Rescuing Rawls’s Institutionalism and Incentives Inequality


G. A. Cohen argues that Rawls’s difference principle is incompatible with his endorsement of incentives inequality—higher pay for certain professions is just when that pay benefits everyone. Cohen concludes that Rawls must reject both incentives inequality and ‘institutionalism’—the view that egalitarian principles, including the difference principle, apply exclusively to social institutions. I argue that the premises of Cohen’s ‘internal criticism’ of Rawls require rejecting two important parts of his theory: a ‘subjective circumstance of justice’ and a ‘shared conception of justice’. These are important parts of Rawls’s ‘constructivism’. Constructivism is the view that a conception of justice is the solution to a practical interaction problem and is solved in part by examining the facts of the problem. Thus, an upshot of my arguments is that this Cohen/Rawls dispute reduces to another dispute over ‘constructivism’.

This is a preview of subscription content, log in to check access.


  1. 1.

    See John Rawls (1971, pp. 302, 75–80) and (2001, pp. 42–43).

  2. 2.

    Cohen developed his criticism of Rawls’s endorsement of incentives inequality in many works spanning two decades. I will focus on his most developed version of it in (2008).

  3. 3.

    Cohen imagines that Rawls would give this response, see (Cohen 2008: chapter 3). For an example of other basic structure responses, see Andrew Williams (1998) and (2008). Cohen’s argument has generated much discussion. For examples of various types of responses to Cohen, see Kenneth Baynes (2006), Thomas Christiano (2010), Joshua Cohen (2002), Norman Daniels (2006), David Estland (1998), Aaron James (2005), Kasper Lippert-Rasmussen (2008), Thomas Pogge (2000), Miriam Ronzoni (2008), Samuel Sheffler (2006), Kok-Chor Tan (2012), and Alan Thomas (2011). For papers defending Cohen’s argument, see Brian Berkey (2015, 2016), Paula Casal (2013, 2015), Liam Murphy (1999), and Serena Olsaretti (2012).

  4. 4.

    For example, see Rawls (1971, pp. 54–55; 1993, pp. 261, 283; 2001, pp. 7, 10, 162–168).

  5. 5.

    See (2008: chapter 3). Cohen’s main response to the objection is that Rawls’s specification of the basic structure is ‘fatally ambiguous’. For responses to Cohen’s challenge, see Julius (2003), Melenovsky (2013), Tan (2012), Louis-Philippe Hodgson (2013), Ronzoni (2008), Scheffler (2006), and Williams (1998).

  6. 6.

    On the institutional approach, see for example, Tan (2012).

  7. 7.

    Williams (1998, p. 246) also charges Cohen’s internal criticism of Rawls’s theory with not really being an internal criticism—that is, its premises actually require the rejection of parts of the theory. However, as stated above, Williams’s argument is a kind of basic structure objection. He argues that the internal conflict is only apparent since Rawls claims that the difference principle applies only to the basic structure. Williams then claims that Cohen’s ethos cannot be part of the basic structure because it cannot be ‘public’, and the basic structure is a public system of rules. Rather than holding that Cohen’s premises ignore a part of Rawls’s theory, I claim that his premises require the rejection of two important, and consistent, parts of Rawls’s theory and therefore cannot show that there is even an apparent conflict within Rawls’s theory. My argument is novel in claiming that Cohen’s premises require the rejection of two parts of Rawls’s just society—the ‘conflicting claims circumstance’ and the ‘shared conception of justice’. I will note how my arguments differ from other responses to Cohen along the way.

  8. 8.

    Rawls uses this name too (Rawls 1971, p. 257).

  9. 9.

    For example, see Rawls (1971, pp. 75–80).

  10. 10.

    Applying the principle does not rely on empirically discerning a benchmark of equality (the argument for it requires a theoretical benchmark, as I explain below). Rawls notes that if we maximize the position of the worst off relative to other social arrangements, and everyone benefits by doing this, the benchmark is irrelevant in practice ‘if not largely impossible to ascertain anyway’ (1971, p. 80).

  11. 11.

    See (1971, pp. 78, 151, 157, 279; 2001, pp. 53, 63).

  12. 12.

    For Rawls’s principles and their ordering, see for example (1971, pp. 302–303; 2001, pp. 42–43).

  13. 13.

    See, for example, Cohen (2008, pp. 383–387).

  14. 14.

    Cohen (2008, pp. 383–387) also thinks that Rawls’s difference principle would allow too much inequality, but his argument for this claim is distinct from his criticism of Rawls’s incentives argument (cf. Cohen 2008, p. 62).

  15. 15.

    For example, see Cohen (2008, p. 56).

  16. 16.

    See, for example, Cohen (2008, pp. 35, 62, 388, 393). My defense of the incentives argument does not endorse all incentives inequality. Some inequality is too great and would be ruled out by Rawls’s theory, and some motivations for wanting incentives inequality are objectionable by Rawls’s lights. For example, some motivations will conflict with the sense of justice which includes the desire to comply with the ‘natural duties’ including mutual-aid (Freeman 2007, p. 123). Motivations regarding relative position would also be unlikely to arise in a Rawlsian well-ordered society (Baynes 2006, pp. 190–195; Cohen 2002, pp. 380–384).

  17. 17.

    See Cohen (2008, pp. 74–75). Cohen cites (Rawls 1999, p. 315). Also see, for example, (Rawls 1971, pp. 138, §69).

  18. 18.

    Cohen often seems to base his argument entirely on the choice/necessity claim. However, he maintains that his only argument for the claim that the talented cannot demand incentives is his ‘justificatory community argument’. See Chapter 1 of Cohen (2008) for the argument and (2008, pp. 121 footnote 11, 122 footnote 15) for instances of his claim that his argument against incentives inequality relies on his justificatory community argument.

  19. 19.

    See, for example, Cohen’s remark (2008, p. 15) that Rawls is committed to the idea of community or as he calls it ‘fraternity’. Cohen explains Rawls’s views on community—including fraternity and civic friendship—in Chapters 1, 3. For his explanation that justificatory community is a requirement of community, see (2008, pp. 44–45).

  20. 20.

    See (2008, pp. 41–48).

  21. 21.

    For example, (2008, pp. 44–45).

  22. 22.

    For this argument, see Cohen (2008, pp. 48–74).

  23. 23.

    See Rawls (1971, p. 127).

  24. 24.

    Rawls claims that this subjective circumstance results when people claim a greater rather than a smaller share of the social product, rather than their claiming a greater share because they want more resources than others (1971, p. 144). He only claims to make weak assumptions about the circumstances and assuming they want more than others is a controversial assumption about the motivations of the cooperating subjects. For simplicity, I will just say that people claim more resources. In (2001, pp. 84–85; cf. 2001, pp. 63–64) Rawls explains that the fact of reasonable pluralism does not mean that people, or the parties who represent them in the original position, want as much as possible. Instead, the parties aim to secure primary goods for the person they represent, and they are not willing to compromise the interests of the person they represent. The important point that gets us a conflict of claims based on securing interests is that people tend to claim more rather than fewer resources to help them most effectively pursue their idea of the good.

  25. 25.

    I discuss this circumstance in terms of distribution (e.g., I noted that moderate scarcity is a precondition of conflicting claims and the conflicting claims are claims to resources) since that is the debate I am addressing (Rawls also explains the circumstance in mainly distributive terms in 1971). Rawls adds to this subjective circumstance and calls it ‘the fact of reasonable pluralism’ in his later works (e.g., 1993, 2001). The circumstance may be interpreted as a deeper conflict in his later works since he abandons his ‘congruence argument’—the argument for the stability of a just society in Part III of (1971) that aims to show the congruence of justice and the good. Instead, he comes to accept that people have irreconcilable ‘comprehensive moral or religious doctrines’ that lead to conflicts in claims to liberties and opportunities (e.g., 2001, pp. 84–85). One’s idea of the good is normally ‘set within’ and ‘interpreted by’ one’s comprehensive doctrine (Rawls 2001, p. 19). These doctrines are incommensurable, which leads Rawls to interpret his theory as a ‘political’ theory and abandon the theory’s ‘partially comprehensive’ elements (elements that presupposed the truth of parts of a comprehensive doctrine), such as congruence. On congruence and Rawls’s ‘political turn’, see for example, Freeman (2003), and Weithman (2013). Rawls (1993, pp. 64–65) also notes that ‘pluralism as such’ and ‘reasonable pluralism’ can be assumed as a circumstance of justice at the stage of principle choice. For my purposes, the subjective circumstance can be expressed either as the ‘conflicting claims circumstance’ (as found in 1971), reasonable pluralism, or as pluralism as such. In other words, I claim that Cohen’s ‘internal criticism’ must reject this circumstance of justice in any of Rawls’s works.

  26. 26.

    For this example, see Cohen (2008: Chapter 5, Section 2).

  27. 27.

    See, for example, Rawls (1993, pp. 269–271).

  28. 28.

    For an argument that a conception of justice is unreasonable if it idealizes away, or is incompatible with, a circumstance of justice, see (Hope 2010).

  29. 29.

    For this response, see Cohen (2008, pp. 71–72). For Cohen’s comments on entitlement and desert, see, for example, (2008, pp. 47, 120).

  30. 30.

    Cohen is quoting (Rawls 1999, p. 347).

  31. 31.

    For another way that the talented might justify their behavior, see (Christiano 2010; Daniels 2006). Their strategies are different than mine. For instance, Daniels (2006, p. 269) argues that the better off can remind the worst off that they are both committed to a ‘division of labor’, that is, the principles of justice apply to the basic structure leaving space for people to pursue their pluralistic ideas of the good and comprehensive doctrines. Their commitment to the liberties and opportunities of a free society leads to the fact of reasonable pluralism, and the division of labor protects a space to pursue one’s idea of the good and comprehensive doctrine. For an objection to Cohen’s interpersonal test, see Lippert-Rasmussen (2008, pp. 435–439).

  32. 32.

    Indeed, Casal (2015, p. 823) claims that the Rawlsian motivation for rejecting most egalitarian ethoses is that many conflict with liberties. Requiring people to choose a certain occupation, for example, restricts liberty.

  33. 33.

    I thank an anonymous referee for this objection.

  34. 34.

    For instance, see (2001, p. 43 footnote 3, §§34–39). He also rejects maximin as justifying the difference principle.

  35. 35.

    For the natural duties and obligations, see Rawls (1971, pp. 333–350).

  36. 36.

    See, for example, Rawls (1971, pp. 9–10, 13, 133, 453–453; 1993, p. 67; 1999, pp. 324–325; 2001, p. 121). Seana Shiffrin (2010) uses this part of Rawls’s publicity condition to argue against inequality inducing incentives. However, I am focused on defending Rawls’s institutional approach, which Shiffrin does not criticize. Tan (2012, pp. 64–66) nevertheless argues that Shiffrin’s argument assumes that institutionalism is unreasonable.

  37. 37.

    Michael Titelbaum (2008) also argues that the parties would not choose Cohen’s ethos. Instead, they would choose a ‘full ethos’—the parties would choose individual correlates of each of Rawls’s principles with their lexical orderings. I only consider the difference principle for individuals—assuming the other principles are satisfied—and argue that the parties would reject it because it requires taking an interest in one another’s interests. This means that my argument also rejects Titelbaum’s ‘full ethos’ since it includes a difference principle correlate for individuals.

  38. 38.

    Cf. (1971, pp. 13, 129, 148–149).

  39. 39.

    Rawls compares mutual disinterest with the reasoning for utilitarianism in (1971: Sect. 6, 30).

  40. 40.

    For other arguments for the difference principle, see Rawls (2001, pp. §§ 35–37).

  41. 41.

    For example, see Rawls (1993, pp. 173–211; 1971, pp. 31–32, 548, 563). Also see Tan (2012, p. 29).

  42. 42.

    Blain Neufeld and Chad Van Schoelandt (2014) argue that the basic structure is coercive and so requires principles that all free and equal citizens can accept. However, they also claim that we should promote principles of ‘ethos justice’, principles of fairness and reciprocity including Cohen’s ethos (2014, pp. 90–91), which apply outside of coercive structures. They note however, that ethos justice is a ‘partially comprehensive doctrine’ that cannot be coercively enforced and so we cannot expect full endorsement in a pluralist society (2014, pp. 93–94). As I argue above, the parties are concerned with adjudicating conflicting claims, so they will not consider ‘ethos justice’ since these principles can conflict with many others.

  43. 43.

    Recall (footnote 32) that Daniels is offering a ‘division of labor’ argument (cf. Tan 2012). My explanation of why my original position argument rules out the difference principle for individuals resembles Daniels’s and Tan’s arguments. However, I am explaining why I think the strict difference principle could not be justified in the original position, which was the conclusion of my previous argument; they instead offer a general argument about the value of pluralism.

  44. 44.

    Joshua Cohen (2002, pp. 380–384) argues that because different basic structure designs will shape the interests people will have in society, the parties could use this tendency to make decisions about how to organize the basic structure. It is important to see the difference between his claim and the claim that applying the difference principle to individuals would bring about a more rational distribution for the parties. The parties can examine the effect of a basic structure design on shaping the interests of people in society because they do not know what the particular interests will be of those who they represent and the principles they choose will partly determine these interests. So they can proceed to choose the basic structure that best secures the fundamental interests of the citizen they represent, whatever the particular interests turn out to be of who they represent. They can therefore determine whether the principles would create a stable society or one in which citizens would tend to be concerned about their relative shares, which may lead to excusable envy, harm to self-respect, etc. The parties can make these decisions while still modeling the conflicting claims circumstance because the decisions do not involve taking an interest in one another’s interests. It only involves securing primary goods, such as wealth, for the fundamental interests of those they represent, whatever citizens’ particular interests turn out to be.

  45. 45.

    See (Korsgaard 2010: chapter 10).

  46. 46.

    This is why, for example, Cohen cannot claim that justice as fairness and his ethos would overcome the conflicting claims circumstance on Rawls’s understanding of a just society. For Rawls, conflicting claims is part of his original position since his constructivist concept of social justice is the response to conflicting claims. Claims conflict because persons aim to protect their ideas of the good by way of rights, resources, and opportunities (2001, p. 85). Indeed, Rawls holds that justice would still be necessary among ‘saints and heroes’ since their interests can be as opposed as any interests (1971, p. 129). He notes that we can imagine a world that corrects for inequalities allowed by his principles but insists that ‘Conceptions of justice must be justified by the conditions of our life as we know it or not at all’ (1971, p. 454). ‘Overcoming’ conflicting claims therefore goes beyond justice to some utopia where people are more than ‘saints and heroes’. For instance, Rawls claims that Marx’s full communist society ‘seems to be one beyond justice in the sense that the circumstances that give rise to the problem of distributive justice are surpassed, and citizens need not, and are not, concerned with it in everyday life. Whereas justice as fairness assumes that, given the general facts of the political sociology of democratic regimes (e.g., the fact of reasonable pluralism), the principles and political virtues falling under justice of various kinds will always play a role in public political life. The evanescence of justice, even of distributive justice, is not possible, nor, it seems, is it desirable’ (Rawls 2001, p. 177; 2008, pp. 321–322). I thank an anonymous referee for this criticism, which allows me to stress further the drastic difference between Cohen’s and Rawls’s concepts of justice.

  47. 47.

    See Cohen (2008, part II).

  48. 48.

    Cf. Korsgaard (2010, p. 325).


  1. Baynes, K. 2006. Ethos and Institution: On the Site of Distributive Justice. Journal of Social Philosophy 37(2): 182–196.

    Article  Google Scholar 

  2. Berkey, B. 2015. Double Counting, Moral Rigorism, and Cohens Critique of Rawls: A Response to Alan Thomas. Mind 124(495): 849–874.

    Article  Google Scholar 

  3. Berkey, B. 2016. Against Rawlsian Institutionalism about Justice. Social Theory and Practice 42(4): 706–732.

    Article  Google Scholar 

  4. Casal, P. 2013. Occupational Choice and the Egalitarian Ethos. Economics and Philosophy 29(01): 3–20.

    Article  Google Scholar 

  5. Casal, P. 2015. Marx, Rawls, Cohen, and Feminism. Hypatia 30(4): 811–828.

    Article  Google Scholar 

  6. Christiano, T. 2010. Cohen on Incentives, Inequality, and Egalitarianism. In Essays on Philosophy, Politics & Economics: Integration and Common Research Projects, ed. C. Favor, G. Gaus, and J. Lamont, 173–200. Stanford: Stanford Economics and Finance.

    Google Scholar 

  7. Cohen, G. A. 2008. Rescuing Justice and Equality. Harvard, MA: Harvard University Press.

    Google Scholar 

  8. Cohen, J. 2002. Taking People as They Are? Philosophy and Public Affairs 30(4): 363–386.

    Article  Google Scholar 

  9. Daniels, N. 2006. Democratic Equality: Rawls’s Complex Egalitarianism. The Cambridge Companion to Rawls, 241–276. Cambridge: Cambridge University Press.

    Google Scholar 

  10. Estlund, D. 1998. Liberalism, Equality, and Fraternity in Cohen's Critique of Rawls. Journal of Political Philosophy 6(1): 99–112.

    Article  Google Scholar 

  11. Freeman, S. 2003. Congruence and the Good of Justice. The Cambridge Companion to Rawls, 277–315. Cambridge: Cambridge University Press.

    Google Scholar 

  12. Freeman, S. 2007. Rawls. New York, NY: Routledge.

    Google Scholar 

  13. Hodgson, L.-P. 2013. Why the Basic Structure? Canadian Journal of Philosophy 42(3–4): 303–334.

    Google Scholar 

  14. Hope, S. 2010. The Circumstances of Justice. Hume Studies 36(2): 125–148.

    Article  Google Scholar 

  15. James, A. 2005. Power in Social Organization as the Subject of Justice. Pacific Philosophical Quarterly 86(1): 25–49.

    Article  Google Scholar 

  16. Julius, A. J. 2003. Basic Structure and the Value of Equality. Philosophy and Public Affairs 31(4): 321–355.

    Article  Google Scholar 

  17. Korsgaard, C. 2010. The Constitution of Agency: Essays on Practical Reason and Moral Psychology. Oxford: Oxford University Press.

    Google Scholar 

  18. Lippert-Rasmussen, K. 2008. Inequality, Incentives and the Interpersonal Test. Ratio 21(4): 421–439.

    Article  Google Scholar 

  19. Melenovsky, C. M. 2013. The Basic Structure as a System of Social Practices. Social Theory and Practice 39(4): 599–624.

    Article  Google Scholar 

  20. Murphy, L. 1999. Institutions and the Demands of Justice. Philosophy and Public Affairs 27: 251–291.

    Article  Google Scholar 

  21. Neufeld, B., and C. Van Schoelandt. 2014. Political Liberalism, Ethos Justice, and Gender Equality. Law and Philosophy 33(1): 75–104.

    Article  Google Scholar 

  22. Olsaretti, S. 2012. The Inseparability of the Personal and the Political: Review of G. A. Cohen’s Rescuing Justice and Equality. Analysis 72(1): 145–156.

    Article  Google Scholar 

  23. Pogge, T. 2000. On the Site of Distributive Justice: Reflections on Cohen and Murphy. Philosophy and Public Affairs 29(2): 137–169.

    Article  Google Scholar 

  24. Rawls, J. 1971. A Theory of Justice. Cambridge, MA: Harvard University Press.

    Google Scholar 

  25. Rawls, J. 1993. Political Liberalism, Expanded edn. New York, NY: Columbia University Press.

    Google Scholar 

  26. Rawls, J. 1999. Kantian constructivism in moral theory. In: Collected Papers, ed. S. Freeman. Cambridge, MA: Harvard University Press.

  27. Rawls, J. 2001. In: Justice as Fairness: A Restatement, ed. E. Kelly. Cambridge, MA: Harvard University Press.

  28. Rawls, J. 2008. Lectures on Marx. In: Lectures on the History of Political Philosophy, ed. S. Freeman. Cambridge, MA: Harvard University Press.

  29. Ronzoni, M. 2008. What Makes a Basic Structure Just? Res Publica: A Journal of Moral, Legal and Social Philosophy 14(3): 203–218.

    Article  Google Scholar 

  30. Scheffler, S. 2006. Is the Basic Structure Basic? The Egalitarian Conscience: Essays in Honor of G. A. Cohen. Christine Sypnowich, 102–129. Oxford: Oxford University Press.

    Google Scholar 

  31. Shiffrin, S. 2010. Incentives, Motives, and Talents. Philosophy and Public Affairs 38(2): 111–142.

    Article  Google Scholar 

  32. Tan, K.-C. 2012. Justice, Institutions, and Luck: The Site, Ground, and Scope of Equality. Oxford: Oxford University Press.

    Google Scholar 

  33. Thomas, A. 2011. Cohen’s Critique of Rawls: A Double Counting Objection. Mind 120(480): 1099–1141.

    Article  Google Scholar 

  34. Titelbaum, M. G. 2008. What Would a Rawlsian Ethos of Justice Look Like? Philosophy and Public Affairs 36(3): 289–322.

    Article  Google Scholar 

  35. Weithman, P. 2013. Why Political Liberalism? On John Rawls’s Political Turn. Oxford: Oxford University Press.

    Google Scholar 

  36. Williams, A. 1998. Incentives, Inequality, and Publicity. Philosophy and Public Affairs 27(3): 225–247.

    Article  Google Scholar 

  37. Williams, A. 2008. Justice, Incentives and Constructivism. Ratio 21(4): 476–493.

    Article  Google Scholar 

Download references


I am grateful to Cynthia Stark, Chrisoula Andreou, Leslie Francis, Erin Beeghly, Darrel Moellendorf, and two anonymous reviewers for very helpful comments on earlier drafts of this paper.

Author information



Corresponding author

Correspondence to Edward Andrew Greetis.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Greetis, E.A. Rescuing Rawls’s Institutionalism and Incentives Inequality. Res Publica 25, 571–590 (2019).

Download citation


  • John Rawls
  • G. A. Cohen
  • Difference principle
  • Incentives inequality
  • Institutionalism
  • Distributive justice
  • Constructivism