Where Every Ruthless Cutthroat is Above Average

Scott Alexander at SlateStarCodex recently put up a post asking what was so bad about meritocracy:

The intuition behind meritocracy is this: if your life depends on a difficult surgery, would you prefer the hospital hire a surgeon who aced medical school, or a surgeon who had to complete remedial training to barely scrape by with a C-? If you prefer the former, you’re a meritocrat with respect to surgeons. Generalize a little, and you have the argument for being a meritocrat everywhere else.

The Federal Reserve making good versus bad decisions can be the difference between an economic boom or a recession, and ten million workers getting raises or getting laid off. When you’ve got that much riding on a decision, you want the best decision-maker possible – that is, you want to choose the head of the Federal Reserve based on merit.

This has nothing to do with fairness, deserts, or anything else. If some rich parents pay for their unborn kid to have experimental gene therapy that makes him a superhumanly-brilliant economist, and it works, and through no credit of his own he becomes a superhumanly-brilliant economist – then I want that kid in charge of the Federal Reserve. And if you care about saving ten million people’s jobs, you do too.

As Scott sees it, the problem isn’t with meritocracy per se, but with a kind of sham credentialism that only poses as meritocracy. When we force people to study British novels for four years before we let them be doctors, or give ourbest jobs to dutiful toadies who pad their resumes with silly clubs and extracurriculars, or tolerate a system that hands out plum appointments to the guy who’s friends with the boss’s son, that’s not true meritocracy at all. In fact, all this waste and corruption is the main impediment to genuine merit-based advancement. So when people complain about this sort of ersatz meritocracy, they’re implicitly praising genuine meritocracy.

I’m skeptical.

Scott’s argument creeps awfully close to that old trick of political philosophy where you say something like, “Communism just means a fair distribution of resources–if you think resources should be distributed fairly, congratulations, you’re a communist.” Or, “Liberalism is about granting freedom to individuals–if you’re an individual and you like being free, congratulations, you’re a liberal.” Or, “Patriotism just refers to the natural human affection for one’s homeland–if you feel an attachment to your home, congratulations, you’re a patriot.”

In each case, the definition transfers the burden of argument onto a word too vague to support it. What do you mean by fair? What do you mean by freedom? What do you mean by homeland?

And so it is with meritocracy. The whole problem is that people disagree about what “merit” means.

Take Scott’s accomplished surgeon. Let’s tweak the example. Would you rather have your surgery done by a brilliant surgeon who’s also a violent white supremacist in his spare time, or by a mediocre surgeon who’s a wonderful person? The honest answer, for most people, I suspect, is: “I would like to have my surgery done by the brilliant surgeon, and I would like not to know that he’s a violent white supremacist in his spare time.” Or possibly, “I would like to have my surgery done by the brilliant surgeon, and then I would like for him to be fired because he’s a horrible person.”

What about Scott’s hypothetical head of the Fed? Suppose it comes out that he beats his wife and molests his children? Do we still let him be chairman, because he’s just so darn good at it? Should we throw him in jail, but ask him to keep on making economic decisions? Or do we quietly choose to ignore his treatment of his family, because ethics are a bitch and if you save ten million jobs, that will probably result in less domestic abuse in the long run, and mumble-garble-arble-utilitarian-ends-justify-the-greatest-good-for-the-something-or-other? Or do we convince ourselves that no child molester could possibly be a good economist, because to be a smart Fed chairman you also have to be a decent family man?

And surgeons and economists, those are the easy cases! What does it mean to identify the best Congressperson for the job? Is she the one who knows the most politics? The one who’s most loyal to her constituents? The one with the most party loyalty? The craftiest negotiator? The most charismatic campaigner? The one who does the most legislating, whatever exactly that means? The one who happens to champion policies that you think are good policies? The one who in some way has made the world a better place fifty years down the road? How can we even determine that?

What about the best cop? Most busts? Toughest warrior? Liked by the community?

What about jobs like day-care worker, where the big concern is weeding out horribly unqualified candidates–like the ones who might hurt children–not making finicky, fine-grained distinctions among candidates whose main qualifications are personality traits that are notoriously hard to test for, such as patience and decency and conscientiousness?

Best CEO? One who maximizes profits? Treats employees well? Adopts green policies? Spurs innovation?

Best teacher? If you know the answer to that one, please tell me, because I’m not sure I’ve ever met two people who hold the same opinions on the subject.

It’s awfully easy to throw out words like “merit” and “qualifications” and assume that people agree about what they mean. A lot of the commenters on Scott’s pose are computer engineers, and they seem to collectively assume that computer programming is a field in which metrics of merit are clear and well-established. “Can you code? Great, take this test. Ace it, and you’re hired.”

Maybe that’s true. I don’t doubt that merit is relatively well-defined in that field. But I still wonder. If we asked programmers what makes a good programmer, and then asked the wider public what they want from programmers, would the definitions agree? Should the insider definition take priority? Should we rely on professionals to establish all their own standards? Why? Don’t we want to get some public benefit out of our programmers? Isn’t that, in a way, the point of things like markets and democracies, to give broad populations of nonspecialists ultimate control over standards of merit? If someone’s a brilliant programmer, and all the other programmers agree that he’s one of the best around, but he spends all his time writing programs that let a couple of day traders screw over small investors, is it fair for the public to say, “We don’t care how talented that dude is; he sucks”?

For that matter, suppose we took a bunch of programmers who absolutely agreed that meritocracy is great and that the best programmers ought to get the best jobs and so on, and then asked them to spell out in detail exactly what they meant by merit and what qualifications and qualities they thought programmers ought to have? Would they all agree? My guess is that they’d agree on a few core traits and principles, but argue incessantly over a broader set of secondary characteristics. “Programmers ought to be really smart and have a solid grasp of basic math, abstraction, and logic.” “Agreed, and they also need to extremely independent.” “No, it’s better for them to work well in teams.” “But they definitely need to know calculus.” “No, no, better for them to be good with language; that’s too rare a skill in comp-sci fields.” “What’s really important is that they know [whatever the latest trendy programming language is].” “Please, that’s just a passing trend. A programmer with deep knowledge of a few old-school languages can pick up a new one, easily.” “What’s really important is for a programmer to be a neat and detail-oriented, follow best practices, comment frequently …” “No, I’ve worked with people like that; they’re all hidebound college grads who want to be micromanaged and earn gold stars … Give me a big-picture thinker and I’ll teach him my own best practices …”

And so on. I would bet it’s the same for surgeons, or economists, or any group of specialists who can be broken down into groups of subspecialists with slightly different values, and again into smaller groups with still other values, and finally into throngs of idiosyncratic souls, each of whom operates according to complicated sets of private principles. Once you start shaking that word, “merit,” all kinds of complicated assumptions fall out.

And that’s where the debates about meritocracy come from. It’s all well and good to say, “Let’s put the best people in charge,” but what the heck does that word “best” really mean? Even if we can decide what it means, how can we be sure we’re using the right measurements? The right systems for sorting and selecting candidates? How can we ensure we’re punishing bad traits as well as promoting good traits?

It isn’t just that any real-world meritocracy is an uncontrolled social experiment in the limits of applied measurement theory. It’s that any truly meritocratic system comes with an intrinsic flaw.

Meritocracies are inherently competitive: the whole idea is to develop a system for measuring and ranking people, then give more social power to those with higher ranks. No matter how you tweak, skew, shape, slice, or dice your particular system–no matter what qualities you choose to measure–this will always be true.

Consider a hypothetical meritocracy, one very much like our own. Let’s assume that each job in our meritocracy requires different innate talents. To be a top-ranked surgeon, it helps to have:

High IQ


Ability to focus

Taste for competition



To succeed as a politician, one needs:


Taste for competition



To succeed as an athlete, one needs:

Strength, agility, endurance

Taste for competition



To succeed as a writer, one needs:


Verbal fluency

Taste for competition


You get the idea. We can try to minimize the extent to which a taste for competition is rewarded, but I don’t see how we’ll ever eliminate it as an advantageous attribute in a system that is by design highly competitive. No matter what sorting system you put in place, folks still have to show up and take tests, put in practice, pump their percentages, go through the grind. Get a bunch of people who are closely matched in other abilities, and drive to compete will become a decisive factor. That’s to say nothing of the likelihood that any human sorting system can be at least partially rigged or gamed, and that in a pool of uniformly superlative candidates, those adept at rigging and gaming systems will have a natural advantage.

Which means that over time, a meritocracy starts to churn up an elite class of people whose most notable shared trait is a penchant for coolly calculating self-serving competitiveness. Over in one corner you have a bunch of mental lightweights who are nevertheless great at singing and acting and looking good in underwear. In another corner you have some people who happen to be great at running or throwing small objects. You have some not-too-bright folks who are nevertheless super-talented at charming and manipulating people. You have a few eccentric psychopaths with extraordinary technical gifts. You have surgeons who go home and abuse their spouses. You have a whole lot of people who are good at “leading,” whatever the hell that means.

You have, in short, a class of people with a whole host of gifts and talents and eerie natural endowments, who nevertheless are likely in the aggregate to be characterized by one salient trait: a burning desire to succeed at the expense of others.

How are all these people going to get along? Can they get along? What do they have in common? What defines them as a class?

The superhumanly cunning CEO takes a lunch date with the superhumanly risk-tolerant hedge-fund manager who brings along his superhumanly intelligent top quant. At the club they run into a superhumanly charismatic state representative who introduces them to her superhumanly conscientious aide and the superhumanly talented actress who donated to her campaign. This is a lucky break for the CEO, whose company is negotiating a licensing agreement with the superhumanly fleet-footed athlete whom the actress met through her superhumanly loquacious publicist–and for the hedge-fund manager, who wants to make connections with the superhumanly creative journalist who’s been putting together a profile on the representative. At the end of this tiring day, the CEO goes home to the superhumanly beautiful model who recently became his third wife, the representative goes home to the superhumanly dogged academic who’s currently her husband, the others go home to their superhuman surgeons and artists and machers and accountants, and they all talk about the superhuman accomplishments they’re planning for the next day.

How does it all work? How can these rare, hypercompetitive, and utterly dissimilar superhumans function as a coherent elite? How can they make the connections they need to make, maintain the relationships they need to maintain, without emphasizing the few experiences and traits they have in common? What gives them a class identity if not a shared fascination with narcissistic self-promotion?

And this is the problem I think we have now, or the problem we’re developing. It’s not that meritocracy doesn’t work. It works too darn well. The longer and better it works, the more likely you are to develop a ruling class of self-obsessed individualistic strivers. And what’s scariest is that there’s no good solution. To solve this problem, you have to actively penalize personal ambition. And that cure sounds worse than the disease.

This entry was posted in Uncategorized. Bookmark the permalink.

Leave a Reply