Philosophers have a habit of approaching political problems in a certain way. They start by imagining a perfectly just world, then suggest ways of making our imperfectly just world as similar to this perfectly just world as possible. Such a philosopher might say ‘In a perfectly just society protected characteristics like race would have no bearing on whether a person is selected to attend a prestigious university; therefore, in our imperfectly just world policies which aim to remove discrimination in the selection process are justified.’ This sort of reasoning seems appealing at first glance, but there are reasons to doubt its cogency.
John Rawls is one of the most famous philosophers known for this form of reasoning. While he is hailed as a champion of left-wing liberal political theory, Rawls’ distinction between ideal and nonideal theory has come under fire from thinkers in the Black radical tradition. In A Theory of Justice (1971), Rawls defines ideal theory as the study of principles of justice that regulate a ‘well-ordered society’ under conditions of ‘strict compliance’ and nonideal theory as the study of principles that regulate injustices under conditions of ‘partial compliance.’ He claims that it is necessary to begin with ideal theory because it provides ‘the only basis for the systematic grasp of these more pressing problems’ and that a ‘deeper understanding [of nonideal theory] can be gained in no other way’ than by first envisaging the nature and aims of a perfectly just society. In other words, without ideal theory as a reference point, it is impossible to determine what the just course of action is under nonideal circumstances.
In the early 2000s, a fierce debate ensued between Tommie Shelby and Charles Mills over whether this approach to political philosophy—let’s call it the Rawlsian method—is an effective way of understanding and combatting racial injustice.
According to Shelby, nonideal theory is logically dependent on ideal theory because the latter serves as an essential guide or benchmark for the former. Shelby states that Rawls considers injustices to be deviations from ideal principles of justice; this means that charges of injustice presuppose ideal theory because it offers a necessary and reasonable standard for judging when social arrangements are oppressive and sets a goal to strive towards when resisting oppression. Furthermore, Shelby interprets Rawls’s fair equality of opportunity (FEO) principle as having a built-in corrective component—if the FEO principle was institutionally realized, it would effectively remove many of the socioeconomic burdens that racial minorities suffer. Thus, on Shelby’s reading, Rawlsian theory is capable of helping us understand and combat racial injustice without modification.
In contrast, Mills argues that nonideal theory is separable from ideal theory, and the former can be conceived of without the latter. Mills considers ideal theory to be ideological in the pejorative sense—it is a group of ideas that reflect and contribute to perpetuating the illicit group privilege of upper-middle-class white men, who are massively overrepresented in professional philosophy. Mills contends that ideal theory abstracts away from the realities of historical injustices and rejects the claim that Rawls’ FEO principle has a built-in corrective component. Instead, Mills suggests that ‘the best way to bring about the ideal is to recognize the nonideal’, and that issues of racial justice should start from or be framed in terms of nonideal theory because this approach is better able to attain political ideals, by ‘realistically recognizing the obstacles to their acceptance and implementation.’ Thus, on Mills’ reading, Rawlsian theory is only capable of helping us understand and combat racial injustice with modification.
This disagreement between Shelby and Mills centers around their confidence in the Rawlsian method. While Shelby argues the Rawlsian method can usefully account for questions of racial justice, Mills criticizes it, arguing that nonideal theory can and should be separated from ideal theory to address questions of racial injustice because the latter reinforces ideological hegemony. So, which of these readings should be favored?
The general theory of the second best was originally presented as an economics theory but has a significant implication for the Rawlsian method. It states that ‘if one of the Paretian optimum conditions cannot be fulfilled a second-best optimum situation is achieved only by departing from all other optimum conditions.’ Applied to Rawls, the general theory of the second best suggests that when all our ideals cannot be realized simultaneously, we should be cautious about naively assuming that the best solution is to simply realize as many ideals as possible. We ought to instead contemplate making systematic revisions that depart from most, if not all of these ideals. Put simply, if the first option (ideal theory) is unattainable, then the second-best option (nonideal theory) may only share a few, if any, characteristics with the first. If my first best option for dessert is a chocolate chip cookie, for example, it does not follow logically that my second-best option is a chocolate chip; my second-best option might look completely different, like a slice of cheesecake.
The general theory of the second-best favors Mills’ reading of Rawls because it illustrates that nonideal theory is separable and not dependent on ideal theory. The first-best option should not necessarily serve as a reference point or benchmark for the second-best option. This implies that Shelby’s defense of the Rawlsian method is fallacious—it is not necessary to have a clearly defined ideal in order to determine the nonideal course of action.
To illustrate how this point is relevant to racial injustice, consider the US Supreme Court’s recent decision to ban the use of affirmative action in college admissions, on the grounds that they violate the Fourteenth Amendment’s ‘equal protection clause.’ The equal protection clause is an example of ideal theory—it is a legal principle that assumes strict compliance and requires that applicants have ‘equal protection of the laws’ in the candidate selection process. Affirmative action policies are incompatible with the equal protection clause because they require that (racial) minorities be given preference or special consideration in the candidate selection process. Affirmative action policies therefore fall under the remit of nonideal theory because they involve governing historical injustices under conditions of partial compliance, which result in (racial) minorities not having a fair and equal opportunity in the selection process. In economic terms, the equal protection clause is the first-best policy option and affirmative action policies are the second-best policy option—but is the equal protection clause a necessary reference point to determine whether affirmative action policies are justified?
Shelby’s reading of Rawls implies that the equal protection clause is a necessary reference point to determine whether affirmative action practices are justified. Mills’ reading of Rawls implies that the equal protection clause is not a necessary reference point to determine whether affirmative action practices are justified. The general theory of the second best gives us reason to favor Mills’ reading of Rawls over Shelby’s because it indicates that the justifiability of the affirmative action policies can be determined without recourse to the equal protection clause. In other words, affirmative action, which is the second-best option, or nonideal theory, is separable from and not dependent on the equal protection clause, which is the first-best option, or ideal theory.
But if the equal protection clause is not a necessary reference point for determining the justifiability of affirmative action, how then should we determine whether affirmative action is justified? Amartya Sen points out that determining the second-best option involves taking a comparative approach which only requires that we choose the best amongst the feasible set of options available and does not require having an identifiably perfect alternative. Affirmative action policies are justified when they are the best available policy amongst a feasible set of alternatives, even if they differ more from the equal protection clause than these alternatives.
The justifiability of a nonideal policy can and should be determined by comparing it with the feasible set of policy options available, regardless of whether it shares few or no characteristics with the ideal policy. The Supreme Court’s recent decision ultimately ignores this comparative justification for affirmative action. Interestingly, there seems to be a covert appeal to the Rawlsian method in the Supreme Court’s decision to ban the use of affirmative action in college admissions. The Court justified their decision on the grounds that affirmative action policies are not justifiable as the second-best way of ensuring that all candidates have a fair and equal opportunity in college admissions because they differ too much from the first-best option, which involves administering the equal protection clause uniformly amongst candidates. Some commentators advocate for alternative admissions strategies, such as targeting recruitment of underrepresented candidates and using community college transfers as a path for boosting racial, ethnic, and economic diversity on campuses. These alternative strategies would be the second-best option if we deploy the Rawlsian method. This is because targeting recruitment of underrepresented groups and promoting college transfers are more similar to the equal protection clause than affirmative action, in that they do not require selectors to show any preference in the candidate selection process. Instead, these alternative strategies aim to promote diversity by improving the accessibility of pathways into college.
The general theory of the second best gives us reason to doubt whether these alternative strategies are more justified than affirmative action. It implies that the degree of similarity between the equal protection clause and both affirmative action and these alternative strategies has no moral relevance on which is more justified. In other words, the fact that these alternative strategies are more similar to the equal protection clause than affirmative action is morally irrelevant to determining which of the two is more justifiable. The justifiability of affirmative action should be determined by comparing it directly with these alternative strategies (and other feasible alternatives), and the similarity between these alternatives and the equal protection clause should not be used as a justification for choosing one alternative over the others.
Another way of understanding this problem is in terms of framing. The Rawlsian method frames political problems in a way that abstracts away from the nonideal historical and socio-economic realities that structure society. If we, as Mills suggests, frame political problems in terms of nonideal theory, there is more reason to believe that the use of affirmative action in college admission is justified. Even though affirmative action is very different from the equal protection clause and would not be justified (or necessary) under ideal circumstances, under nonideal, real-world circumstances, affirmative action can be justified on the grounds that it is the best policy amongst the set of viable alternatives. Tackling racial injustice requires taking a pragmatic approach to political problems, instead of naively hoping that we will one day live in a world where there is strict compliance with the principles of justice.
Hamza King
Hamza King is the founder and editor-in-chief at Phlexible Philosophy. He recently completed a MSc in Philosophy & Public Policy at the London School of Economics and Political Science, writing his dissertation on the ethics of climate displacement.
Hamza is also interested in business and organizational ethics, the postliberal movement, and the philosophy of technology.
