Desert, Responsibility and Luck Egalitarianism

Abstract

Desert and responsibility are key concepts in political philosophy, most notably in discussions on justice. It is just that people get what they deserve, and what they deserve seems to have something to do with what they are responsible for. This tenet is as close to a fundamental constant as one can get in practical philosophy, so that even some egalitarians, luck egalitarians, make room for exceptions dictated by it: only differences people are not responsible for should be equalized, differences people are responsible for are not unjust, because they are deserved. In this paper I shall contest the second part of this tenet that what people deserve is somehow linked to what they are responsible for. To this end, I shall give a detailed account of the concept of desert in the first half of this paper. In the second half, I shall consider the implications of this for luck egalitarianism, and conclude that while luck egalitarianians can counter some criticisms that are grounded on a wrong understanding of the concept of desert, they cannot rest content in relying on the purely formal notions of responsibility and desert, but need to provide substantial arguments to support their conclusions.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

8.1 Desert and Responsibility

Regarding desert, astonishingly many people think that only persons can deserve something, and only in virtue of actions for which those persons are responsible. This is astonishing, to my mind, as in everyday language all sorts of things are said to be deserving in virtue of all sorts of properties. For example, we do say that criminals deserve punishment, and that good pupils deserve to get good marks, but we also say that the most beautiful contestant deserves first prize at the beauty contest, that great paintings deserve to be admired, and that unique landscapes deserve to be preserved.

In this paper I shall examine the conceptual connection between desert and responsibility, and I shall argue that there is none. That is, I shall argue that those who claim that it is a conceptual truth that only persons can deserve something solely for actions for which they are responsible are mistaken. I have absolutely no idea whether or not the claim that only responsible action can ground desert is true substantially. Personally, I am inclined to doubt it, but that is neither here nor there, as that is not the question I will pursue in this paper. I shall not argue for the truth or falsehood of any given desert judgment. My aim is just to clear up a conceptual confusion.

Since it is the concept of desert I aim to elucidate, and not the concept of responsibility, I shall have a lot to say about the former and rather little about the latter. I will take a lot for granted as far as responsibility is concerned although I realize that things are far from clear in that regard. Most importantly, I will take for granted that we all know what it is to be responsible for an action. Basically, this is already all I have to say about responsibility.

On to desert then.

8.1.1 Desert: The Basics

The first thing to be said about desert is that it is a normative notion, that is, the fact that somebody deserves something implies that she ought to get it, although not unconditionally, but only pro tanto.¹ So, the fact that somebody deserves something is always a reason for giving it to her, although not always a conclusive reason. Something more important could count against it. The best team deserve to get the cup, but if the best team do not actually win, they ought not to get it.

The next thing to be said is that desert is a three-place relation, “x deserves y in virtue of z”. Let us call x the desert subject, y the desert object, and z the desert base.² There are many desert judgments that do not explicitly state a desert base.

Sometimes, we simply say, for example: “The team deserve the cup.” This is elliptical, though. One cannot deserve something for no reason at all, but only in virtue of something, the desert base, as Joel Feinberg has shown in his seminal paper “Justice and personal desert”, giving the first and most influential analysis of the concept of desert. Moreover, the desert base must be attributable to the desert subject.³

Hence, desert judgments like “The team deserve the cup.” need to be understood as abbreviated and as always implying a desert base which needs to be attributable to the desert subject, the team in this case. So, we have to supplement the statement with a desert base, for instance like this: “The team deserve the cup because they have played so well.”, because having played well is attributable to the team, whereas we cannot supplement the statement like this: “The team deserve the cup because of water’s boiling point.”, because water’s boiling point is not an attribute of the team’s.

Here the trouble begins, because it is not quite clear what is “attributable” to a desert subject. Me, I do not mean to say anything more with this than that the desert base has to be an attribute of the desert subject’s, be it an action or a property. It just has to be theirs, nothing more, nothing less.⁴ However, there are others who think that only actions for which the desert subject is responsible are “attributable” to them in the requisite sense.⁵ This also means, of course, that only persons can be desert subjects, because other sorts of things cannot perform such actions.

This is the view I shall contest in this part of the paper. I will call it the responsibility view. Before doing so, however, let me explain why this view is so curiously widespread, even though the concept of desert clearly works differently in everyday language.

8.1.2 Feinberg and Rawls

I think the seed for the responsibility view was already laid in Feinberg’s analysis of desert. In it, Feinberg is concerned with a particular sort of desert objects: prizes, grades, rewards and punishment, praise and blame, compensations, in short: he is concerned with certain forms of treatment as desert objects only. And the only desert subjects he is concerned with are persons.⁶ He exclusively looks at persons who receive these desert objects of prizes, grades, rewards, etc.

Concentrating as he does on persons as desert subjects suggests that only actions for which the desert subjects are responsible can be desert bases, because if only persons can be desert subjects this must be due to something that is peculiar to persons, something like actions for which they are responsible which no other animal or object can lay a claim on.

This is corroborated by the particular set of desert objects Feinberg considers. Only persons deserve these sorts of things, and only in virtue of actions for which they are responsible. It just would be nonsensical to assign punishment, say, to any other sort of desert subject than persons, and it would be unfair to do so on any other basis than an action for which they are responsible.

After having sprung from this seed, the responsibility view throve and prospered further with the publication of Rawls’s A theory of justice a few years after Feinberg’s paper which enormously influenced the subsequent treatment of the concept of desert in philosophical discussion.⁷ Rawls as well considers only actions for which persons are responsible as desert bases. As he says in A theory of justice:

The precept which seems intuitively to come closest to rewarding moral desert is that of distribution according to effort, or perhaps better, conscientious effort.⁸

Evidently, a conscientious effort consists in an action for which a person is responsible.

So, it seems as if Feinberg and Rawls subscribe to the responsibility view, and, influential as their theories have been, as if Feinberg and Rawls really are responsible for the responsibility view being so widespread. I think the latter of these claims is true. I do think that Feinberg’s and Rawls’s analyses of desert are responsible for many people holding the responsibility view.

The former claim, however, is false. Neither Feinberg nor Rawls really held the responsibility view. It is true that they limit their discussion of desert to cases in which persons deserve something in virtue of actions for which they are responsible, but this is due to the fact that they are concerned with examining the particular connections between desert and justice. Still, this restriction does not mean that other things than persons cannot be desert subjects, nor that other things than responsible actions cannot be desert bases. Neither Feinberg nor Rawls excludes the possibility of a painting’s deserving admiration. The thing is just that the question whether or not a painting deserves admiration is not a question pertaining to justice.⁹

8.1.3 Against the Responsibility View

Even so, many philosophers hold the responsibility view. So, what is wrong with it?

What is wrong with it is that it puts the cart before the horse. Instead of examining the concept of desert and then deriving from that which individuals may permissibly replace the variables x, y, and z, they try to derive the concept from the individuals which they already know must be the only permissible ones. This is methodologically unsound, and it begs the question whether those individuals really are conceptually the only permissible ones.

Consider, by way of analogy, the predicate “x is a flightless bird”, and consider a zoologist, a newbie on the field of ornithology, who is convinced that penguins are the only flightless birds there are. Now one day our zoologist is confronted with an ostrich. There are two ways of reacting open to our newbie-zoologist: first, he can correct his belief and say that penguins evidently are not the only flightless birds there are; or second, he can flatly refuse to recognize an ostrich as a flightless bird and qualify it as something else instead on the grounds that, obviously, it is not a penguin. Clearly, the first way of reacting is the adequate one. Unfortunately, regarding desert, the proponents of the responsibility view take on the equivalent of the second way.

As Wojciech Sadurski, one of the many champions of the responsibility view, puts it representatively:

When we are pronouncing judgments of desert, we are inevitably making judgments about persons whom we hold responsible for their actions. It makes no sense to attribute desert… to persons for actions or facts over which they have no control.¹⁰

This is rather a curious statement about desert judgments. What exactly does Sadurski mean by saying that in talking about desert we “inevitably” talk about persons responsible for their actions, and that talking about other things in connection with desert “makes no sense”? It is obviously not inevitable in the sense that we cannot but, or do not, talk about desert in any other way, because we do so all the time. Let us say, for example, that a man may deserve something good in virtue of his noble birth, as the Ancient Greeks believed. In what sense does it “make no sense” to say this?

Sadurski leaves it at that and does not elaborate what he means. More is the pity, since it needs elaborating, because clearly, that judgment does not “make no sense” such that we could not possibly understand what it means. We do. We might be inclined to disagree, but we understand it alright. There are several other possibilities, though: Sadurski might mean that it is analytically false to say this, or that there is some kind of Strawsonian truth-value gap, or a category mistake, or quite literally, that this sentence is meaningless.

However, as of yet, we have not been offered a concept of desert such that the very meaning of the word excludes such desert judgments or makes them meaningless, or which implies a truth-value gap or a category mistake in such cases. Hence, to say that the desert judgment that a man deserves something good in virtue of his noble birth is analytically false, or meaningless, or that there is a truth-value gap regarding this judgment, or that there is a category mistake, would be to beg the question in a zoologist-reaction-number-2-kind of way as just described.

The desert judgment in question may very well be false, and moreover, it may very well be that only those desert judgments are true that have persons as desert subjects and actions for which they are responsible as desert bases. My point here is purely conceptual. These desert judgments may be false, but not analytically so, and in no sense are they meaningless. If they are false, they are not so because the concept of desert does not allow for these kinds of desert judgments. It does, as I shall go on to show. If they are false indeed, they must be so because of substantial reasons rather than conceptual ones.¹¹

The proponents of the responsibility view might mean something else entirely, though. They might say that only desert judgments about persons and the actions they are responsible for are “real” desert judgments, whereas any other desert judgments are merely metaphorical, or derived, just loose talking, or manners of speech. We understand them well enough, but they are not to be taken seriously since they are not to be taken literally. Alternatively, they might want to distinguish different concepts of desert, say, one that pertains to questions of justice and that deals only with persons and responsible actions, as Feinberg and Rawls have it, and other concepts which may be useful in other contexts, all of them “real” concepts of desert and to be taken seriously.

The trouble with this proposal is that it presents us with a fragmented view of desert. On the face of it, the concept of desert works the same way in any of the many varied contexts in which it may be used, it is only the individuals whose names replace the variables that are vastly heterogeneous. So, an analysis of desert as a single, unified concept which covers all cases would be, on grounds of simplicity, superior to an analysis that chopped desert up into different concepts according to context. After all, an analysis of a concept is supposed to take seriously the different ways the concept is actually used and make sense of them. An analysis which resulted in saying that there is no way to make sense of the different uses, that they have nothing in common even though the same word is used, or an analysis that said that most people use a concept in a metaphorical way only, or indeed one that came to the conclusion that people do not know what they are saying really, would be a poor analysis if there was another one available.

So, it is really three problems the responsibility view has: First, it begs the question regarding whether there can be other desert subjects than persons and other desert bases than actions for which those persons are responsible. Secondly, it has to treat other desert judgments as meaningless, where at worst they are false. Thirdly, it fragments the concept of desert.

8.1.4 The Concept of Desert

So, what is the concept of desert?¹² What does it mean to deserve something? To deserve something means that it is appropriate to get it. And what does this in turn mean, that it is appropriate? Propriety, as I will introduce the term, consists of two components: the one is a relation I shall call fittingness, and the second is a certain normative element I shall call requirement.

Fittingness is a pretty straightforward thing. Puzzle pieces, for example, fit each other. But also states of affairs may be said to fit each other. Another way of putting this would be to say that what is fitting is a “response” to what it is fitting to. Returning the ball is a fitting response to having been served in tennis; having been asked what time it is, it is fitting to tell; going for a walk is fitting to the weather’s being good; and so on, and so forth. Obviously, unlike particular cases like puzzle pieces, there may be more than one fitting response to given states of affairs. Hence, instead of going for a walk when the weather is good, it may also be fitting to go for a swim, or to hang out your laundry to dry, or to get inside if you are prone to get sunburnt. What is fitting to what really depends on the particular case and its circumstances.¹³ To come back to the responsibility view for a second, though: it should be evident that it is not only actions for which persons are responsible that can have fitting counterparts. Puzzle pieces and the weather’s being good are cases in point.

In the case of desert, the desert object needs to be fitting to the desert base. This is not all there is to it, though, because as I have pointed out at the beginning, desert is a normative notion. So, the second component I have mentioned needs to come in, the normative element of requirement. Desert is a case of not only fittingness, but of propriety, and by that I mean that it is not only a fittingness relation, but one where the fitting counterpart is required.

Clearly, not everything that is fitting is also required, that is, not everything that is fitting is also appropriate. Two puzzle pieces may be put together, but they just as well may not. They do not require being put together as if that were their natural state they belonged in. Just so, the weather’s being good does not require that I take a walk. It would just be a good idea, a fitting thing to do. I may as well not. Nothing goes wrong if I do not. By contrast, something does go wrong if what is appropriate does not occur.

Thus, to deserve something means that it is appropriate to get it, and this in turn means that it is fitting and required that the desert subject get the desert object in virtue of the desert base.

However, this still is not all there is to it, because there are cases of propriety which are not cases of desert. That is to say, there are cases where something is fitting and required, but not deserved. Consider for example a major scale. Anyone who has ever played seven notes of any major scale on any musical instrument will know that the eighth note is not only fitting, but required too. That is so because a major scale’s seventh note is a leading note that leads on to the eighth note. If the eighth note does not follow the leading note leads to nothing, and that is just not right. Something goes wrong if a major scale’s octave is not completed. This is evidenced by the fact that most listeners can hardly bear to hear it so that they add the eighth note either mentally or by singing it. This unbearableness for listeners is explained by the impropriety of the eighth note missing. Hence, it is not only fitting, but appropriate to add the eighth note after having played the first seven of a major scale. Yet, it is not the case that the first seven notes of a major scale deserve that the eighth be added.

To return again to the responsibility view: as requirement is the only difference between fittingness and propriety, and since not only actions for which people are responsible have fitting counterparts, these are also not the only things that require a fitting counterpart, as this last example of the scales shows.

The question now, though, is: as not all cases of propriety are cases of desert, which ones are? Here it would be very easy for the proponents of the responsibility view to jump in and say: only those cases of propriety are also cases of desert where the subject is a person and the base is an action for which she is responsible. To which I respond, as I have already at length in the last section, that this leaves out too many judgments of desert to be taken seriously as a good analysis of the concept.

Instead, I propose that only those cases of propriety are also cases of desert which are based on something fitting or unfitting, or in other words: desert is that propriety that is appropriate in virtue of something “fittingness-affecting”. In still other words, desert is something like “second-order fittingness”: deserving something means, utlimately, that it is fitting (and required) to get it because one has done or is something (un-)fitting.

I realize of course that at this point at the latest things are getting just a tiny bit intricate, so let me elucidate this by way of some examples. I will use some of those I have mentioned at the beginning:

• 1. Criminals deserve punishment in virtue of their crimes. This means that punishing criminals is appropriate, and this in turn means that it is fitting and required to punish criminals. What makes it so? Committing crimes, the desert base, is itself something unfitting; let us say breaking and entering is unfitting to the concept and right of property. And this is what makes it fitting and required, that is appropriate, that they be punished. So, punishing criminals is fitting (and required), because their committing crimes is unfitting. Hence, criminals deserve punishment. If committing crimes was not itself fitting or unfitting to anything, if punishing criminals was only fitting (and required), because of their committing crimes period, their punishment would not be deserved but appropriate. The normative force is the same, the punishment is required in just the same sense both times, it is just that in the first case we can say that it is “deserved” whereas in the second it is “appropriate”.

• 2. The most beautiful contestant deserves first prize in the beauty contest. In this case, being the most beautiful is the desert base, so this is what must be fitting to something. Well, the context is a contest and this contest is about beauty, so presumably, being the most beautiful is fitting to the concepts of contest and beauty. If this is so, this is what makes it appropriate that the most beautiful contestant get first prize, and hence she deserves it. Again, if things are not so, that is, if being most beautiful is not fitting to anythings, it is just appropriate period that the most beautiful contestant get first prize.

• 3. Great paintings deserve to be admired. What makes great paintings great, let us say, is that they exemplify to an extraordinarily high level what art is about. To be honest, I am making this up, I do not know the first thing about art and what makes paintings great. But remember that I am not concerned with the truth or falsity of this desert judgment, but with what it means. So, whoever says that great paintings deserve to be admired says something like this: their greatness is the paintings’ desert base; their greatness is fitting to something, let us say to the idea of what art is about; so this is what makes it appropriate to admire them, and therefore they deserve it.

• 4. Certain landscapes deserve to be preserved. For instance, the UNESCO talks about Natural Heritage Sites like this. The idea is that these landscapes have particular features that are these landscapes’ desert bases, and so that these features are fitting to something. Let us say that, whatever those features are, they are unique, and their uniqueness is fitting to Earth’s marvelous variety of landscapes. (I am speculating again.) Hence, it is the fittingness of those unique features that make it appropriate to preserve those landscapes, and so they deserve preservation.¹⁴

Obviously, not all propriety relations are based on something fitting or unfitting. The example of the major scale was a case in point. Hence, even though it is appropriate to add the eighth note to the first seven, it is not deserved, as playing seven notes of a major scale is not fitting to anything. But if it were, if there was some obscure story in which we could say that, then we would have to say that the seven notes deserve to have the eighth note added.

8.1.5 Conclusion

To wrap it all up: desert is that propriety that is based on something fitting or unfitting. So, propriety is a subset of fittingness; those fittingness relations are also propriety relations in which the fitting counterpart is not only fitting, but also required. In turn, desert is a subset of propriety; those propriety relations are also desert relations which are based on something fitting or unfitting.

This analysis makes clear that actions for which persons are responsible are not the only things there are that have fitting counterparts, or that require the fitting counterparts as appropriate, or in virtue of which one can deserve something. To think so is wrong for the three reasons I have given: it begs the question, it renders too many desert judgments meaningless, and it fragments the concept of desert.

So, the responsibility view of desert is wrong. We have to understand desert as I have explained, and there is nothing in that account that precludes other things than persons to be desert subjects nor other things than actions for which they are responsible to be desert bases.

8.2 Desert and Luck Egalitarianism

Now that the concept of desert and the role responsibility does, or rather, does not, play in relation to it are clear, it can be used to untangle misunderstandings in discussions in political philosophy in which both the concepts of responsibility and of desert figure crucially. One such discussion is the one on luck egalitarianism. It is outside the scope of this paper to engage in a fully fledged analysis of this debate, examine in detail how any given authors use the concepts of desert and responsibility, show how they go wrong on the basis of the foregoing considerations, and set them on the right track. However, I shall give a very rough and exemplary outline of how applying the correct concept of desert can help further the discussion substantially: I shall discuss two critics of luck egalitarianism, Serena Olsaretti and Nicole A Vincent, and show, first, how, with the proper account of desert expounded in this paper, luck egalitarianism can counter their criticisms, but also, secondly, that relying on this account of desert is still not enough to vindicate luck egalitarianism.

Luck egalitarianism, to quote Olsaretti, is the position that holds that those inequalities between people are unjust that are traceable to “circumstances that individuals could not reasonably foresee and avoid. By contrast, individuals are justly held responsible for, that is, they are liable to pick up the costs and reap the benefits of, events they could reasonably foresee and avoid”.¹⁵ This position is motivated by a powerful intuition Vincent calls the responsibility-tracking intuition, “i.e. the intuition that people should take responsibility for those things for which they were responsible, and that no one is entitled to expect others to take this responsibility for them”.¹⁶ So, the idea is if you make your own free choices and act accordingly, you have to suffer the consequences for better or worse. Hence, if you recklessly drive your motorcycle at high speed without wearing a helmet and have an accident, or freely gamble away all your savings, it is your own fault and you cannot expect your health insurance to pay your hospital bills and social security or anybody else to help you out.¹⁷

As common and as powerful as the responsibility-tracking intuition is, it is not quite clear what it means. Vincent argues that there are actually two distinct notions of responsibility at play here, one she calls outcome responsibility which involves attributing a particular state of affairs to a particular person, and one she calls liability responsibility which concerns the question “who should now do what” in consequence of that state of affairs being attributable to that person.¹⁸ So, the responsibility-tracking intuition, precisely understood, should really read like this: you have to take liability responsibility for what you are outcome responsible for.¹⁹

At this point, two questions arise, and though Vincent and Olsaretti both consider both questions they each specially focus on one of them: first, does liability responsibility really follow from outcome responsibility, and secondly, if so, how do we determine what consequences exactly one is liable for given one’s outcome responsibility.

8.2.1 How to Determine the Consequences One Is Liable For

To start with the second question, Olsaretti shows that, even assuming that liability responsibility does follow from outcome responsibility it is not as easy to determine what consequences exactly one is liability responsible for given one’s outcome responsibility as luck egalitarians relying on the responsibility-tracking intuition would have it:²⁰

consider Bert’s [the reckless motorcyclist’s driving at high speed without a helmet and subsequently injured in an accident] case more closely. On reflection, it appears that these consequences are not so self-evident after all. For example, do these consequences include being left to the side of the road? Even if not, is the strength of the obligation on passers-by an obligation to take him to a hospital conditional on the gravity of Bert’s condition and/or on the costs, to them, of taking him to a hospital? Should Bert pay for treatment only of those injuries that resulted from the accident itself, or also for medical conditions that resulted from the unforeseeable effect of the accident on certain hitherto unknown predispositions to illnesses? Or even for any medical treatment he will need henceforth? And at what price should the treatment be charged, so that that price may also be deemed “a consequence of his action”? (May a hospital have a policy of charging imprudent motorcyclists more than others?) Are the consequences of Bert’s action also that passers-by may appropriate his motorbike from the side of the road? May he lose his job if, once he has recovered from his accident, his limpness makes him a less attractive employee? May he be denied life insurance henceforth?

The list of questions could go on.²¹

So, the responsibility-tracking intuition by itself does not determine what consequences should follow from outcome responsibility. All sorts of consequences issue from a given outcome, and we need to rule out those that are “unduly harsh towards those who end up in dire straits through their own choices”²² as well as those that “are of the wrong, because irrelevant, sort”.²³ Only if we can do this, Olsaretti maintains, are the inequalities that result from people’s choices justified on luck egalitarian terms. Hence, in addition to the responsibility-tracking intuition luck egalitarians need what Olsaretti calls a “principle of stakes” which does just that.

The trouble, though, is, according to Olsaretti, that no account of a principle of stakes works, and that, therefore, luck egalitarianism should be rejected.

I am sure Olsaretti is right about most of the candidates for a principle of stakes she considers and that they do not serve to supplement the responsibility-tracking intuition in the requisite way. However, she also considers and rejects desert as a principle of stakes, and here, naturally, I beg to differ from her. Though she clearly sees that there is an advantage to the desert view, namely a “proportionality constraint” which rules out unduly harsh consequences, this is to “deliver too little … for too high a price”,²⁴ because desert by itself cannot determine what consequences outcome responsibility might justifiably have. Rather, outside considerations are needed, so that this commits us to a view of responsibility as “[uniquely entailing] one’s own favoured account of stakes”.²⁵ In the case of desert as a principle of stakes, this means that, since, “as even defenders of desert have been willing to grant”, what someone deserves “is settled by the institutional context in which desert claims are made, rather than by the notion of desert itself”.²⁶ Hence, adopting desert as luck egalitarianism’s principle of stakes would mean understanding responsibility as “entailing” institutions that settle desert and, by extension, responsibility, which is said “too high a price”, therefore we should not adopt it, and so luck egalitarianism fails.²⁷

There are several problems with this line of argument, though. First of all, here is a defender of desert who is not willing to grant that what someone deserves is in every case settled within an institutional context. Sometimes this is the case, but more often than not, it is not. And even the institutional cases of desert are not intelligible without understanding desert preinstitutionally. So, no, desert is not essentially institutional.²⁸

Secondly, as I have explained in the foregoing sections, desert involves far more than “proportionality”, by which I take Olsaretti to mean what I call “fittingness”. While this is a central component, as should be evident by now, it is not all there is to it. Desert is not only fittingness, but required fittingness, that is propriety. Hence, to adopt desert as the principle of stakes for liability responsibility allows us not only to find out which consequences are fitting that a subject bear for her outcome responsibility, but also requires her bearing them. So, the link between outcome responsibility and liability responsibility provided by desert also comes with the requisite normative force. Hence, adopting desert as the principle of stakes does not provide “too little”, but just the thing required to serve luck egalitarianism’s purposes.

Finally, Olsaretti has a problem with desert as a principle of stakes, because she mistakenly believes that what someone deserves is settled by institutions “rather than by the notion of desert itself”, which leads her to conclude that one needs to understand responsibility as entailing institutions. Presumably, if what someone deserved was settled by “the notion of desert itself” the problem would not arise. However, as should be clear from the preceding discussion of desert, “the notion of desert itself” is a formally normative relation that by itself does not determine anything. What desert object a desert subject deserves is in every case, institutional or not, settled by the desert base and nothing else. To say that the subject deserves it is not to settle anything, it is just to say that it is appropriate that she get it. Hence, to demand that what someone deserves to be settled “by the notion of desert itself”, because otherwise it cannot serve as a principle of stakes for luck egalitarianism, is unreasonable.

Incidentally, as Vincent shows, the same is true of the notion of responsibility: it “only provides a formal structure within which … other normative considerations determine how people may be treated, but contrary to what most people seem to think responsibility does not generate practical demands of its own”.²⁹ So treating the question of the principle of stakes as “a question of what responsibility itself requires”,³⁰ as Olsaretti does, is equally unreasonable.

Thus, that other considerations settle what someone deserves, and by extension what they are liability responsible for, is not to understand responsibility as entailing those other things. This, indeed, would be “too high a price” and absurd too, but since both responsibility and desert are just formal relations that provide normative links between different things, the problem does not arise. No relation conceptually entails the individuals it relates to each other.

These considerations certainly do commit luck egalitarianism to “a particular view about the principle identifying the grounds of responsibility”, namely the desert view.³¹ I fail to see, though, how this constitutes “too high a price”, because this is just what a principle of stakes is supposed to do on Olsaretti’s own terms. A problem arises only if we do not keep in mind what sort of responsibility we are talking about here. Obviously, desert cannot ground outcome responsibility. To say that I am only outcome responsible for what I deserve to bring about is bizarre. However, we are not talking about outcome responsibility here, but about liability responsibility. So, in looking for a way to normatively link consequences that are neither unduly harsh nor irrelevant to certain states of affairs someone is outcome responsible for, we are indeed looking for a principle of stakes that identifies the grounds of liability responsibility, and desert can fulfill this need. Far from being too high a price then, this is just what Olsaretti said luck egalitarianism needed.

So, if luck egalitarianism adopts desert as the principle of stakes, then, contra Olsaretti, it can rule out consequences that are unduly harsh or irrelevant as consequences someone should be liability responsible for given her outcome responsibility. However, luck egalitarianism still needs to establish that liability responsibility really does follow from outcome responsibility in the first place.

8.2.2 How to Derive Liability Responsibility from Outcome Responsibility

As I have already said, Vincent shows that outcome responsibility and liability responsibility are two different responsibility concepts. Not only do they refer to different objects, but they also are differently orientated in time: while outcome responsibility looks backwards, liability responsibility looks forward. Hence, outcome responsibility refers to a state of affairs one has brought about in the past, whereas liability responsibility refers to consequences one will bear in the future.

Since outcome responsibility and liability responsibility are two quite different concepts of responsibility, Vincent argues that the one does not follow from the other “automatically”, that is it is not logically entailed, as the responsibility-tracking intuition would have it. Rather, it would need additional normative premises to “bridge the inference gap”, for example something like: “those who are outcome responsible for X should take liability responsibility in manner Y”. Vincent calls such additional normative premises “reactive norms, since they are norms that govern our reactions to outcome responsible parties”.³² However, luck egalitarians do not offer any reactive norms as bridging premises, but simply assume that liability responsibility automatically follows from outcome responsibility. Since this is wrong, Vincent concludes that the responsibility-tracking intuition must be rejected.³³

This conclusion is surprising in its abruptness, because, first, we could just have luck egalitarians read Vincent’s analysis and surely they would immediately recognize the need to offer reactive norms and do so. Secondly, Vincent herself thinks that there are reactive norms that bridge the inference gap since there are “normative considerations that … play a key role in validating the transition from claims about a person’s outcome responsibility to conclusions about their liability responsibility”.³⁴ Thirdly, the responsibility-tracking intuition is a very powerful intuition many people share, and not just luck egalitarians. We should be wary of discarding it just like that, but rather see whether there is a way of retaining it without running into the problems Vincent points out.

Regarding the first point that luck egalitarians might just agree with Vincent and belatedly offer reactive norms to supplement their theory, Vincent might reply that this essentially is to give up luck egalitarianism, because, to borrow from Olsaretti, the idea behind luck egalitarianism presumably is to see only those inequalities as justified that can be derived from a liberal concept of freedom of choice and the concept of responsibility which are compatible with luck egalitarianism.³⁵ Thus, the idea seems to be to start out from concepts as thin as can be so as to get as widespread approval as possible and justify inequalities from there. Suggesting to simply add some reactive norms really amounts to abandoning this underlying idea, because all of a sudden the starting points do not seem to be so thin and universally approvable anymore.

Well, if luck egalitarians really are as inflexible as all that, so much the worse for them, of course. However, I am not convinced that there might not be a more yielding kind of luck egalitarian who would not mind a spot of extra justification for the odd reactive norm or two supplementing the responsibility-tracking intuition. Still, even if there is not, luck egalitarians are not the only people who want to derive liability responsibility from outcome responsibility. In fact, and this is the second point, Vincent herself seems to want to do just that, and she introduces reactive norms to do the very trick.

And to be sure, reactive norms will serve. Vincent correctly argues that we need some additional premises to bridge the inference gap from outcome responsibility to liability responsibility in the responsibility-tracking intuition, and her reactive norms do that. But she also sees that “this now raises the question of where such reactive norms might come from”. So, she recognizes the need to justify those norms, because they may allow treating people harshly, and this is where normative considerations of, say, justice, utility, caring, beneficence and so on come in that inform those norms.³⁶

However, this will not do. The problem with Vincent’s account of reactive norms as bridging premises is that they are arbitrary. Reactive norms would not only provide a link between outcome responsibility and liability responsibility but also spell out the consequences one is liability responsible for. However, as we have seen in the last section, it is important that this be a link such that it connects only the right sort of consequence to outcome responsibility.

Informing reactive norms by normative considerations like utility or caring does not help in this regard. Say we care about our fellow citizens’ safety so much that we really want them to wear helmets when riding their motorbikes so as to spare them the ghastly consequences of possible head injuries. Moreover, those head injuries are a real strain on our health insurance system. In order to achieve this we institute a reactive norm that motorcyclists who have an accident while riding without a helmet be ordered to scrape chewing gum from underneath school desks as soon as they recover. This is both caring and not at all too harsh, as almost every motorcyclist is bullied into wearing a helmet by the prospect of this distasteful task resulting in a significant drop in those specific head injuries, and those who are not at least do something useful to pay something back to society. The trouble is that this consequence of gum-scraping is not covered by the responsibility-tracking intuition. What is powerful about the intuition, even though it employs two different responsibility concepts, is that we intuitively see that there is something to the idea that people who are in a situation through their own fault should deal with it themselves, that they should suffer the consequences. However, in the example it is all too natural to ask, “Why gum-scraping?” So, the intuition does not cover any consequences that may arise out of situations people are in through their own fault, but only those that are linked to them in some special way. Linking them through a reactive norm does not work, as there is a reactive norm in place in the example, but we still do not intuitively see that a reckless motorcyclist really should scrape gums. This might be covered by the reactive norm, but it is not covered by the responsibility-tracking intuition.

Therefore, while reactive norms can serve as normative premises bridging the inference gap from outcome responsibility to liability responsibility, they can do so only in an arbitrary way, and thus fail as an explanation of the normative force of the responsibility-tracking intuition.

In view of this, Vincent might be tempted to reject the responsibility-tracking intuition altogether: we cannot derive liability responsibility from outcome responsibility automatically, we need reactive norms as additional bridging premises. Reactive norms fail to do the trick, so let us chuck the responsibility-tracking intuition and face it – liability responsibility cannot be derived from outcome responsibility at all.³⁷ But this would be too easy a dismissal.

Whether or not Vincent believes in the responsibility-tracking intuition, many, if not most, people do. It is such a powerful intuition, and this is the third point mentioned at the outset of this discussion, that we should be loath to go to such lengths as rejecting it if not necessary. And it is not, for we can tell which specific consequences are responses to certain states of affairs they issue from, and which are not. In the first case, the consequences are fitting to what happened before, in the latter, they are not. And the force of the responsibility-tracking intuition, its being so widespread and powerful, indicates that they are not only fitting, but required, hence appropriate. So, as easy as that, we can derive liability responsibility from outcome responsibility: a person should take liability responsibility for a state of affairs she is outcome responsible for, because that is the appropriate thing for her to do.

So, there is a certain link between reckless, helmetless motorcyclists getting into accidents and their paying their hospital bills themselves, namely the latter being a response to the former, while there is no such link – other things being equal – between their getting helmetless into accidents and scraping gum from underneath school desks. Hence, with the help of the notion of a “response” we get what consequences are fitting to what states of affairs. And the responsibility-tracking is evidence for our believing that they ought to be thus linked, that they are not only fitting but appropriate.³⁸

We can probably even say that, given the dangers of motorcycling and the fragility of the human body, motorcyclists do something unfitting when they drive without helmets, and therefore, that it is not only appropriate but that they deserve to pay their own hospital bills. Obviously, though, whether or not we can say this, whether the person outcome responsible deserves to take liability reponsibility, depends on the responsibility-tracking example. At the least, however, it is appropriate for her to do so.

Some might worry that this is nothing but an exercise in terminology, that I have invented a new name for the responsibility-tracking intuition, but have not justified it. The last bit of this is true, I have not justified the responsibility-tracking intuition, but that is because that is not my business here. Rather, I have tried to make sense of how exactly to derive liability responsibility from outcome responsibility. In order to do this, I have simply taken for granted that liability responsibility should follow from outcome responsibility. I have done so on evidence of the responsibility-tracking intuition’s plausibility and force. The trouble was, as Vincent showed, that the intuition is not enough to derive liability responsibility from outcome responsibility, it needs to be supplemented. Vincent tries to do this with reactive norms, but that will not work. The foregoing considerations show that, rather, the responsibility-tracking intuition needs to be supplemented by the notion of propriety. This is not to give just a new name to the game. It is not to say that persons outcome responsible ought to take liability responsibility because they ought to, it is to say that they ought to because it is appropriate. And that this is not the same is what I have shown in the first part of this paper.

8.2.3 Two Questions or One?

Now it seems as though in the last two sections I have given two rather similar answers to the two quite distinct questions Olsaretti and Vincent focus on, that is on the one hand the question how the transition from outcome responsibility to liability responsibility works, and on the other hand the further question, assuming the transition does work, what consequences one should be liable for given one’s outcome responsibility. In answer to both questions I have argued that desert will do the respective trick. So, does it not seem as though I have become confused about what the difference between the two questions is somewhere along the way? Well, no. Rather, the concept of desert is so great that it is capable of doing both tricks at once. Let us see how this works.

Remember that the concept of desert is made up of two components, fittingness and a normative element which makes what is fitting appropriate, that is which requires that the desert subject, characterized by the desert base, get the fitting desert object. The second question, what consequences one should be liable for given one’s outcome responsibility, is the question what desert object is a response to a particular desert base a particular desert subject has set. The first question, how liability responsibility follows from outcome responsibility, is the question of the normative force of propriety in this case. I have not specially argued that there is such normative force there in this case, but have simply taken the power of the responsibility-tracking intuition as sufficient evidence for its holding here.

In other words, in the desert relation “x deserves y in virtue of z”, the second question concerns the fittingness between y as liability responsibility and z as outcome responsibility of subject x, whereas the first question concerns the transition marked by “deserves”.³⁹

So, rather than having confused the two questions by employing the concept of desert in the discussion of luck egalitarianism, I have killed two birds with one stone.

8.2.4 Luck Egalitarianism

Now, what does all of this mean for luck egalitarianism? Vincent and Olsaretti reject luck egalitarianism, or the responsibility-tracking intuition it rests on the, grounds that luck egalitarians cannot provide answers to the two questions mentioned. I have shown that luck egalitarianism cannot be rejected for that reason, because, if luck egalitarianism is supplemented by the concept of desert as explained, it can provide such answers after all. However, this does not mean that desert is all in the luck egalitarians’ court. Luck egalitarians may employ this concept in order to draw the conclusions they want to draw, but that does not mean that desert may not equally well be employed by luck egalitarianism’s opponents.

This is so because, as I have now said repeatedly, the concept of desert by itself does not determine who deserves what for what, and because, as I have emphasized all along, the concept of desert does not presuppose the concept of responsibility.

So, while it is reasonable to suppose that an unemployed gambler who has lost all his savings in a casino (who is outcome responsible for this) deserves to be barred from welfare aid (ought to take liability responsibility in this manner) it is equally reasonable to suppose that the gambler’s situation now being thus that he has no adequate access to food, medicine, and shelter, he deserves to get welfare aid. This being equally reasonable means that in both cases the consequences are equally good responses to the antecedent states of affairs, so that we can say that the gambler both deserves to get welfare aid in virtue of his situation and deserves to be barred from welfare aid in virtue of his outcome responsibility.

Luck egalitarians, therefore, cannot simply point to the gambler’s desert in order to draw the conclusion that he ought to be barred from welfare aid, because there are other considerations of desert to be taken into account, and since all of these are equally considerations of desert none of them are more weighty than the others simply on the grounds of their being deserved. So, in this regard the concept of desert will not help the luck egalitarian.

Neither can luck egalitarians point out that it is the gambler’s own fault, the gambler is liability responsible and so he should be barred from welfare aid no matter what other desert considerations count against that, because those other desert considerations are not based on responsibility. The answer to that would be: so what? What on Earth is so very special about responsibility that it should break the tie between different considerations of desert, the normative force of which I have already shown does not rest in the slightest on responsibility? We would need additional arguments for that.

Nor can luck egalitarians point to the concept of justice and say that the gambler’s deserving to be barred from welfare aid is a matter of justice whereas the gambler’s deserving welfare aid is a matter of charity, and justice outweighs charity. Again, this would need some substantial arguments, that justice is only concerned with what people are responsible for, which have as of yet not been given, and that would be a counterintuitively thin notion of justice anyway.⁴⁰ After all, it is not implausible to suppose that giving people what they deserve is always a matter of justice. Moreover, even if the one really was a matter of justice and the other a matter of charity – whoever said that justice was more important?

Well, the luck egalitarian might say, all of this is well and good, and if this was an ideal world in which there was enough for everyone no problem would arise. Alas, our resources are scarce, there is not enough for everyone, not even enough for everyone deserving, so that unfortunately we need to limit distribution in some way. Surely, it is preferable to do so in a just way. And distributing resources according to desert based on outcome responsibility is such a just way after all. So, given the scarcity of resources, such distribution is justified since at least I have granted that it is just.

Certainly, but this would be just to settle the matter by pragmatic considerations. This does not make picking out desert based on outcome responsibility of all desert considerations any more warranted in terms of justice, and it does not settle the matter regarding other possible just ways to distribute scarce resources, that is, this consideration by itself does not show that from among various just ways to distribute scarce resources we should pick the luck egalitarian option.

8.3 Conclusion

So, there is no conceptually necessary link between desert and responsibility. The responsibility view of desert is wrong. Nevertheless, if luck egalitarians avail themselves of the concept of desert, they thereby have an instrument with the help of which they can both explain how to derive liability responsibility from outcome responsibility and what specific consequences one is liability responsible for given one’s outcome responsibility. However, it is important for them to note that this is only an instrument with the help of which they can explain these things. The concept of desert is not there to do all their argumentative work for them. They still need to provide substantial arguments to support their conclusions, and so far, these arguments have not been forthcoming.