On the random origin of Merit

Deserve's got nothin' to do with it - Unforgiven

1.

The meritocracy debate is one of those issues. You know, the ones that never end, that seem like they can’t be resolved, ever. Euthanasia, gun control, Israel-Palestine and the like. The issues that seem like we need them to exist more than they need us to solve them. We sure act as if we need to agree on a solution, but everyone’s secret conviction is that we’ll fight over it for show but in the end, nobody’s opinion will be scathed.

But, in a way, that debate is more fundamental. It basically tries to answer questions such as: “are people who have a good life worthy of it?” “Are people with bad outcomes unworthy of them?” Most other debates can be boiled down to the question: “who deserves what?”.

And yet every discussion on this topic ends up going nowhere (just look at the nigh-infinite comment section of Scott Alexander’s blogpost) 

That’s because meritocracy is everyone's favorite double-edged sword. It's easy to see why: if it can be made to work, almost anyone would be all for it. But it doesn’t work, it can’t work according to critics, because there is no decoupling an individual from his environment. There is no way to distinguish his achievements from the culture which produced them and him.

This is the essence of every critique of merit that I found, from “The Meritocracy Trap” to “The Tyranny of the Meritocracy” and “Meritocracy and Economic Inequality”. From Michael Sandel to Freddie de Boer. Some people go even further and deny that Merit is a thing at all. Rawls (A Theory of Justice, 1971) is the epitome of this. As is explained in “Meritocracy and Economic Inequality” “Innate talent and socioeconomic background are equally arbitrary forms of luck, which in themselves merit no reward.”

The cardinal sin is always the same: there is no way to evaluate merit and any attempt to do so is yet another way for the elites to keep their power. 

Unfortunately, that accusation brings us nowhere close to a resolution because it’s extremely easy to make, no matter where you stand on the issue.

A typical debate goes like this:

-"I got into this college based on merit. You got it based on affirmative action and quotas."

-"Quotas are just there to give me a minimum fighting chance, I still beat everyone else that's from my background in the entry exam. You only got this because your parents are rich."

-"I beat a lot of rich kids too and besides I had great grades and that's an objective measure."

-"It's easy to have great grades with all the private tutors your parents pay for. They've been training you for that type of exam for years. I had to study on my own."

-"I have a high IQ since childhood."

-"If IQ is genetic then that's on your parents and you didn't deserve them. If it's your environment, then you didn't deserve that either."

-"What part of me did I deserve? What part of you do you think you deserved and got you here?"

There’s an uninteresting component to this debate and that’s when corruption is clearly at the core of the “Merit” system. It’s when no-one -not the corruptor or the corrupted- actually believes in the meritocratic system. For example, when your teacher gives you good grades because your parents are important. It’s not what interests me here.

The critics of the books and blogposts I’ve mentioned have this in common: they reject the very notion of merit (either practically or even theoretically) and think scandals of corruption at universities obfuscate the issue because they make it seem like the only problem is getting meritocracy right. Whereas, to their mind, it’s the whole concept itself that’s unfair.

The core problem is philosophical in nature: there seems to be no consensus on when, if ever, is a reward for effort actually warranted and legitimate. It’s because of that vacuum of a definition, that all these voices have risen in the past decade.

The whole point of this post is to insist that it is in fact possible to pinpoint exactly what most people intuitively mean, when they say that someone deserves something. And further that this notion is essential for the development of our societies.

Why does it matter? Because, as I said earlier, almost all current debates can be boiled down to a question of merit:

-Do trans athletes deserve their win?

-Do rich capitalists deserve their money?

-What punishments do criminals really deserve?

-Do smart people deserve their high intelligence/IQ/high grades

-Should Overweight people pay more on airplanes?

2.

Let’s start with a simple case where merit is to be assessed all the time, without too many tears: competitive sports, for example, boxing. I will use this as an analogy throughout this post.

You have a group of people (mostly fit) of which you know everything (physical characteristics, background, psychological tests etc.) You want to know who is better at knocking out opponents according to standard boxing rules.

You have no idea what makes a good boxer, so your data is useless for now (by you). Your best bet is therefore to have everyone duke it out and repeat, tournament after tournament, until you start seeing some patterns.

If no pattern emerges, someone random comes out on top every time, it obviously means that boxing -as a selection process- doesn’t correlate to anything remarkable in human characteristics. Of course, that’s not what happens in boxing.

·       After a while, you start noticing that huge guys almost consistently win over smaller guys. The competition becomes uninteresting, because you often can tell who’s going to win if the weight difference is too big.

·       So, you solved the problem of who wins when physiques are too different. But not if they are equivalent. Next step: create weight classes to filter out that selection tool. Basically, you say, winning because of your weight is unfair! There’s no skill involved. Let’s control for that.

·       Next you notice that people that had a lot of protein intake during their childhood and teen years consistently outperform others. So, you separate between diet classes.

·       Next you realize that some people have better access to healthcare and recover much better between fights. Also, unfair. You control again for medical classes.

Finally, your analysis becomes so good (and remember, your data is perfect) that every single person is in a class of his own. You can always tell who will win a fight before it begins, based on a complex combination of factors you devised. Your competition is boring now and you forget about it. But your new project, based on that experiment is creating the ultimate fighter since you now know everything that factors into it and democratizing the process, so everyone has access to that!

What I hope this metaphor has made clear is that Merit is relevant precisely when we are ignorant. It is, in fact, the way in which we get information about what factors into which skill. And that’s extremely valuable information, especially if we can control for it and replicate it.

The next section expands on this point.

3.

Here’s a typical comment which I think stems from a misconception about competition:

“That guy didn’t deserve his win. He had an undeserved edge because of rich background/luck/sun wasn’t in his eyes etc.” The competitors were not on “equal footing” and so the competition was “unfair”.

Of course it was! That’s the whole essence of competition. People are not clones. Even clones are not entirely identical. The whole point of competition is that there is some difference between the two (there always is) but we’re not sure -beforehand- how those differences will express themselves in skills.  We already knew they were not on equal footing because, in the end, one will win and the other will lose.

The only thing that truly matters is: can we determine the factors that will predict the outcome beforehand or not? If not, then the competition is interesting (that is, it gives us information). Otherwise, if the relevant factors are predetermined, then the way forward, is to equalize those factors, effectively filtering them out and see if something else can come into play.

What I am suggesting is that the concept of “fair competition” is meaningless. Someone always comes out on top. This criticism (unfairness) makes a fundamental mistake in thinking that there could be a fair competition on equal footing. But that is non-sensical because if they were really, truly on equal footing, the outcomes would always be a draw -by definition.

I will insist on this point because it can be confusing: there’s no such thing as a “fair competition” and if there was, it would be a bad thing.

The goal of competition (in sports, education, or society at large) is not and never was to have everyone be as skilled as everyone else. It’s to find out what makes someone better. There’s enormous value in that because if it can be reproduced, then everyone could benefit.

Think of the boxing example above. After we determine that weight classes matter and separate them accordingly, fighters can’t win based on their body strength alone anymore. Now, they actually have to develop other technical skills to best their opponents. This means that both heavyweights and featherweights have to improve their knowledge and techniques to continue winning. In fact, this benefits featherweights more than heavyweights. But no matter what new strategies are developed in each weight classes, everyone benefits from these discoveries, including newcomers, who can now train based on this new knowledge.

If we follow that logic, then the “Meritocratic system” works as follows:

·       While we don’t know why some people succeed, keep analyzing the competition process until we see a pattern and can start making predictions on factors that are key to success.

·       If some factors are figured out, try filtering them out by segregating classes or by disseminating them to everyone (obviously not easy to do outside of sports)

·       Start the competition again with these new parameters while keeping a measure of randomness between the new competitors. In other words, the rest of the parameters that aren’t filtered out should be kept to as “free” or “open” as possible to allow new patterns to emerge.

4.

 I haven’t precisely answered the question I initially asked (what is Merit), so let me do just that: A person deserves his achievements if the reasons for his success are not clearly identified and if there is no way to control them.

That may seem paradoxical, but it follows naturally from what I outlined above: When we know (and can control) exactly what allowed someone to win (even if it is down to a particular gene or physical characteristic that he has) there will always be a feeling that the success is not deserved.

The question really comes down to that last line in the fake dialogue I wrote above: “What part of me did I deserve?”

On this point, there is a recurring misconception that I saw in several of the books and articles about Meritocracy: people don’t deserve their parents. People don’t deserve their childhood environment or the economic state of their country. People don’t even deserve their genes as those are inherited as well. That’s why the meritocratic system is “unfair”.

But this is a non-sensical argument. There is no version of “You” that didn’t have these parents, this particular combination of genes, and this childhood. That wouldn’t be “You”. It would be someone else. This is the old Descartes-inspired vision of the self which implicitly supposes that you somehow exist before you were born, in the air, as a spirit.

That you, somehow, inherited this body and these genes and these parents.

But you are this body and these genes. It makes no sense to talk of “deserving” bodies and genes. There is no “You” outside of that.

Any merit-based argument can only be made sometime after childhood when some part of the person is already formed.

However, my point here is that, ultimately, this doesn’t matter. What you consider “You” as a fixed part and “You” as incidental is not going to affect how the progress of society functions.

As I suggested above, we need the lottery of birth (genes+ childhood environment) to create people who are different. That’s how we ensure that new knowledge and progress are even possible within the free parameters of competition – the ones we don’t understand and haven’t controlled for.

Another sports example that synthesizes this nicely is the Fosbury flop:

In the late 60s, the Olympic high-jump scene was rather hostile to the backwards flop jumping technique. Although it had been tried, most experts strongly doubted it’s potential. The dominating technique was the straddle jump.

Unfortunately, Fosbury was “not a naturally talented high jumper” (Olympics website) with the straddle. He found it complex to perform and experimented with various different techniques until he decided on the backwards flop which now bears his name.

After his gold at the 1968 Olympics, a revolution began in the sport and today almost no-one uses another technique. The last straddle jump was at Seoul 1988.

What if instead, we had forced everyone to be on equal footing? What if we had strict regulations that prevented any free styling? Well in effect, everyone would be worse off. Fosbury would have been unknown since he would be forced to use a technique he wasn’t good at. And everyone else wouldn’t have benefited from that revolutionary flop.

Notice that Fosbury didn’t know that this strategy was the best. He simply tried several of them after he found the straddle complicated.

This is what I meant in the last step of the meritocratic system. Once you’ve controlled and segregated competitors by classes, you need to allow for a random component within a class to discover new improvements.

In other words, you need arbitrary unfairness! This is the only way to, again, sort out patterns, from within a class! That’s why it makes no sense to talk of “equal footing” once you’ve controlled and filtered the classes with every measure you can determine. What is left, by definition, is something you can’t predict. And within that class, no further segregations make sense, at least for now. The talent of competitors is basically random and only further competition will highlight whether some other parameter can be controlled for.

Now someone might object that this conception of Merit might work for sports but in education, job markets or society in general it’s completely unfeasible.

There is also the fact that sports are not a right in the same way education is thought of as a right. Failing at a sport is not the same as failing at society.

That may be true, but my worry here is that we have stopped even trying to find out why people succeed and instead started blaming the system and declaring that a difference of outcome is proof of something going wrong. It’s not. Not necessarily. 

Without difference, there can never be any discovery or progress.

5.

I would like to summarize here because at this point, it looks like there’s a big IF in my views:

1-    On the one hand, when we are ignorant of the factors of success, we need arbitrary production of skills (meaning “unfair” competition i.e. by genes, socioeconomic background etc.) because that’s the only way to actually get the best results and thereby acquire information on how to achieve those results.

2-    On the other hand, you should kill that part of the competition by sharing the “secrets of success” with everyone IF you can.

3-    If you can’t, then segregate those who have access to the “secret of success” with those who don’t (create classes) and have each class evolve on its own.

A critic might say: “That’s all fine for sports but we know for instance that private tutors help a lot with grades. Except we can’t replicate that to other kids because their parents aren’t rich. And yet all kids compete -in the end- in the same class. Universities, job market, life”. Yes, well there’s your segregation process. It’s an indirect one but it has the same function. Remember I’m trying to describe how to system already works. I’m just stating that it cannot really work in another way if we want innovation and progress.

I understand however that there is a specific point of contention with the fact that rich parents can simply throw money at a problem until they solve it. But that criticism is one of the Family Unit as a basic building block of society, not of the concept of Merit. If you think that a parent shouldn’t affect his child’s education and the State alone should handle the opportunities of children, say so. But don’t hold that against the notion of Merit itself.  

I was always struck by how different the American Dream is perceived by Americans vs foreigners who dream of going to the US to start a new life.

In America, the dream is often summarized as : Everyone can make it. However, myself and most immigrants I know see that dream differently.

It’s more like: Anyone could make it.