Does systemic injustice justify Robin Hood strategies?
Many injustices arise because of patterns of behaviour, single instances of which seem harmless or at least pardonable. For example, if professors help the kids of their friend get access to university programs – and given the fact that professors and their friends tend to come from the same socio-economic background – this can lead to structural discrimination against applicants from other backgrounds (as discussed by Bazerman and Tenbrunsel here, p. 38-40). Other examples concern implicit biases against women and ethnic minorities. Much work has been done recently that helps us to understand how these mechanisms work (see e.g. here). Given how pervasive these mechanisms are, it is understandable that they cause moral outrage. The question is, however, what individuals should do in reaction to them.
Imagine that you are in a situation in which you have some amount of power, for example as a reviewer or as a member of a search committee. You might be tempted to use a “Robin Hood strategy”, i.e. a strategy that breaks the existing rules, for the sake of supporting those who are treated unjustly by these rules. Given how structural injustices work, many such “rules” are not formal rules, but rather informal patterns of behaviour. But it is still possible to work against them. For example, could it be justified to reject male applicants not because of the quality of their applications, but because they are white and male and come from a rich Western country?
One has to distinguish two levels of what such a strategy could imply. The first concerns correcting own biases that one might have, despite all good intentions (to check them, the various tests offered by Harvard University on this website can be helpful). The best way to do this, if possible, seems to be anonymity. When this is not feasible, the alternative is to scrutinize one’s patterns of thought and behaviour as best one can. The more power one has, the more it seems a requirement of justice to do this.
This is different from a second level of Robin Hood strategies, for which the name seems more appropriate: these concern not only own biases, but biases of the system. The idea is to work against them on one’s own, in one’s little corner, maybe hoping that if enough of us do this, the problems can be solved or at least attenuated. Could this be a defensible strategy?
The problem is, of course, that one risks introducing new injustices. One consciously deviates from what are supposed to be the criteria of selection, for example a candidate’s performance in previous jobs or the likelihood of being a good team member. In some cases, however, it is reasonable to assume that if a candidate comes from a group that suffers from discrimination, achieving the same level of merit as a candidate from another group takes much more effort. So according to this argument, and as long as these problems are not recognized by the official selection criteria, it seems defensible to privately factor in these previous structural inequalities.
But one’s epistemic position in judging such cases is often a weak one. For example, standard application material for many jobs includes a CV and some letters of reference. These materials are often insufficient for understanding the details of a specific case and the degree to which discrimination or stigmatization might have had an impact on the candidate’s previous career. One risks making mistakes and importing one’s own subjective biases and prejudices; taken together, this can make things worse, all things considered.
Robin Hood strategies do not provide what seems most needed: good procedures and public accountability. They do not get at the root of the problem, which is to create collective awareness of the issues, and to find collective mechanisms for addressing them (the gendered conference campaign is an example). Collective mechanisms are not only likely to be more effective, they also bring things out into the open, and create a public discourse on them. Although public discourses also have their weaknesses, there is at least a chance that the better argument will win, and there are opportunities for correcting what end up misguided strategies. Robin Hood strategies, in contrast, fight fire with fire: they remain within a logic of power, trying to find ways in which one can use counter-power to subvert the dominant power elites. But this does not change the fundamental logic of the game.
Thus, our preferred strategies should be different ones: strategies that really change the logic of the game, openly addressing problematic patterns of behaviour and looking for collective – and maybe formally institutionalized – solutions. Nonetheless, and despite all the reasons mentioned above, I cannot bring myself to thinking that Robin Hood strategies can never be justified in today’s world. Of course one has to be very careful with them, not only with particular cases, but also with regard to the slippery slope one might get onto. But are they ruled out completely? What do you think?
Thanks for this, Lisa. The post raises some interesting points about the relationship between unjust procedures and legitimate authority, and also about what we ought to do about this.
I want to begin by mentioning a small caveat. In the post, you move very quickly from the idea the procedures are unjust to the conclusion that they lack legitimate authority. However, under at least some conditions, this is not obvious to me. On at least some conceptions of authority, there may remain weighty principled reasons to obey the procedure even though it is unjust. (Perhaps your post is consistent with this, and if so then there is no disagreement.) Having said this, I acknowledge that this looks less plausible when the procedure is *very* unjust, which is often how it is.
Second, could you please explain in a little more detail how you understand the second level of Robin Hood strategies? You mention that it is concerned with the 'biases of the system' but this seems like a huge category as systems will be biased in lots of very different ways. Consider a system that biased (i) those who are good at performing the job because of natural talent; (ii) those are good at performing the job because of the education that they unjustly received; (iii) those are not good at performing the job. Which kind of case are you (principally) concerned with?
Hi Tom, thanks for these comments. On your first point: I am uncertain how the concept of legitimate authority would apply to systemic patterns of behavior that are not formally institutionalized (for example biases against minorities). This is what I was mostly thinking about. Your point might still apply in cases in which such informal phenomena occur in combination with formal institutions, and there might be questions about the latter's legitimate authority.
You are right that the second category is rather broad. I was mainly concerned with your second point: advantages that accrue to some individuals, but not others, because of biases, stereotype threats, etc. Such advantages can even accrue to individuals who fall under your third category, i.e. they are not actually good at performing the job, but they receive (premature) praise.
That's helpful, Lisa. Thanks.
On the first point: We could know that, given our biases, the current procedure for selecting applicants is unjust – that is, it fails always to select the applicant that is in fact entitled to the job. But, it doesn't always follow from this that we don't have principled reasons to obey the result it produces. Perhaps procedures can be unjust but still command legitimate authority. (I have in mind the kind of position discussed by Rawls in his reply to Habermas.)
I'm afraid that I'm still not sure on the second point. (It could well be that I'm just very dense this morning!) Perhaps I am confused by the phrase 'biases of the system'. Is the thought that I should be concerned with neutralising the effects of not only my own biases, but also the biases of others on the selection committee?
Hi Tom, what's at stake is in a way the tension between your two paragraphs. Formal procedures (for example a rule along the lines of "only pay attention to merit, nothing else") has, on the face of it, a certain legitimacy – but often we can know that in combination with certain patterns of behavior it will result in injustices. As to the second paragraph: correcting the biases of others on a specific selection committee would be one thing; one might even be tempted to try to correct (at least a bit) for all the biases that candidates have experienced in their previous careers, by various members of the profession in question.
To put it into a concrete example: as a member of a search committee, there are several short-listed candidates. They have very different backgrounds, so you have reasons to assume that they have been favored and disfavored in very different ways in the past, because of systemic biases. The first thing to do would be to try to get rid of one's own biases. Then you might try to discuss the problem of biases with the other members of the selection committee. But let's assume they are not very open to it. You might then use other means – for example, rhetoric, bargaining with other committee members, or the leaking of additional information – in order to influence the decision making process. But such measures are exactly what is problematic in many decision making processes – so should you use these problematic means in order to counteract systematic injustices in the system as a whole?
Thanks for the post, Lisa. Could I add a little to the discussion on Tom’s first point here. As I understood it (correct me if I am wrong, Tom), the thought was that one distinction we might want to draw would be between injustices that are within and beyond reasonable limits. The distinction might be relevant because it is sometimes thought that the different kinds of injustice warrant different types of response – for example, these different cases might help distinguish when one would be justified in engaging in protests and demonstrations and when one would be justified in engaging in civil disobedience or revolutions. It also appears that it might be quite relevant to your case. We might, for example, say that pushing for public dialogue is most appropriate if the injustice is within reasonable limits, but that Robin Hood strategies (as their namesake might suggest) are appropriate in cases of more extreme injustices. Perhaps trying to correct for biases in the system (your second type) looks most plausible when the injustice is extremely glaring (e.g., if different genders are treated legally equally, but social patriarchy is so strong that women can never reach any positions of import), perhaps even because it is one of few means to get dialogue moving. Whereas, if there is objectionable patriarchy, but not of this level, perhaps working through public dialogue if more appropriate. That seems as though it might have relevance to the case. I wonder whether you could say a bit more about where your concerns lie on this spectrum of injustice and whether you think the different scenarios would alter the argument?
Andrew, this is what I wanted to hint at in the last paragraph, so thanks for the opportunity to expand on this. I fully agree that the degree of injustice plays an important role in evaluating strategies. The question is: do we think that the injustices caused by systemic patterns of behavior (within systems that are formally non-discriminatory) can justify such action? I.e. can *informal* (or maybe I should say: informally caused) forms of injustice be a reason for breaking formal rules? As I suggest, normally we should prefer other strategies, but I wonder whether some cases might be grave enough to go for other ones? I don't have a clear-cut rule, and I think a lot has to be decided by case-to-case judgments.
Another factor that weighs in is the degree of epistemic certainty that we have for understanding these mechanisms and seeing that someone was discriminated against. Maybe one should think about this as a two factors that both need to obtain (think of their multiplication, maybe): a high degree of injustice, and a high degree of certainty. Taken together, they might lead beyond the threshold that justifies Robin Hood strategies. Would you agree with this?
Thanks for your post Lisa and for forcing me to confront my own unreflective appreciation for 'Robin Hood' strategies. While clearly collective transparent structural mechanisms to address such inequalities and injustices are preferred these are often impossible if those who benefit from them remain in power. I was wondering if you might classify quotas as a type of Robin hood strategy? If so, couldn't we argue that quotas are justifiable temporal mechanisms that are necessary to change an unjust system which refuses to correct itself?
Lisa, thanks for the further thoughts. I think I am with you on some of the relevant considerations here – epistemic/publicity constraints and level of injustice seem important, as does the case you make in the text about personal biases in attempts to correct (similar, I guess, to Locke’s thought about the rationale for centralised institutions). I am a little unsure about the distinction between formal (rules) and informal (patterns of behaviour) injustices. I guess I might tend to talk about ‘institutions’ that have certain kinds of regulating control over people’s lives. These can be formal (e.g., the law) or informal (e.g., widely held views about appropriate behaviour or common biases of the kind you mention), but I am not sure I see an independent worry about trade-offs (i.e., problems internal to that distinction with violating the former to correct injustices in the latter). If I understood you correctly, the crucial thing about Robin Hood strategies is that they are attempts to work through isolated instances of action, the major problems with which are that isolated instances of action do not score so well in terms of some of the factors mentioned above (they are less likely to produce long-term change, avoid the public process, and so forth), although we may want to relax these standards if injustice was so grave that, e.g., we should simply address it however possible. (So, counterfactually, if Robin Hood strategies did score higher in terms of epistemic/publicity conditions and so forth, they would be preferable to the alternatives; it would not matter that they were attempting to correct informal injustices by breaking formal rules.) If that reading is correct, then I think we are in broad agreement here. But if I have missed something further about the formal/informal distinction, please do correct me.
I'm not sure quotas would come undo Lisa's definition since they are public rules. Though of course if you applied quotas privately in your admission choices they would a Robin Hood strategy.
Andrew, you are right that the question about formal / informal is at the core of things here. What you call the "independent worry about trade-offs" is as follows, in my eyes: in the situations I have in mind, the formal rules are such that we could endorse them (e.g. calling for objectivity, neutrality, etc.). It is the combination of these formal rules – which are usually quite vague – and informal patterns of behavior that creates the problem. So if we (privately and secretly) favor candidates disadvantaged by the informal patterns, we are actually acting against (formal and public) rules which, if interpreted correctly and not combined with systemic biases etc., we could endorse. To use the Robin Hood analogy again (although I'm not sure this is historically accurate): property rights could be a good thing if they were distributed reasonably evenly; but in the current situation, the distribution of property is so unequal that some individuals really lose out, so I violate rules that would otherwise be legitimate in order to correct the current injustice.
One fear one might have, and which could be a reason why the informal/formal distinction matters, is that by using Robin Hood strategies, one might undermine what is otherwise a legitimate system (or maybe the best system we have). This worry becomes less relevant, however, if we think that the current system lacks legitimacy anyway, and that it is a mere smokescreen for the power plays that go on behind the scene. This may well be the case in certain situations. But I tend to think that on the whole, we should try to maintain and strengthen legitimate formal institutions (e.g. selection procedures). So there is a question about how to weigh the injustices of specific cases against the value of having a formal system which, while certainly imperfect, may still be better than a sort of institutionalized war fare.
Anya and Bruno, yes, public quotas wouldn't fall under my definition, private quotas would (and there may be an intermediate area when you try to convince your fellow committee to introduce quotas). I agree with Anya's point that quotas can be a justified as a necessary means in order to change an unjust system (although I think that if they are introduced top down, and those who have to enforce them do not see their point, their effectiveness can be limited – so one still has to (try to) do the work of convincing others). But if we don't have public quotas, we might indeed think about using the little power we have, as individuals, in order to push for a certain gender balance. Sometimes this may even be possible without violating any formal rules. The question I am interested in is whether we may also be justified in doing it if we violate rules (see also my reply to Andrew above – I assume that the rules as such would be legitimate, and that proceduralism has a certain value).
Lisa, thanks for the further comment and clarification. I think I understand a little better now. I suspect that where we are differing here is a difference we have found previously about assessing frameworks. To wit, I am generally more disposed to think about systems rather than modules – how institutions combine rather than what they do individually. So, I am less sure we can separate our analysis of a legal regime from the informal regime with which it combines. In that respect, I suspect I do not see quite the same dilemma. But I think we are on similar lines about the conclusion anyway, so perhaps no matter.
Andrew, insofar as we might disagree here, it could also be a question of ideal vs. non-ideal – in the sense that the circumstances under which these questions become salient are (often extremely) non-ideal, but also in the sense that in a non-ideal world, the importance of proceduralism might in fact be greater than in ideal circumstances.
I must confess that I have become quite confused by the introduction of the ideal of publicity to the definition/account of Robin Hood strategies. Lisa, perhaps it would help me if you were to offer a canonical definition of what you understand by Robin Hood strategies.
Second, and more generally, I'm interested by Lisa's reply to one of Andrew's early comments. You suggest that proceduralism may have greater importance in an unjust world. I'd be really interested to hear more about why you think. In contrast to your position, you might think (as Rawls indeed does!) that living in an unjust society might corrupt the legitimacy of certain procedures, and so diminish the legitimacy of these procedures.
Hi Tom, not sure I have a canonical definition yet, but here are the key ingredients: a) it's done in order to correct a grave injustice, b) it breaks a formal rule although that rule would be fine in itself, but it is a rule which, together with widespread biases in the system, contributes to the injustice, c) it thereby foregoes the official channels of complaint, using, instead, discretionary room that one happens to have. c) has to do with publicity, although one might argue that there could also be public Robin Hood strategies. These might have some similarities with acts of civil disobedience, and presumably the main aim would be to draw attention to the injustice. Whereas what I see as an important feature of Robin Hood strategies is that they directly aim at correcting (one instance of) a systemic injustice. Often, pursuing them secretly may be the only opportunity for pursuing them at all, because otherwise there would be sanctions.
On your second point: I don't think there is a contradiction here; if procedures are already corrupted, there might be no point in sticking to them. But if they still work reasonably well, we might have reasons to want to preserve them. The reason that I said they might be more important in an non-ideal than in an ideal situation is that in an ideal situation, one might be able to decide about each case individually, and one would maybe have enough time to discuss until one arrives at an agreement, and all participants to the procedure would be willing to follow good reasons rather than just push through their own interests. In a non-ideal situation, fair procedures might be the lowest common denominator one can agree on, and the only way to prevent even more unfair outcomes. That's all I meant, and that's compatible with some procedures being so corrupted that they do not possess legitimacy any more.
Thanks for the post Lisa. Perhaps looking at an example where the epistemic problems of judgement is absent can help clarify the issue. Take the case of a teacher who knows her students very well. She is asked to write recommendation letters by two students applying for the same position. One of the students belongs to a group against whom there is an implicit bias. Let's say she's a female. The teacher knows that her recommendation will have significant influence on who takes the job. The question then is, would it be justifiable or morally permissible for the teacher to write a stronger recommendation letter for the female student and for instance put her in contact with key people at the department she's applying to but hide this information from the male student. This, I guess, would be a typical Robin Hood strategy.
I think Andrew is right that such instances might not do much to change the system which is the most desirable option (although one could argue that the more women you have in a department the more chance that they bring about change because they act from within). So yes under ideal circumstances; that is circumstances where institutional reform is possible; we should direct all efforts to rendering institutions more just (both their formal and informal mechanisms) . I wonder, however, whether under non-ideal circumstances individuals who are faced with the impossibility of affecting direct institutional changes might be justified in taking some form of retaliatory measures even if the likelihood that these measures end up changing the system is not very high.
Hi Siba, thanks for this example, it is spot on. Assuming that the teacher does not actively discriminate against the male applicant (this might depend on what „hiding information“ amounts to), it would be a relatively morally unproblematic kind: giving extra support to those who are in a weaker position as a result to systemic biases. It could be compared to a sort of private affirmative action. The line between this and actively undermining the positions of members of stronger groups can be a fine one in practice, I assume. Justifying the latter raises larger questions.
I find the term „retaliatory measures“ interesting here. I am not sure I would want to argue along these lines, because the problem there is that we are talking about groups, not individuals, and about structural injustices, not individual wrong-doing. One might think about cases in which the model of retaliation fits well (e.g. because the male student, in your example, took advantage of his privileged position in a situation in which he should have supported his class mate, e.g. by introducing her to an important person in the field), but because what I was mostly focussing on were *systemic* injustices, this wasn’t what I had in mind. I might be more optimistic than Andrew, in fact, about the possibility of slowing changing the system by systematically supporting disadvantaged groups, so I could run with a consequentialist justification.