Hostname: page-component-7c8c6479df-p566r Total loading time: 0 Render date: 2024-03-27T05:49:05.927Z Has data issue: false hasContentIssue false

CONTRACTUALISM'S (NOT SO) SLIPPERY SLOPE

Published online by Cambridge University Press:  11 July 2012

Aaron James*
Affiliation:
Department of Philosophy, University of California, Irvineaaron.james@uci.edu

Abstract

Familiar questions about whether or how far to impose risks of harm for social benefit present a fundamental dilemma for contractualist moral theories. If contractualism allows “ex post” objections by considering actual outcomes, it becomes difficult to justify the risks created by most public policy, leaving contractualism at odds with moral commonsense in much the way utilitarianism is. But if contractualism instead takes a fully “ex ante” form by considering only expected outcomes, it becomes unclear how it recommends something other than aggregative cost-benefit decision-making. Focusing on T.M. Scanlon's version, this paper develops this basic choice of interpretation and recommends the ex ante version. The paper explains how contractualism is inconsistent with John Harsanyi–style utilitarianism and how contractualism supplies a principled framework for walking a careful line between the “bad aggregation” characteristic of utilitarianism and the “good aggregation” that is both unavoidable and fully appropriate in public life.

Type
Research Article
Copyright
Copyright © Cambridge University Press 2012

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1. John Rawls, A Theory of Justice (1971); T.M. Scanlon, What We Owe to Each Other (1998).

2. Scanlon, supra note 1, at 229.

3. Derek Parfit, 2 On What Matters (2011), ch. 21.

4. Ashford, Elizabeth, The Demandingness of Scanlon's Contractualism, 113 Ethics 273302 (2003)CrossRefGoogle Scholar, focuses on the ex post interpretation. Fried, Barbara, Can Contractualism Save Us from Aggregation, 16 J. Ethics 3966 (2012)CrossRefGoogle Scholar, available at http://www.springerlink.com/content/f8v117544h876204/; and Barbara Fried, Is There a Coherent Alternative to a Cost/Benefit Calculus for Regulating Risky Conduct? (unpublished manuscript), emphasizes the choice between ex ante and ex post perspectives. Gibbard, Allan, Reconciling our Aims (Stroud, Barry ed., 2008)CrossRefGoogle Scholar, presents an ex ante–style challenge to explain why we should reject a Scanlon-Harsanyi marriage, which I consider in detail below.

5. The issue is perhaps part of the “legislative stage” of Rawls's “four stage sequence,” though it amounts to a significant further step of application. Rawls, supra note 1, at 195–201.

6. This is suggested in Scanlon, T.M., Contractualism and Utilitarianism, in Utilitarianism and Beyond 122123 (Williams, Bernard & Sen, Amartya eds., 1982)Google Scholar. Parfit's contractualism is especially clear about addressing actual rather than expected outcomes; Parfit, supra note 3, at 1:162. See also Parfit, Derek, Justifiability to Each Person, 16 Ratio 368390 (2003)CrossRefGoogle Scholar.

7. This is suggested in Keating, Gregory C., Irreparable Injury and Extraordinary Precaution: The Safety and Feasibility Norms in American Accident Law, 4 Theoretical Inquiries Law (2003)CrossRefGoogle Scholar; Keating, Gregory C., Pressing Precaution beyond the Point of Cost-Justification, 56 Vand. L. Rev. (2003)Google Scholar; and Keating, Gregory C., Pricelessness and Life: An Essay from Guido Calabresi, 65 Md. L. Rev. (2005)Google Scholar. Keating puts contractualist argument in terms of expected values but without an explicit limitation to them. Rahul Kumar, Risks and Wrongs (unpublished paper), defends a fully ex ante version of contractualism.

8. Rawls's theory is ex ante in several respects. First, a decision made from behind the veil of ignorance is ex ante in the hypothetical sense that it is assumed to be made “in advance” of knowing one's actual situation. Second, even looking beyond the veil, outcomes are to be evaluated as just or unjust only in terms of the justifiability of the institutional procedures that produce them. Third, Rawls's theory is epistemically constrained to the exclusion of an esoteric morality. As Rawls puts the idea, “Conceptions of justice must be justified for the conditions of our life as we know it or not at all”; Rawls, supra note 1, at 398, italics added.

9. The ex ante contractualist must also firmly deny the possibility of “moral luck.” T.M. Scanlon, Moral Dimensions: Permissibility, Meaning, Blame (2008), 148–151, develops an account of how the phenomenon might be explained away. This may be seen to flow from ideas of self-governance and mutual recognition, as discussed in Aaron James, The Significance of Distribution, in Reasons and Recognition: Essays in Honor of T.M. Scanlon (2011).

10. Scanlon, What We Owe, supra note 1, at 204.

11. Principles of conduct are thus agent-relative in the sense that some agents but not others may have the assumed informational frame of reference. The risk that I might kill someone on a trip to the store (when I comply with all established regulations) might be too small for me to take account of. But that same risk imposed over thousands of trips to the store might be relevant for how regulations themselves are set or adjusted when the larger pattern of risk imposition can be known by a regulatory authority.

12. An ex post version can rely on the available evidence in any particular decision, but the independent facts of the case will settle when the right thing is done.

13. The ex ante contractualist can, however, identify who is ultimately wronged with who is ultimately injured. The objection is then that the agent acted while in a position to know that someone in the same (appropriately generically described) situation would likely get hurt. Note also that other forms of ex post decision are not at issue, as when a court is deciding whether someone should be compensated for some harm already done. On the ex ante version, the question here is still about what compensatory efforts could be reasonably expected going forward, given the available evidence about what might address the prior wrong done. For the ex post variant, what is at issue the harmful outcome and restoration of the victim in and of itself.

14. Reibetanz, Sophia, Contractualism and Aggregation, 108 Ethics 296311 (1998)CrossRefGoogle Scholar; and Lenman, James, Contractualism and Risk Imposition, 7 Pol. Phil. & Econ. (2008)Google Scholar.

15. Scanlon, What We Owe, supra note 1, at 208.

16. Id. at 209.

17. Id.

18. Ashford, supra note 4; and Fried, Can Contractualism, supra note 4.

19. Broome, John, Trying to Value a Life, 9 J. Pub. Econ. 91100 (1978)CrossRefGoogle Scholar; which is cited in Fried, Can Contractualism, supra note 4.

20. Scanlon, What We Owe, supra note 1, at 209.

21. I assume the overall argument of this paragraph would not be undercut by normalizing relevant objections. That is, it might seem that any principles for the regulation of conduct can be only so specific about which risks an addressed agent is supposed to track and manage, in which case potential objections to harm cannot be narrowly tailored to specific cases, especially when the risks of harm are exceedingly low. This has considerable force on the ex ante version of contractualism but not on the ex post version. On the ex post version, the potential objection to harm is still the harm that actually happens, while the potential demands of self-governance (concerning gathering information, forgoing valuable opportunities, and so on) are merely countervailing objections. They will not carry the day when premature death is in question.

22. Id. at 232; following Frances M. Kamm, 2 Morality, Mortality (1993), at 116–117.

23. Ashford, supra note 4, takes this to support utilitarianism, which is then not “too demanding,” at least not by comparison to contractualism.

24. The example is due to Kamm, supra note 22, at 303.

25. As in “Ambulance II,” from Frances M. Kamm, Intricate Ethics (2007), at 273.

26. The justified background exemption from traffic rules means the driver could not have been expected to avoid creating the dangerous situation. He can justify having set in motion the process that may turn out to have a fatal moment of carelessness; in that case, it is only the moment of carelessness that is unjustified.

27. Lenman, supra note 14, at 116, suggests that reasonable ex post objections will take into account considerations of what the agent could have known and the ensuing strains of commitment, so as to blunt the contrast between ex post and ex ante objections. As suggested in note 21supra, such burdens of governance would still only be countervailing considerations, and it is hard to see how they would be decisive in the face of powerful ex post objections to premature death. They gain greater force within a fully ex ante framework that compares risks of death rather than the bare fact of death itself.

28. Versions of this question are pressed in different ways by Reibetanz, supra note 14; and Fleurbaey, Marc, Assessing Risky Social Situations, 118 J. Pol. Econ. 649680 (2010)CrossRefGoogle Scholar.

29. Johann Frick, Contractualism and the Ethics of Risk (unpublished manuscript), nicely develops this thought.

30. For development of this theme in light of the idea of “moral recognition,” see James, supra note 9.

31. Harsanyi, John, Cardinal Utility in Welfare Economics and in the Theory of Risk Taking, 61 J. Pol. Econ. 434435 (1953)CrossRefGoogle Scholar. See John Broome, Weighing Goods (1991), at 51–58, for discussion and criticism.

32. Fried, Can Contractualism, supra note 4, suggests this at several points, though in conjunction with her several challenges to proposed limitations on aggregation.

33. Harsanyi, supra note 31.

34. For discussion, see Rawls, John, Justice as Fairness: A Restatement (Kelly, Erin ed., 2001), at 106109Google Scholar. When a veiled decision concerns one's whole life prospects, and there are grave prospects of profound disadvantage under pervasively influential basic institutions (e.g., being enslaved in a utilitarian society), what is socially acceptable from a moral point of view is not adequately represented by a rational gamble under conditions of risk when probabilities are known (to be equal, in this case) and one is willing to take one's chances (with a gambler's modest aversion to risk). The more appropriate model is what one would choose under conditions of fundamental uncertainty (the parties simply do not know who they are), with a correspondingly heightened aversion to winding up in a permanently disadvantaged state. Rational choosers then follow the maximin rule of choice and thus reject an aggregative principle of social choice in favor more egalitarian principles. Rawls's claim is that only the more cautious representation will explain to those who wind up in the worst position why, save the demands of impartiality, they are being given their full due.

35. Gibbard, supra note 4.

36. Scanlon, Contractualism, supra note 6, at 120–128.

37. Harsanyi, John, Cardinal Welfare, Individualistic Ethics, and Interpersonal Comparisons of Utility, 63 J. Pol. Econ. 309321 (1955)CrossRefGoogle Scholar.

38. Gibbard, supra note 4, at 50–52, 59–60.

39. See the “Complaint Model,” which Scanlon, What We Owe, supra note 1, at 229, discusses and rejects, as well as the discussion in Reibetanz, supra note 14. Scanlon, What We Owe, supra note 1, ch. 3, rejects the assumed welfarism, admitting as morally relevant only aspects of welfare, along with other kinds of complaints.

40. Broome, Weighing Goods, supra note 31, chs. 8 and 10; Gibbard, supra note 4, at 42–43, 64–66; and Broome's formulation at Broome, Weighing Goods, supra note 31, at 114–115. It is worth emphasizing that in any case, a workable basis for interpersonal comparisons may have little relation to traditional utilitarian concern with personal welfare.

41. Broome, Weighing Goods, supra note 31, at 58, summarizes his extended development of the point as follows: Harsanyi's equiprobablity argument “does not really yield philosophically interesting information about the nature of morality. Instead, it makes a philosophically unsupported assumption about the nature of morality.” For a similar criticism, see Brian Barry, Theories of Justice: A Treatise on Social Justice (1989), at 334–335.

42. Thus, as Scanlon explains, in considering relevant grounds for mutual aid, we ignore the fact that the fortunate Joneses are unlikely to at some point themselves require assistance. We instead consider their relevant interests under more generic descriptions such as “the importance of being able to get aid should one need it” and “the degree of inconvenience involved in giving it, should one be called upon to do so.” Scanlon, What We Owe, supra note 1, at 207.

43. See Broome, Weighing Goods, supra note 31, at 94–95.

44. I am grateful to A.J. Julius for drawing my attention to this line of objection.

45. Diamond, Peter, Cardinal Welfare, Individualistic Ethics, and Interpersonal Comparisons of Utility: Comment, 75 J. Pol. Econ. 765766 (1967)CrossRefGoogle Scholar.

46. For detailed explication, see Broome, Weighing Goods, supra note 31, at 111.

47. Id. at 111–112.

48. Lenman, supra note 14, at 100.

49. Id. at 107.

50. Among “aggregative decisions” we might include any aggregative cost-benefit calculation, including policies that require them of officials (perhaps in particular specified forms), whether or not the decision's licensing principle is itself aggregative in either its content or underlying justification.

51. For discussion, see Aaron James, Fairness in Practice: A Social Contract for a Global Economy (2012), ch. 7.

52. Fried presses this objection in Fried, Can Contractualism, supra note 4, sec. 3.2.

53. Keating, Irreparable Injury, supra note 7.

54. Reibetanz, supra note 14, at 301–302, explains how Scanlon's limits on aggregation can be rephrased entirely in terms of expected benefit and harm: taking regulative principles in light of their expected consequences over time, one's prospects are assessed in light of one's overall chances, over time, of falling into groups of larger or smaller numbers of persons who could require assistance.

55. Scanlon's transmission towers case might thus be different from the case of Jones the World Cup technician, because Jones is not assumed voluntarily to take on risks of getting stuck in the transmission machinery. Either the technical problems are assumed to be a freak accident and are not simply known to come with the job, or Jones is assumed not to have adequate alternative employment options. Scanlon, What We Owe, supra note 1, at 236.

56. This is not to justify any outcomes that result; the continuous reduction of workplace risks through sustained safety standard improvement with international support can still be required, because such measures are reasonably affordable. This is an instance of the point above that the permissibility of using an aggregative decision rule on Harsanyian grounds depends not only on general ex ante benefit but also on the assumption that no one can mount a reasonable complaint against the risks he or she is subjected to. For detailed discussion of the case, see James, Fairness, supra note 51, ch. 10.

57. One might thus admit the relevance of numbers in Parfit's “Case Four,” in Parfit, On What Matters, supra note 3, at 197; and “Case Six,” id. at 200–201. In the specified institutional contexts, we can fail to give one person a much greater benefit in order to give equal or comparable but much smaller benefits to a much larger number of people (and without abandoning Scanlon's restriction to personal reasons). Though Parfit intends such cases also to apply outside of our specified institutional setting.

58. Scanlon, What We Owe, supra note 1, at 237 n.42, mentions this basis for aggregation.

59. This is “Case 6” in Parfit, On What Matters, supra note 3, at 200.

60. “Threshold deontology” offers one version of this claim. I mean to suggest a more general line of argument, which may or may not assume that claims are always sensitive to some aggregative opportunity cost threshold.

61. The point also applies to Parfit's “Case 7” (id. at 203), in which the options are either equally permissible or underdescribed. Parfit's “Case 5” (id. at 199) may be different because of implicit reference to normalcy of life span in a sense that is broadly subject to societal influence. So, e.g., among people with life-shortening medical conditions, we can use scarce medical resources either to give one person a normal life span (forty more years, instead of an early death at thirty years of age) or marginally to extend a larger number of normal life spans (from seventy to seventy-five total years). Here one might lay a claim of justice to a measure of normalization. But the claim need not be decisive. Aggregative concerns of Institutional Mandate may still apply.

62. This may require rejecting what Parfit calls Scanlon's “Greater Burden Claim” (id. at 192). Given that claim, reasonable rejection of a principle depends in part on its “burden” as compared to an alternative principle's being in force, where this always includes any benefits one would forgo. The suggestion in the text is that we instead treat forgone benefits as relevant only in appropriate contexts, as supported by context-specific generic reasons claims. Relaxing the Greater Burden Claim as a sweeping constraint relieves some of the pressure Parfit places upon Scanlon's restriction to personal reasons. For this and other grounds for aggregation, Parfit may therefore be mistaken to suggest (id. at 200) that “It is clearly Scanlon's Individual Restriction which is making Scanlon's Formula go astray.” On the contrary, the Individual Restriction is not clearly the culprit.

63. Louis Kaplow & Steven Shavell, Fairness versus Welfare (2002), emphasize this point. From a contractualist point of view, they mistakenly assume that considerations of greater benefit are decisive without offering necessary moral arguments.

64. Waldron, Jeremy, Moments of Carelessness and Massive Loss, in Philosophical Foundations of Tort Law (Owen, David G. ed., 1995)Google Scholar, offers a Rawlsian defense of a no-fault scheme.

65. I assume there are alternatives, such as following established administrative practice (as distinct from relying on cost-benefit constructions) with sensitivity to political pressure or special interests. Officials could also simply muddle along, deciding by moral intuition or assumed principle. There is also, of course, a range of cost-benefit approaches for any particular case, only some of which may be justifiable.

66. Scanlon, What We Owe, supra note 1, at 221, allows that personal claims, e.g., against environmental degradation might be bolstered by the impersonal value of environmental preservation. Thus the fact that the overall cost of aggressive climate change mitigation is relatively small (according to Björn Lomborg, The Skeptical Environmentalist (2001), at 323, it amounts to one in fifty years of economic growth) might strengthen personal objections to anything less than aggressive measures. Personal objection might also arise from the personal interest we each have in the nature of the society in which we live, including, e.g., a personal interest in societal goods such as national security or national wealth. For the case of the national “gains of international trade,” see James, Fairness, supra note 51, ch. 6.

67. So, e.g., in the unexploded mine case discussed in Reibetanz, supra note 14, at 302, we can treat the case as one of easy rescue: one knows that each worker (in a group of one hundred) has a substantial chance of death, while one is merely sure to become ill by a rescue effort. Because the claims of each worker to assistance are not discounted (according to the specific probability that he or she among the one hundred will be the one injured), the substantial chance of each worker's death thus defeats one's appeal to one's own high risk of mild illness. This suggestion is similar to the objection mentioned in id. at 303 n.12, and I believe it is not undermined by the suggested reply, which appeals to an ex post reading of contractualism.

68. Scanlon, What We Owe, supra note 1, at 209.

69. One could still imagine a further case in which only a forced medical experimentation lottery would reliably spur medical advances that would save millions from death. The issue here is still not one's probability of harm but how much one's interest in bodily self-determination weighs against the possibility of saving lives—that is, whether or to what extent that interest has a silencing role. Claiming that it does have a silencing role would not require relaxing Scanlon's restriction to personal reasons, as Parfit, On What Matters, supra note 3, at 209–210, recommends in light of a similar example.

70. The lottery could run in two stages: a first stage that decides whether someone will be executed, and a second stage that decides who. The first stage is heavily weighted against anyone being chosen, and by chance has never yet moved to the second selection stage.

71. But will contractualism guide decision-making if it is not made less judgment-dependent somehow? It arguably still can. First, policy-making is invariably shaped by value-judgment. Second, contractualism can in any case offer principles that frame and in that sense guide policy choice. Third, the possibility of an aggregative cost-benefit calculus that provides “determinacy” without judgment is largely illusory: unless any decision rule can be appropriately morally grounded, in part by way of moral judgment, it is no more justified than a determinate toss of a coin.

72. Although Scanlon has come to reject the relevance of intention to permissibility in Scanlon, Moral Dimensions, supra note 9, ch. 1.