Skip to main content

Interpersonal Comparisons of the Good: Epistemic not Impossible


To evaluate the overall good/welfare of any action, policy or institutional choice we need some way of comparing the benefits and losses to those affected: we need to make interpersonal comparisons of the good/welfare. Yet sceptics have worried either: (1) that such comparisons are impossible as they involve an impossible introspection across individuals, getting ‘into their minds’; (2) that they are indeterminate as individual-level information is compatible with a range of welfare numbers; or (3) that they are metaphysically mysterious as they assume the existence either of a social mind or of absolute levels of welfare when no such things exist. This article argues that such scepticism can potentially be addressed if we view the problem of interpersonal comparisons as fundamentally an epistemic problem – that is, as a problem of forming justified beliefs about the overall good based on evidence of the individual good.

Hide All

1 Robbins, Lionel, An Essay on the Nature and Significance of Economic Science (London, 1962), p. 140, and see paper VI for the general discussion.

2 Jevons, Stanley, Theory of Political Economy (London, 1871), p. 21 .

3 Nozick, Robert, Anarchy, State and Utopia (New York, 1974), p. 41 .

4 Arrow, Kenneth, Social Choice and Individual Values (New York, 1963), p. 9 .

5 Arrow, Social Choice, p. 11.

6 Harsanyi, John, ‘Cardinal Welfare, Individualistic Ethics, and Interpersonal Comparisons of Utility’, The Journal of Political Economy 63 (1955), pp. 309–21, at 317.

7 Connee, Earl and Feldman, Richard, Evidentialism (Oxford, 2004), p. 83 . Evidentialism would, prima facie, treat informationally identical evidence as supporting the relevant hypotheses identically.

8 The type of multi-parameter cases where uniform Bayesian priors are problematic, such as that of forming priors over a factory's cube sizes, do not apply in the case of beliefs about the good/welfare since this represents only one parameter. For the famous cube example see van Fraassen, Bas C, Laws and Symmetry (Oxford, 1989), p. 303 .

9 Scanlon, Timothy, What We Owe to Each Other (Cambridge, MA, 1998).

10 Rawls, John, A Theory of Justice (Cambridge, MA, 1999). Note that unbiased beliefs about the overall good need not for a Rawlsian entail that maximin be rejected; the point of maximin is a rejection of maximizing the overall good, not of the possibility of the overall good.

11 This idea is of course famously in Bentham, Jeremy, An Introduction to The Principles of Morals and Legislation (Oxford, 1907), well discussed in Sidgwick, Henry, The Methods of Ethics, 7th edn. (London, 1907) and expanded to non–human animals and further defended as a basic commitment of morality in Singer, Peter, Practical Ethics (Cambridge, 1993).

12 A proof: E[R / E] = E[R] + yn(r +) − yn(r −); E[B / E] = E[R] + yn(b +) − yn(b −); as y>0 and, from PI, E[R] = E[B], therefore E[R / E] > E[B / E] iff n(r +) − n(r −) > n(b +) − n(b −). QED.

13 See notably Von Neumann, John and Morgenstern, Oskar, The Theory of Games and Economic Behavior (Princeton, 1944).

14 This, for instance, is an enabling premise of falsificationism. Popper, See Karl, Conjectures and Refutations: The Growth of Scientific Knowledge (London, 2002).

15 Hammond rightly, but perhaps over-diplomatically, characterizes using monetary value as an implicit welfare proxy as ‘almost certainly unethical’. Hammond, Peter, ‘Interpersonal Comparisons of Utility: Why and How They Are and Should Be Made’, Interpersonal Comparisons of Well-Being, ed. Elster, Jon and Roemer, John (Cambridge, 1991), pp. 200–54, at 201.

16 See for instance Broome, John, Weighing Lives (Oxford, 2006) or, for a famous outline of the general temporal difficulties, Parfit, Derek, Reasons and Persons (Oxford 1984), esp. chs. 8, 16, 17, 18 and 19.

17 I am very grateful for extremely helpful comments on this article from participants of a 2011 NYU Political Theory group meeting, a 2011 LSE Choice Group session, a 2012 Harvard-MIT special political theory seminar, and a 2015 Paris seminar in Normative Political Philosophy at the Ecole des Hautes Etudes en Sciences Sociales, and also to detailed comments from Russell Hardin, Sean Ingham, Michael Kates, Dimitri Landa, Christian List, Bernard Manin, Michael Rosen, Kai Spiekermann and Lucas Stanczyk.

18 Proof: As (h ∧ ¬h) → h therefore Cr(h ∧ ¬h) ≤ Cr(h). As h → (h v ¬h) therefore Cr(h) ≤ Cr(h v ¬h). (In fact, given Additivity, not only does Coherence entail Boundedness, but the converse is true too. Proof: As Y → ¬ (¬Y) therefore Cr(Y) + Cr(¬Y) = Cr(Y v ¬Y) [From A1]. If X → Y then Cr(X v ¬Y) = Cr(X) + Cr(¬Y) [From A1]. Thus Cr(X v ¬Y) + Cr(¬X ∧ Y) = Cr((Xv ¬Y) v (¬X ∧Y)) = Cr(Y v ¬Y). Therefore Cr(X) + Cr(¬Y) + Cr(¬X ∧ Y) = Cr(Y) + Cr(¬Y). As Cr(¬X ∧Y) ≥ 0 [From Boundedness] therefore if X → Y then Cr (Y) ≥ Cr (X) QED.)

Recommend this journal

Email your librarian or administrator to recommend adding this journal to your organisation's collection.

  • ISSN: 0953-8208
  • EISSN: 1741-6183
  • URL: /core/journals/utilitas
Please enter your name
Please enter a valid email address
Who would you like to send this to? *


Full text views

Total number of HTML views: 0
Total number of PDF views: 0 *
Loading metrics...

Abstract views

Total abstract views: 0 *
Loading metrics...

* Views captured on Cambridge Core between <date>. This data will be updated every 24 hours.

Usage data cannot currently be displayed