Autonomy and Agent-Centered Options," Analysis 51 (1991): 244-54. It is also the central theme of Estlund's critique of Cohen; and so I will not develop it further.
56. A Rawls-inspired picture of this sort has been earnestly filled in by David A. J. Richards, A Theory of Reasons for Action (Oxford: Oxford University Press, 1971),
57. and was criticized for its "righteous absurdity" in Bernard Williams, Moral Luck (Cambridge: Cambridge University Press, 1981), p. 16.
58. See TJ, p. 266, and examples in §§34, 35, 37, 38.
59. As Cohen recognizes: "Appointment to a given job of candidate A rather than candidate B might be judged unjust, even though it occurs within the rules of a just basic structure. But injustice in such a choice is not the sort of injustice that the Rawlsian principles are designed to condemn" (SDJ, p. 11).
60. A point like the one I am making here (with 'basic structure' replaced by 'government') was apparently pressed on Cohen by Ronald Dworkin in a seminar in 1994 (SDJ, p. 13n20). Cohen's response concludes that this "attempt to reduce the distance between Rawls's position and my own threatens to render the former incoherent" by committing it to "three strikingly incongruous" propositions: (1) the difference principle is an egalitarian principle of distributive justice; (2) it requires the basic structure to promote an egalitarian ethos; (3) it is not for the sake of enhancing distributive justice in society that it is required to promote this ethos. (These three propositions are nearly verbatim from Cohen [SDJ, p. 13]) Now why exactly is the combination of these three propositions supposed to be incongruous? Pursuant to (2), let us imagine a society whose basic structure the difference principle requires to be designed so as to promote an egalitarian ethos. Pursuant to (1), this requirement obtains because that egalitarian-ethos-promoting design of the basic structure produces a better distribution than any alternative design would. Having said all this, how can one still deny - Cohen asks - that the basic structure, by optimally promoting D, is "enhancing distributive justice in society" or enhancing the justice of the distribution? To deny this is once more to deny the challenge from mastergoal monism: Rawls offers a theory about what makes a society's basic structure just or unjust. To be just, a basic structure must work so as to generate no worse a distribution, in the social system it organizes, than any feasible alternative would. Rawls's standard for assessing alternative basic structures is then parasitic on a standard for assessing distributions as better or worse. But the asserted validity of this latter standard need not extend beyond the role it plays in Rawls's theory. What counts as a better distribution for purposes of ranking alternative feasible basic structures need not count as a better distribution tout court.
61. This raises the question whether Rawls's view might itself have totalitarian implications of the sort discussed above. This question I must leave for another day.
62. We have seen that talk of what people ought to do/promote is ambiguous between a performative and an inspirational interpretation. Focusing on an objection that either type of monism might make to dualism, the present section deliberately tolerates this ambiguity.
63. On this point I differ from Rawls who - while agreeing that "it seems plausible to hold that, when the distinction is clear, negative duties have more weight than positive ones" (TJ, p. 98) - explicitly classifies as positive our natural duty to uphold justice, placing it alongside natural duties of mutual aid and mutual respect and opposite negative natural duties not to injure and not to harm the innocent (TJ, p. 94). For further discussion of this point, see my "The Bounds of Nationalism," in Jocelyne Couture et al., eds., Rethinking Nationalism, Canadian Journal of Philosophy Supplementary Volume 22 (Calgary: University of Calgary Press, 1998), pp. 463-504.