1 Introduction

One of the primary threats to a meritocratic conception of distributive justice is the concern that the accumulation of merit is an unavoidably lucky process. At minimum, the accumulation of merit—which will require all that is necessary to adequately develop one’s abilities (care, education, various public goods and opportunities, etc.)—depends on one’s own efforts, but also on a favorable early environment and innate talent. The concern, famously pressed by Rawls, is that early environment and innate talent are purely matters of luck.Footnote 1 But if one’s accumulation of merit is a matter of luck, then it does not seem like they are particularly deserving of anything on the basis of those merits. And if no one is or can be particularly deserving of anything on the basis of their merits, then it’s hard to see how distributive justice could be meritocratic. We can call this the “luck objection” to a meritocratic conception of distributive justice.

The luck objection suggests the “Luck Constraint”:

Luck Constraint: To the extent that one’s merits are had as a matter of luck, then those merits do not ground one’s deserts.

Even if merit accumulation is not a purely lucky process, the concern is that the luckier the process is, the less deserving one is of economic benefit, and thus the moral force of desert claims is weakened.Footnote 2 But for Mulligan (and many, but not all meritocrats) the normative value of having income (for example) distributed in accordance with the distribution of merit is derivative on the value of people getting what they deserve. Thus, if, in the context of distributive justice, the moral force of desert claims is not particularly strong, then either the demands of distributive justice are, themselves, quite weak, or else distributive justice is not meritocratic.

In his Justice and the Meritocratic State (2018a), Thomas Mulligan gives a novel defense of meritocracy against the luck objection. His strategy is to argue that, contra Rawls, the accumulation of merit in a proper meritocracy would not be particularly lucky. First, meritocracy requires fair equality of opportunity, and so in a proper meritocracy, society would be arranged in such a way that there will not be major differences in the accumulation of merit that trace back to differences in social class, or the quality of one’s early environment more generally.Footnote 3

But what about differences in merit that are due to differences in innate talent? Mulligan’s answer here is more surprising: those differences will remain, but the differences will not be, and currently are not matters of luck. His thought is that, given Kripke’s essentiality of origin thesis—the idea that we could not have originated from any sperm and egg pair other than the one that we did originate from—and given that our innate talents are determined by our genetic makeup, our innate talents are essential to us.Footnote 4 In that case, it is not a matter of luck that we have those talents, because we could not be any other way. Luck, presumably, requires contingency, and while we might say that we are lucky to exist in the first place because we might have failed to exist, it is not lucky that we have the innate talents that we have, because they are essential to our existence. Thus, if we are not lucky (or unlucky) to have our innate talents, then the merits that we accrue as the result of developing and exercising those innate talents are not had as a matter of luck either. The meritocratic conception of distributive justice is thereby rescued from the luck objection.

The aim of this paper is to argue that the appeal to the essentiality of origin fails to save meritocratic accounts of distributive justice from the luck objection. There are two primary problems with the defense. First, even if our origin is essential to us, it does not follow that the accumulation of merit on the basis of our innate talents is not, to a significant degree, a matter of luck—even in an idealized meritocracy of the sort that Mulligan envisions. I give this objection in Section 2. In Section 3, I present a second, deeper objection to Mulligan’s appeal to the essentiality of origin. In short, the essentiality of origin thesis misses the underlying motivation for the luck objection. The connection between luck and desert that is spelled out in the luck constraint is not primitive, but rather follows from a deeper connection between desert and responsibility. Thus, even if the possession of our innate talents is not, strictly speaking, lucky, it is clearly not the case that we are responsible for the possession of those abilities. But absence of responsibility weakens desert claims as much as luck does, and so the appeal to the essentiality of origin thesis misses the point of the deeper objection to a meritocratic account of distributive justice.

2 The Luckiness of Economic Contribution in a Meritocracy

The basic idea of Mulligan’s defense against the luck objection, again, is that given the truth of the essentiality of origin thesis, it is not a matter of luck that we resulted from the particular sperm and egg pair that is our origin. Luck, the thought goes, requires contingency. From here, Mulligan moves to the idea that because our innate talents are determined by our genetic endowment, and because our genetic endowment is determined by our sperm-and-egg-pair origin, it is not lucky that we have the innate talents that we have. This much can be granted. We can skip over, for instance, the questions of developmental psychology and biology—whether there are any sort of necessary connections between one’s origin, their genetic endowment, and their innate cognitive and physical qualities, etc. Whatever these connections are, they are surely not a matter of metaphysical necessity, but since these sources of contingency are presumably relatively minor, we can ignore them for the sake of argument.

The problem for Mulligan comes in the next step. He concludes from the fact that our innate talents are not had as a matter of luck, that the merits that we accumulate on the basis of those talents are not had as a matter of luck. But even when we assume fair equality of opportunity, this is not the case, because what counts as an economic contribution, and thus what counts as merit, depends on the contingencies of the market and one’s social environment more generally.

Now, some care is required here, since Mulligan is not giving a meritocratic justification of market economies as they currently operate.Footnote 5 Rather, his aim is to sketch the basics of a meritocratic economy in which incomes are deserved, and thus in which income is not so heavily dependent on luck.Footnote 6 Accumulating merit in Mulligan’s envisioned, meritocratic economy is, to be sure, a significantly less lucky process than it is in our current world. Nevertheless, I argue, even when we assume the essentiality of origin, in addition to the meritocratic adjustments to the distribution of income that Mulligan proposes, the process of accumulating merit on the basis of our innate talents is still a substantially lucky matter.

In order to get a better sense of the kind of meritocratic economy that Mulligan envisions, however, it will be useful to start with a discussion of certain kinds of luck that will be ruled out in a meritocratic economy.

2.1 How A Meritocratic Economy Reduces the Role of Luck

Given that we are assuming that fair equality of opportunity would obtain in a meritocracy, we can ignore the luck that is currently involved in having the opportunities to develop one’s talents. Fair equality of opportunity requires that “those who have the same talent and ability and the same willingness to use [their] gifts...have the same prospects of success regardless of their social class of origin.”Footnote 7 We assume, for example, a robust and equal educational system which favors no particular social class, and we also assume meritocratic hiring, so that only one’s “abilities and willingness to use them” determine who gets to use their talents to make economic contributions via some kind of employment.

Mulligan also rules out what Miller (1999, 143-144) calls ‘integral luck’. Integral luck concerns the gap between one’s abilities and efforts on the one hand, and the result of the effort on the other. For example, a foolish investment might—because of unforeseen events—pan out, while a smart investment might prove disastrous, just as the fastest runner will not always win the race because of the occasional occurrence of some freak accident which causes them to lose. Importantly, for Mulligan, what grounds the desert of income is not mere economic contribution, but ‘meritorious economic contribution” (2018a, 130). Roughly, the idea is just that what grounds desert—i.e. what merit is—is not one’s economic contribution, but rather the excellence of one’s economic actions, where this is a function of the actor’s ability and their diligence in using their ability in the service of making an economic contribution.

Finally, Mulligan argues that economic rents are not deserved, and so in a meritocratic economy, there will be no rent-based income.Footnote 8 Economic rents are “returns to factors of production in excess of what is necessary to maintain them in economic use” (2018a, 135). In short, rents are incomes which reflect no actual economic contribution. Examples might include someone who commands a large salary because they have some skill which is in demand, but which is quite rare. If the skill was in greater supply—i.e. if there was adequate competition among those selling that skill—then they would earn far less.Footnote 9 Their large income, in this case, does not reflect their great merits or the magnitude of their contribution so much as it reflects the absence of competition. That Mulligan takes rents to be undeserved is important for the argument of this paper, since it eliminates major sources of luck from a meritocratic economy. In the above case, for instance, while (given essentiality of origin) it might not be lucky that someone possesses a certain talent, which happens to be rare and highly desirable, the fact that it is rare (because there are not others who possess the talent) is very lucky.

2.2 Why Talent-Based Luck Still Remains

The accumulation of merit in Mulligan’s meritocratic economy would be significantly less lucky than in modern market economies. Nevertheless, as I will argue next, there are significant elements of luck that would remain, and it would still be the case that people with certain sets of talents are luckier than others. The problem is not that anyone is lucky to have the talents they have, but rather that they are lucky that they live in a world which favors their talents, or unlucky that it does not.

The first problem stems from the fact that how great or how small an economic contribution is—and indeed, whether some activity counts as an economic contribution at all—depends on the demand for those activities, and thus, the preferences of everyone else. Thus, while some of our talents might be essential to us, the fact that they are in demand is not. It is lucky if they are, and unlucky if they are not. Consider, for example, the recent ”chess boom” which has resulted from the popularity of the hit series The Queen’s Gambit. Prior to the release of the show, interest in and demand for chess was relatively low. Since then, however, demand has exploded, as have the economic contributions of some of the world’s great chess players, as they have capitalized on the surge of demand for chess on online streaming platforms like Twitch and YouTube. Or to take a much more far-reaching example, those who have mathematical talents are, on balance, currently luckier than those who have talents in, say, theater, given the differences in demand for these talents in the present economy.Footnote 10

Crucially, fair equality of opportunity does not rule out this sort of luck. What fair equality of opportunity ensures is just that people of identical talents and willingness to use those talents should have equal chances to use their talents to make an economic contribution of a certain magnitude. It does not require that all talents should afford the same opportunities for success.Footnote 11 This, of course, is as intended, as many talents are plainly useless with respect to their social contribution, and so cannot ground the desert of income. Appealing to the distinction between economic contribution and meritorious economic contribution does not help either. The struggling artist may be as skilled and diligent as the thriving computer programmer in their respective fields. Nevertheless, the economic contributions of the computer programmer are more meritorious, as it is plain to both of them that, due to differences in demand, the programmer’s actions result in a greater contribution.

The second way in which people of certain talents might be luckier than others has to do with the fact that, even if you are not lucky to have your own talents, you might be lucky that others have or lack certain talents, even when we assume that economic rents cannot be deserved.

Consider, for example, “superstar incomes”, which Mulligan argues are a kind of economic rent, and are, therefore, not deserved.Footnote 12 The idea is that, in some fields, the very best will earn much, much more than those that are otherwise exceptional or very good, but not the best. A superstar musician, for example, might only be a little more talented or diligent than another musician, and yet might have an income that is many times greater than the slightly less meritorious musician. Much of the superstar’s income will qualify as economic rent on Mulligan’s view, and does not reflect the minor differences in merit, and so will not be deserved.Footnote 13

But while a superstar does not deserve an astronomically greater income relative to their slightly less able peers, it is not denied that the superstar should earn proportionately more than their peers. In that case, being the best at something can make you deserving of a greater economic reward, as you are in a position to make a greater economic contribution than you otherwise would. But of course, “being best” is a relational property, and depends on facts about the abilities of everyone else. Assuming there is a fixed demand for great music, for instance, then the second best musician is unlucky that there is someone better, who, because of their greater merits as a musician, deserves more.Footnote 14 If the better musician did not exist, then the currently second-best musician would deserve more than they currently do, as they would then best be able to satisfy peoples’ desire to consume great music.Footnote 15 This point is reminiscent of Rawls’ claim that “We do not deserve our place in the distribution of native endowments, any more than we deserve our initial starting place in society.”Footnote 16 The point, for our purposes, is not that our native endowments are a matter of luck, but that our place in the distribution of native endowments is. Whatever our particular talents may be, it can often help—with respect to the accumulation of economic merit—that others do not have those talents to the same degree.

A third source of luck simply has to do with the relational nature of ability. The kinds of actions one can perform—and especially the kinds of actions that count as economic contributions—depend on features of one’s external environment. Ability and disability, or what one is capable of, depends on the obstacles and infrastructure that may prevent or enable various actions by an individual.Footnote 17 The effectiveness and availability of eyeglasses, for instance, makes it the case that those with poor vision can engage in just about all of the same activities that they could if they had perfect vision. Those with poor eyesight are lucky that they live in a society in which poor eyesight is not debilitating, while those who are debilitated are unlucky that they live in a society that does not or cannot enable them. In short, how useful one’s innate endowments will be in the pursuit of merit depends greatly on whether one’s society enables the kind of actions which will allow one to translate their talents (or qualities, generally) into merit.

Or consider again our chess example. There might be great demand for high level chess play, but in the absence of online chess platforms which allow for the availability of high level competition at all hours of the day, and the streaming platforms which allow for the easy broadcast of the play to a massive audience, the economic contribution one could make through playing chess would be greatly limited. The kinds of actions which can count as economic contributions are greatly circumscribed by the infrastructure and activities of others that makes those actions possible.Footnote 18 To the extent that the existing infrastructure enables people to use their abilities to make economic contributions, they are luckier than those for whom it is more difficult to make contributions via use of their talents.

There are likely other sources of luck that we could identify (for instance, the extent to which one enjoys developing and employing the talent which would make the greatest contribution), but I take it that the foregoing is sufficient to establish the pervasive influence of luck in the accumulation of merit as a result of differences in innate talents, even in an idealized meritocracy of the sort Mulligan envisions, and even when we assume the essentiality of origin thesis. In the next section, however, I argue that there is a deeper problem with the appeal to the essentiality of origin to rescue meritocracy from the luck objection, which is that it ignores the crucial connection between desert and responsibility.

3 Luck, Responsibility, and Desert

The second, more fundamental reason that the appeal to essentiality of origin fails to rescue the possibility of meritocracy from the luck objection is that it ignores the underlying motivation for the luck constraint, and thus misses the underlying point of the objection. Plausibly, the luck constraint is not primitive, but instead follows from the “responsibility constraint” on desert bases:

Responsibility Constraint: S deserves O on the basis of B only if (and to the extent that) S is responsible for B.

The responsibility constraint enjoys fairly wide support among desert theorists.Footnote 19 It explains, for instance, why one’s voluntary actions, one’s cultivated skills or character traits, are all paradigmatic desert bases.Footnote 20 It also explains the luck constraint, since if some desert base B is had purely as a matter of luck, then clearly a subject is not responsible for the desert base obtaining.

The appeal to essentiality of origin in response to the luck objection relies on the fact that the luck constraint does not entail the responsibility constraint. So long as we think of luck as depending on the idea that things could have been otherwise, then no necessary truth can be lucky, despite the fact that its obtaining is outside of anyone’s control. Nevertheless, if what matters for desert is that you are responsible for your desert bases, then the appeal to essentiality of origin is of no help to the meritocrat, since—lucky or not—no one is responsible for their origin.

Mulligan, unsurprisingly, does not endorse the responsibility constraint. Since he accepts the luck constraint, however, he cannot adopt a view of desert which is more along the lines of Miller (1992), who argues, for example, that the best candidate deserves the job, regardless of whether their being the best is a matter of luck or not.Footnote 21 The question, then, is whether the responsibility constraint should be adopted in conjunction with the luck constraint, or whether one can reasonably accept the latter without the former. I will now argue that it would be better to accept the responsibility constraint.

The first reason to prefer the responsibility constraint in conjunction with the luck constraint is just that, again, the responsibility constraint explains the luck constraint. Without the responsibility constraint, the luck constraint must be taken on as a primitive constraint on desert. One wants to know why, however, the notion of contingency bears such an important connection to a normative concept like that of desert. The notion of responsibility provides the link, since one is not responsible for things that—as a matter of metaphysical necessity—could not be otherwise. Contingency, in other words, is a necessary condition for responsibility. Taking on the responsibility constraint, therefore, fills an explanatory gap, and makes more coherent the connections between desert, luck, and responsibility.

The responsibility constraint also figures in explanations of a range of robust intuitions concerning desert. For example, someone who favors a retributivist view of punishment—on which the aim of punishment is, at least in part, to ensure that the wrongdoer gets what they deserve—will be highly sensitive to mitigating factors which undermine one’s responsibility for the wrongdoing.Footnote 22 If, for instance, one commits a crime because they are in the grip of some hallucination, through no fault of their own, we will tend not to think they deserve as severe a punishment as they would deserve if there were seeing the world clearly. In such a state, we tend not to think one is as responsible for their actions as they would normally be. To the extent that one is less responsible for the wrongdoing, they are deserving of a lesser punishment than they would deserve if they were fully responsible.

More relevant to the meritocratic conception of distributive justice, the responsibility constraint helps to explain why fair equality of opportunity is necessary for meritocracy. For example, it is intuitive that in a race, if one runner starts halfway to the finish line, then when they win the race, they likely do not deserve the prize. Likewise, Mulligan argues, the most meritorious applicant will not always deserve to be admitted to the prestigious school. After all, it is well established that, for example, children raised by wealthy families will often have considerable advantages over their less wealthy peers when it comes to accumulating the sort of merit that is relevant to college admissions. Without fair equality of opportunity, then those with the most merits will typically be less deserving of what they would deserve if fair equality of opportunity obtained (assuming the most meritorious also have disproportionate access to opportunity).

Once again, however, we can ask for an explanation of the connection between equality of opportunity and desert. In this case, Mulligan does not take the connection to be primitive. Rather, he thinks that it follows from the “aboutness constraint” on desert bases:

Aboutness Constraint::

S deserves O on the basis of B only if B is about S.

The thought here is just that you can’t deserve something because of some fact or property of something else. One cannot deserve punishment, for example, because somebody else committed some wrongdoing, and a student cannot deserve a good grade on the basis of somebody else’s good work.

The aboutness constraint is uncontroversial. However, it’s not so clear that it explains the connection between fair equality of opportunity and desert. Mulligan uses the following example to motivate the connection. Suppose Daisy gains admission to a prestigious school over a less meritorious, but equally talented and hard-working peer, where Daisy’s greater merits are primarily the result of her advantageous upbringing, which gave her access to higher quality schools, more enriching extra-curricular activities, and so on. Had Daisy’s peer had the same opportunities for educational development, let’s suppose, they would be equally meritorious by the time they apply for admission. Mulligan claims that, in this case, Daisy does not deserve admissions over her peer (despite her greater merit) because “these historically-grounded components of Daisy’s merit are not really about Daisy; they cannot ground her desert-claim here”. (2018a, 72)

But it is not clear, then, how we should interpret aboutness in these cases. The purported desert basis for Daisy’s admission to the school, in this case, is her merits—her grades, her attendance at a prestigious high school, and so on. All of these are properties of Daisy, and not anyone else. Mulligan does not deny this, of course, but he argues that the sense in which these are facts about Daisy is a “weak” sense of aboutness. But without further elaboration of what the stronger aboutness principle comes to, it’s not clear why the facts about Daisy are not about her in the relevant sense.

Of course, the explanation for why she has those merits must appeal to all manner of facts, only some of which are about Daisy. For instance, Daisy’s efforts are obviously an important part of the explanation, but so are facts about her family, and the fact that she lives in a society where the wealthy are better able to secure favorable educational opportunities for their children. Obviously, when we ask why she has more merit than a less advantaged student, her fortunate upbringing becomes the salient answer.

But aboutness, it would seem, is completely unconcerned with explanations of why a particular desert basis obtains (in this case Daisy’s merits). The responsibility constraint, however, is very much concerned with explanations of why a particular desert basis obtains. It requires that, to be deserving of some object, one is responsible for the fact that the desert basis obtains. The problem, in Daisy’s case, is not that her merits are not about her, but rather that she is only partly responsible for her merits. More to the point, when we compare Daisy to another student who works just as hard and who is equally talented, but who is less meritorious than Daisy, we find that Daisy is not responsible for the additional merit that she possesses due to her advantageous upbringing. Thus, we conclude that she is not more deserving of admissions than the less advantaged student, despite the difference in merits.

Consider another of Mulligan’s examples which is intended to illustrate the stronger sense of aboutness that he appeals to. Suppose that Veronica, as a result of her being mind-controlled by Anna, commits some crime. Intuitively, Veronica does not deserve to be punished. For Mulligan, the reason, again, is that the aboutness principle is violated, as the crime is only “‘about’ her in some weak sense.” (2018a, 72) But the point is that we know that Veronica does not deserve punishment without having to consider the metaphysics of mind control (and thus whether Veronica’s crime is her act, or whether it is “about” her in the relevant sense) because she is clearly not responsible for the crime, and thus does not deserve punishment.

Elsewhere, Brouwer and Mulligan (2019) mention (but do not commit themselves to) the possibility of understanding the aboutness constraint in terms of the responsibility constraint—perhaps being responsible for a desert basis is necessary for the desert basis to be about the subject in the relevant sense. Obviously, that kind of move does not help with the luck objection, since it would simply accept the responsibility constraint. But this proposal is ill-advised anyhow. Properly understood, neither the aboutness principle nor the responsibility constraint entail the other. Just as there can be facts about someone that they are not responsible for (their birthdate, e.g.), there can be facts that they are responsible for that are not about them (the disappointment of their parents or the applause of the crowd, e.g.).

It seems, then, that in trying to explain and motivate the connection between desert and fair equality of opportunity by appealing to the aboutness principle, the examples simply provide further motivation for adopting the responsibility constraint. Intuitions concerning the desert of punishment, furthermore, are best explained by the responsibility constraint, and are not well explained by appeal to the aboutness constraint. Finally, the responsibility constraint entails the luck constraint, and so provides an account of the connection between luck and desert. In sum, there are very good reasons for thinking that the responsibility constraint is central to a proper theory of desert, and a meritocratic account of distributive justice. But in that case, the appeal to the essentiality of origin misses the deeper point of the luck objection: even if we grant that one’s origin and innate talents are not a matter of luck, we are not responsible for them. Thus, given that the development of our merits is causally dependent on our innate talents, the extent to which our merits ground our deserts is weakened. And indeed, they are weakened just as much as if having our innate talents were a matter of luck.

Mulligan might have a response here, which I will consider before concluding the paper. Part of his response to the initial Rawlsian luck-based challenge to meritocratic conceptions of distributive justice is that the challenge often involves some claim to the effect that no one deserves their talents, their character, or even their efforts. Mulligan argues that such claims are “not just false but unintelligible” (2018a, 173). His explanation for this appeals to his particular conception of persons, on which one’s innate talents, deep character traits, and things like work ethic, are constitutive of one’s identity. Thus, he argues, it doesn’t even make sense to ask whether one deserves those things. Perhaps too, then, my claims to the effect that we are not responsible for our talents are similarly unintelligible, and thus can’t be used as premises in my arguments. In that case, I could not argue that the appeal to the essentiality of origin, even if it helps with the luck objection, fails to deal with the deeper, responsibility-based objection.

But Mulligan’s argument here is too quick. First, it is not at all clear to me that any of the relevant claims, either involving desert or responsibility are unintelligible. Indeed, the very fact that Rawlsians have made such claims and readers have apparently understood them is good linguistic evidence that the claims are intelligible. Second, there is the question of what unintelligibility comes to, and thus why it matters. For instance, does it follow that the claims have no truth value, or are they false? Some answer is needed here to argue that the claims cannot be used as premises in arguments. Third, what is the cause of the unintelligibility? Clearly the claims are syntactically well-formed, and they do not seem like paradigm examples of semantic anomaly like ‘Mary is half tall’, ‘Mary began the rock’, or ‘colorless green ideas sleep furiously’. Without a proper explanation of the cause, it will be difficult to know what the kind of unintelligibility is, or why it should matter if the claims turn out to be unintelligible in some sense or other.

Mulligan’s explanation for the linguistic badness of the relevant claims seems to be that they express propositions which are metaphysically impossible, because one cannot possibly deserve (or perhaps be responsible for) traits which are constitutive of who they are. This might be enough to make the claims “puzzling”, but it does not make the claims meaningless or unintelligible. It seems obviously true to me, for instance, that we are not responsible for our innate talents—indeed, it seems metaphysically impossible that we are responsible for our innate talents. But it is also metaphysically impossible that water consists of copper or that two plus two equals five, but nevertheless, “water consists of copper”, and “two plus two equals five” are perfectly intelligible. When one makes such claims, we answer immediately that they are false. Without further elaboration, then, it’s not clear that the relevant claims are unintelligible, or even if they are, how their unintelligibility bears on the argument I’ve made in this paper.Footnote 23 Indeed, if the point is that it is metaphysically impossible to deserve one’s innate talents, then this is a conclusion I readily welcome, and does not conflict with anything I’ve said in the paper.

In conclusion, then, the appeal to Kripke’s essentiality of origin thesis does not rescue meritocracy from the luck objection. Even in an idealized, meritocratic society of the sort Mulligan envisions, the contingency of the market means that it will be lucky to have the kinds of talents that will more easily translate to merit. Furthermore, the real threat to a meritocratic conception of distributive justice is not merely that, due to differences in innate talents, the accumulation of merit is lucky, but that we are not responsible for the differences in merit that result from those differences in talent. But in that case, the extent to which our merits ground our desert of income and other economic goods is lessened, and so too would be the demands of distributive justice.