Having finished Sher’s Desert last week, the MTSP Discussion is on to discussing HLA Hart’s The Concept of Law, but I’m going to spend the next few weeks hammering out summaries of the last four chapters of Sher’s book, just for the hell of it. I’ve had to break my discussion of Chapter 7 of Desert into two parts, a summary and a critique. This post is the summary; I’ll post the critique when I get a chance.
Chapter 7 of Desert discusses a so-far neglected basis of desert, merit. It seems self-evident or obvious to many people that we deserve things insofar as we have or exhibit the right kind of merit, whether moral or non-moral, to do so. Chapter 7, “Merit and Desert,” discusses contexts where moral and non-moral considerations merge in ways that are hard to entangle. Take for instance the common claim that college admissions be based on candidates’ “merit” with respect to admission. Is that a moral claim or a non-moral one? Does it involve a moral conception of merit or a non-moral one?
Obviously, it depends on what one means by “moral,” and what one means by “merit.” In one sense, the “most meritorious” candidate for admissions is the one who has the greatest ability, right now, to excel at college-level work, even if that ability came about in large part through an etiology for which the candidate deserves little credit, e.g., native endowments and/or fortunate circumstances. In another sense, the most meritorious candidate might be the one who, right now, requires remedial education to be brought up to speed to do even introductory-level college work, but who, having faced duress or adversity en route to the admissions process (and done so with grace and tenacity), might be thought deserving of admissions either in virtue of that past accomplishment, or in virtue of the potential for excellence that the candidate’s tenacity represents (or both).
The first candidate seems to exhibit mostly (but not exclusively) non-moral merit; the second candidate seems to exhibit the reverse (subject to the same qualification). Further, each candidate might excel at something, but excel at different things, differently relevant to “success in college.” So there’s a fair bit of complexity involved even in garden-variety cases of “deserving because of merit.” It’s an open question whether Sher ends up doing justice to all or most of these complexities. (Chapter 8, on desert and virtue, focuses more specifically on moral considerations.)
In any case, the puzzle we face is why the meritorious deserve to be rewarded, and by implication, what constitutes an appropriate reward for the merit they exhibit or possess. Sher starts by considering two initial possibilities, rejects them, and then offers his own independent account.
One initial possibility is that reward is a kind of mirror-image of punishment. If punishment is (as a previous chapter/post suggested) a negatively-valenced instance of diachronic fairness (DF3), then reward is a positively-valenced instance of that same concept. Punishment, Sher had argued, restores the proper balance of burdens and benefits in an organized society; the offender puts this balance out of alignment through wrongdoing, and thus comes to deserve punishment as a means of putting the prior balance back into alignment.
In other words, it’s tempting to think that the display of merit or excellence pushes the balance or burdens or benefits “out of alignment,” but in the reverse direction from wrongdoing; in displaying excellence, the meritorious person assumes too large a burden, and foregoes too many of the benefits due him. So as the offender in effect free rides on the system of liberty, the meritorious person over-contributes to it. The free rider therefore has to be punished; the over-contributor, rewarded.*
There are many problems with the preceding proposal, but the simplest objection is that merit and excellence are not plausibly conceived of as supererogatory assumptions of excessive burdens or self-sacrificial forswearings of the benefits of a system of liberty. There’s no plausible sense in which merit or excellence surpass or exceed the norms that govern us, or put any prior moral balance out of alignment. So the reward of merit is not an instance of DF3, and can’t be accounted for by it.
A second possibility, expressed both by Sidgwick and by Feinberg, is that reward-for-merit is a kind of “gratitude universalized.”** When people display merit, we naturally feel gratitude for their effort or achievement, and it’s natural for the grateful to want to express their gratitude by rewarding the objects of that gratitude. When we generalize our gratitude from our particular or personal circumstances to all those who generally elicit it, we get the principle of reward-based-on-deserved-merit. Thus “gratitude universalized.”
There are many problems here as well, but the simplest one is that the account gets the conceptual cart before the horse. We can see this by asking: who ought, justifiably, to get our gratitude? Unless we trivialize the concept of gratitude, the answer can’t plausibly be “anyone who does anything that we might ever regard in a favorable light.” Gratitude is subject to normative standards. Hence there is justified and unjustified gratitude. We feel justified gratitude when the object of that gratitude has done the kind of thing that makes gratitude a fit or appropriate response to it. But that’s just a long-winded way of saying that justified gratitude presupposes a judgment of desert. By contrast, a person’s desert or merit cannot be certified by the fact that we feel gratitude to him; we could in principle feel gratitude for almost anything. In any case, a person might deserve something and elicit no gratitude for it at all.
If so, gratitude can’t be the ground of desert. Things are the other way around: desert is the ground of gratitude. We therefore can’t invoke gratitude to explain why the meritorious deserve reward; we need a prior account of why the meritorious deserve reward, from which we infer that gratitude is among the rewards they may deserve.
So much for the initial proposals. How to proceed?
One route to the solution of our problem is to focus less on desert than on reward. On reflection, reward involves two separate elements, one assertive and one expressive. The assertive element acknowledges the merit of the meritorious person according to some objective standard, and thus expresses a virtue of veracity. When we acknowledge that the person satisfies the relevant standard, we are, simply, acknowledging the truth about her. The expressive element presupposes this prior acknowledgement of merit, and seeks to respond to it in an appropriately fitting way, expressing what Sher regards as an obligation of admiration or gratitude for the merit.
It’s tempting, in light of Sher’s rejection of the Sidgwick-Feinberg account, to assimilate merit-based-desert entirely to veracity. In other words, truthfully to acknowledge that someone satisfies a standard of merit entails by itself that we reward her. To fail to reward her, on this line of argument, is to fail to be truthful about her–to fail to acknowledge the truth that she satisfied the relevant standard. But this won’t work. For one thing, obligations of veracity are generally negative rather than positive in form: we have a negative obligation not to offend against the truth by lying or misrepresenting it, not a positive obligation to sail the seven seas looking for truths to express out loud. So our obligation to reward the meritorious will always outrun our duty of veracity about them. Put differently, a negative obligation of veracity would yield an obligation far too weak to account for our duty to reward the meritorious.***
More to the point, an obligation of veracity bears no important connection to reward at all. I can wholeheartedly acknowledge your merit and still (justifiably) do nothing about it, whether through incapacity or unwillingness. Consider in this regard how often we encounter panhandlers on the street (or other petitioners for help in other contexts), and decline to help them. A panhandler might tell you the story of her misfortunes, and genuinely convince you that she deserves a helping hand. You might acknowledge the truth of her story and her claim of desert, and yet decide not to lend her a helping hand, whether because you couldn’t afford it, or despite the fact that you could. Acknowledgement of her merit does not, by itself, entail action of any kind.
If this seems hard-hearted, at least in the context of American cities, it simply expresses the brutal truth about cities in the developing world. No one could simultaneously preserve their sanity (or stay solvent) while rewarding every deserving beggar one encountered on the streets of Delhi, Lahore, Hebron, or Managua. But I doubt one could do so even on the streets of New York, Philadelphia, Baltimore, or Bloomfield, New Jersey. Even if you heard every story, weighed the evidence, and acknowledged the truth about every deserving person’s situation, you wouldn’t, by that fact alone, have an obligation to pay them a cent.****
We’re still left trying to account for reward. At this point, Sher either sells out, moves the goalposts, or bows to the inevitable, depending on your point of view: he invokes convention. It’s an obvious social fact that merit is defined in part by conventional standards that regulate conventionally-structured activities.
Such conventional structures are of course extremely varied. They range from the rules and bylaws of the National Football League to the arrangements for impromptu footraces to the regulations and criteria the govern the assignment of grades in universities. But despite their differences, these structures all involve conventions that are public knowledge within the relevant communities. These conventions include rules that specify what performers may and may not do; criteria for success and failure in performance; and procedures for awarding prizes, titles, or grades on the basis of that success or failure. Like linguistic rules, the conventions invest brute occurrences with significance they would otherwise lack. In particular, where the conventions exist, the awarding of any prize, title, or grade is a conventional way of signifying that the recipient has satisfied the appropriate criteria of performance….(Sher, p. 115)
Barring special grounds for doubt,” Sher continues, “it would be unreasonable not to believe” that conventional criteria for merit deserve both recognition and reward.
In effect, then, the meritorious deserve reward because conventions are in place to identify merit and reward it. Sher’s presumption is that these conventions are fundamentally sound. Given this, reward-for-merit involves a conjunction of two obligations: a duty of veracity with respect to the satisfaction of the conventional criteria; and a duty of fidelity or gratitude in granting the reward conventionally attached to the display of the relevant sort of merit (Sher, p. 115).
Granted, the presumption in question requires justification, and can in principle be defeated. But given the fundamental soundness of the prevailing conventions of merit (Sher thinks), we can accept a reasonable division of labor without falling into undue skepticism about our conventions (Sher, p. 117). In other words, barring special grounds for doubt, the conventions of merit can essentially be taken for granted; it’s a separate question (involving a separate inquiry) how to justify those conventions, and what form that justification will take. But our not having such a justification in hand is not, Sher thinks, a problem for his account. Too many of our conventions are too obviously sound to make global skepticism about them reasonable or plausible.
That said, Sher acknowledges that the preceding account will not handle every case of merit-that-deserves-reward. It won’t, for instance, handle one of the clearest paradigms of such a case, i.e., hiring-on-the-basis-of-merit (or generally, accepting-into-a-coveted-position-on-the-basis-of-merit). Here, our moral intuitions clash pretty sharply with convention. There’s a long-standing moral intuition to the effect that a hiring officer (or admissions officer) ought to hire (or admit) the best-qualified of the candidates for a position. But convention does not consistently reflect this, to put it mildly.
Consider the number of “exceptions” to the “rule” that we ought to hire “the best qualified candidate.” Many job candidates (or other sorts of candidates) are hired or admitted on the basis of seniority within the organization, past promises, nepotism, need, pity, externally-imposed diversity requirements, compensation for past injustices, blind faith, hunches or whims, predictions of future promise (whether justified or unjustified), or other, extraneous, non-merit-based considerations (at least as Sher conceives “merit”). Family businesses hire on the basis of family connections. Universities admit legacies. Many institutions hire job candidates exclusively within their social network, simply because they are in that social network.
This last example suggests that being a known quantity can become a substitute for merit simply because it’s easier to detect than merit. Practical exigencies work this way pretty generally in the labor market. Sometimes a business will hire someone–anyone–simply to fill a position, less because it’s known that the person merits the position but because the position has to be filled, and filled quickly. And then there are committee hires, where the basis for the hire is a horse-trading compromise between the members of the committee rather than the expression of any determinate principle, much less the principle of deserved-reward-for-merit. (For an informative account of the contemporary job market, see Steve Dalton’s The 2 Hour Job Search.)
Even in cases where hiring or admissions nominally takes place on the basis of merit, epistemic considerations suggest that there’s a sizable (and probably ineliminable) gap between promise and delivery: it’s very rare that a hiring officer can really know, on the basis of, say, a resume, a LinkedIn page, an interview, and a few phone calls to references (each maybe 30 seconds long), whether or not a person truly “deserves” to be hired on the basis of merit. The fact is, convention dictates that hiring is widely regarded as a gamble in which the candidate’s merit is discovered less by the procedure used to hire her than through the test of the time the person spends actually doing the job. That’s why so many businesses operate by employment-at-will, which makes it easier to boot out those who lack “merit” however defined, or whether defined at all.
So unless we simply assimilate these conventions to our account of desert, our norms of merit, desert, and reward can’t be conventional all the way down. When it comes to hiring ab initio, Sher admits that it isn’t sufficient to say
the best-qualified applicant ought to be chose because existing rules dictate his selection. Rather, what is meant is that the rules ought to dictate his selection–that he has a significant prior claim to be chosen. Thus, the desert of the best-qualified appears to have some further (and more strongly pre-conventional) grounding (Sher, p. 119).
One possible answer is that the best-qualified candidate ought to be hired because the best-qualified candidate is most likely to exhibit the best performance in the relevant position, making it instrumentally rational for the hiring or admissions officer to hire or admit him. Though Sher grants that it is instrumentally rational, its being instrumentally rational for the officer to hire or admit doesn’t explain why the candidate deserves to be hired or admitted, i.e., why the officer has a self-regarding duty of rationality to himself and/or to his institution, but an other-regarding obligation of justice to the candidate.
To break free of both convention and egoistic rationality, Sher invokes a broadly Kantian sort of argument (expressed in somewhat Dworkinian language): hiring by merit–hiring the most qualified applicant for a given position–is an instance of respecting the personhood of the candidate, of taking him seriously, of ensuring that his agency matters.
When we hire by merit, we abstract from all facts about the applicants except their ability to perform well at the relevant tasks. By thus concentrating on their ability to perform, we treat them as agents whose purposeful acts can make a difference in the world. Moreover, by concentrating on abilities that are intrinsically connected to jobs and opportunities that exist to serve independent purposes, we affirm the applicants’ involvement in the wider life of the community. For both reasons, selecting by merit is a way of taking seriously the potential agency of both the successful and the unsuccessful applicants. Conversely, when an applicant is selected on some other basis, there is a recognizable if elusive sense in which he and his rivals are not taken seriously. And this suggests that we may justify selection by merit by arguing that persons ought to be taken seriously in the relevant sense (Sher, pp. 121-22).
Properly understood, the requirement that we select among applicants on the basis of their qualifications is a consequence of the more general requirement that we treat all persons as rational agents….Thus, any requirement that we treat persons as rational agents must demand respect not only for their choices, but also for their ability to do things which advance their own and others’ ends….[W]hen we select among applicants for reasons other than their ability to perform the tasks that define positions, we treat them as passive recipients or largesse or links in causal chains rather than as active contributors to anyone’s ends. by thus disengaging their practical wills from the aims that have generated the positions, we violate the requirement that they be treated as rational agents (Sher, p. 126).
That’s Sher’s view of merit-based desert in a nutshell. The view is essentially disjunctive: rewarding based on merit is either based on justifiable conventions, where we have obligations of veracity to identify cases of merit, and obligations of expression (gratitude, fidelity) to reward on the basis of conventions that demand it; or, where appeal to convention is not available, a matter of rewarding merit because doing so is an expression of respect for persons or rational agency, and failing to do so would disrespect persons or rational agency. In my next post, I’ll discuss some objections to Sher’s account.
*However absurd this account seems, it seems to me actually exemplified by Ayn Rand’s conception of the “Pyramid of Ability” in Atlas Shrugged, which might with complete accuracy be described as a “social hierarchy of non-moral merit, in which those at the top enjoy a privileged position while granting the unearned to their inferiors.” Rand systematically conflates moral and non-moral conceptions of merit throughout her works. I’ll save a fuller discussion of Rand and the Pyramid of Ability for a separate post.
***This strikes me as a basic defect of Leonard Peikoff’s discussion of desert in Objectivism: The Philosophy of Ayn Rand. Having assimilated our obligation of desert to a form of veracity (pp. 276-86), Peikoff then argues that reward is a matter of adherence to the trader principle (pp. 286-91). But he offers no explanation of how the two halves of his account interact. Do our judgments of merit determine who ought to be our trading partners? What if we’re obliged to trade with people who lack merit? Can we judge the merit of every trading partner we engage?
****This latter implication seems to me to contradict the point of Peter Singer’s famous “drowning child” thought experiments. Without saying so explicitly, Singer seems to be suggesting that acknowledging the truth about the drowning child entails an obligation to help the drowning child. I agree with Sher rather than Singer in this respect: even if we have a duty to save the drowning child, the duty to rescue the child is not a direct implication of the sheer acknowledgement that the child is drowning. One of the serious argumentative and ethical defects of Singer’s thought-experiment is to make this obvious logical point harder to make than it ought to be. But the fact remains: “X is drowning” does not entail “I ought to save X from drowning,” even if I ought to save X from drowning.