When gatekeeping is legitimate

April 4, 2019 (2323 words) :: Some amount of gatekeeping in tech is valid, because not everyone is suited for every role. The problem is when the incentives are distorted by economic factors.
Tags: ideology, working-in-tech

This post is day 94 of a personal challenge to write every day in 2019. See the other fragments, or sign up for my weekly newsletter.

I’ve written before about gatekeeping in the tech industry, and why it’s generally a bad thing:

Gatekeeping is a way for those inside the gates to feel better about themselves. They can justify their inflated wages by pointing to how much better they are than those outside, citing failed answers to interview questions that they helped design. At the same time, it allows them to improve their own conditions at the expense of others, by keeping wages high through limiting labour supply rather than through organising with their fellow workers to ensure decent wages for all.

Today, I want to talk about an obvious counter-argument to the position above. Isn’t some amount of gatekeeping actually good, sometimes? If you’re working on a team with other software engineers, surely you want them all to be good at their jobs, right?

Now, it’s completely reasonable to have some sort of filter to screen out candidates for technical roles. Even if the filters that actually get used in modern-day tech companies tend to be really bad and biased, there is still merit to the underlying goal of finding people who would fit the role, in terms of expertise, communication skills, interest in the product, etc. It would be silly to take a random person off the street and expect them to do a good job of managing your Kubernetes cluster. Computers are complicated, and certain technical skills take a lot of time and patience to build, and at least some sort of aptitide or interest in the topic. A lot of the technical skills that are valorised today could probably be more widespread if educational access were improved, but they wouldn’t be universal. There are grades of difficulty, and just as widespread literacy doesn’t mean everybody is now capable of writing literary masterpieces, there will always be particular technical concepts that not everyone will be able to master. Some things will always be harder than others.

Expertise isn’t a bad thing, is what I’m saying. People who are making decisions about important technical architecture should be extremely qualified for their job - they should have some sort of prior experience, and a propensity to work hard, and even some innate talent. This is true from the perspective of anyone who has to work with them, or who has to interact with the systems they build. It’s the same deal as with civil engineers, or doctors, or pilots, etc - as a society, we want buildings to stay up, and surgeries to go smoothly, and planes to not crash.

So does this mean gatekeeping is good, then? Well, not necessarily. There are two different strands to tease out there. Gatekeeping is all about drawing a line around roles that are perceived to be comparatively high value, and in tech, there are two primary dimensions along which this value is expressed.

The first is about having control over technical decisions. This is all pretty straightforward: you want the people working on a product to make the right decisions, whether you’re a user or also working on it. I’m talking high-level philosophical decisions about what the product does, mid-level architectural decisions about how it works behind the scenes (which could impact on development or user experience), as well as really low-level decisions like how to name a variable or where to place a button. So if we want a product/system being developed to be optimal, we also want the people developing it to be up to scratch. This is the reasoning people tend to trot out when they want to justify their gatekeeping: “we don’t want to lower the bar”, “we only hire the best”, etc. Of course, people accused of gatekeeping are usually relying on very skewed notions of what constitutes the “best”, but it is at least theoretically possible to design a system where you only hire the best. Gatekeeping as a means of ensuring product quality is desirable.

The second dimension is a little murkier, because it’s rarely explicitly discussed, but it has to do with money. The roles with the most decision-making power also tend to be the ones that get paid better. (This isn’t just true of the tech industry, obviously, but it’s especially true in tech.) The CTO is almost universally going to be making way more money than the recent bootcamp graduate who’s been tasked with fixing bugs in an internal CRUD app - this is so overwhelmingly true that even stating it seems silly. Of course the CTO makes more money: they have to make tough decisions about technical architecture and maybe product direction, which requires a lot of different competencies and knowledge. Fewer people could fill that role, so market rates are higher, but even going beyond the arbitrary outcomes guaranteed by the market, it feels right. It feels fair.

Similarly, zooming out to look at the industry as a whole, it feels reasonable that wages for technical roles (even for recent bootcamp graduates) are higher than for other sorts of roles, or in other industries. The recent graduate’s job may not be as difficult as the CTO’s job, but it still requires enough specific background knowledge that you wouldn’t expect the average (non-tech) worker to have. So the relative distribution of wages within the economy as a whole feels somewhat fair, too. Maybe tech workers shouldn’t be making that much more than the average worker, but it’s still reasonable for them to make quite a bit more.

The sort of reasoning above is typical to modern-day, meritocracy-inflected capitalism. In this ideological desert, is very difficult to challenge the idea that people who are more skilled, or harder working, or otherwise exhibit “merit”, deserve to be paid more. We can argue about how to define merit, we can argue about what sorts of merit are most useful for society (see, for example, post-2008 backlash against the high salaries on Wall Street), but the link between skill/talent/merit/whatever and reward is basically unquestionable.

So bear with me, please, as I attempt to question it. (If you already think meritocracy is bullshit, then none of this will be particularly new, and I won’t be hurt if you stop reading now.) I propose that tying compensation to skill is bearable only if the correlation is very small, ideally zero; as soon it gets to extreme levels, like what we’re seeing today, it spirals completely out of control.

One reason is that it’s bad in terms of economic efficiency: if salaries are really high in tech but low everywhere else, then some people will enter tech simply because they don’t want to be poor, even though they’d rather be elsewhere; others will flock to the industry not because they’d rather do something else, but out of greed. The first scenario is bad for the individuals involved (they’re sacrificing potential happiness) and maybe even society as a whole (their talents could be much more useful elsewhere, at least if tech is overvalued, which I think is true for historical economic reasons). The second scenario is bad for both the tech industry and society as a whole (people motivated by greed, rather than a conscientious desire to serve society, should not be making big decisions).

But it’s not just about efficiency - there’s an ethical problem, too. And this is trickier, because it gets to the heart of meritocracy’s entire ideological legitimation. Consider the people who do the best under systems that are purportedly meritocratic: is their merit innate? Were they simply born with good genes, fated from birth to be good CTOs? I don’t think anyone fully believes that, partly because it’s obviously not true, and partly because if it were, then our economic system would be rewarding people based on unchangeable personal attributes, which feels manifestly unfair. And capitalism’s all about seeming fair (equality of opportunity, etc).

So let’s suppose, instead, that merit isn’t entirely innate. Maybe some people are better at some things than others, but it’s not entirely nature; there is an element of nurture, as well. One’s social/economic habitus has some impact on one’s abilities and achievements: if you’re born to a poor family in a bad neighbourhood, your life chances are markedly diminished because you won’t have the same educational opportunities or connections that someone born a Rockefeller would, totally independent of your personal abilities. I think pretty much everyone accepts this. For innate capacities to be expressed as “merit”, there needs to be an equitable social system, and right now, our social system is anything but equitable, because certain things are gated by wealth.

What this means, though, is that our social system is unfairly depriving some of the rewards they deserve. And like, until we’ve fixed that, maybe we should hold off on this whole “meritocracy” thing. If you think meritocracy is a fair mechanism for allocating resources, then you should want to correct for the ways in which individuals are prevented from reaching their potential “merit” level due to factors outside of their control. You should want to constantly tweak the system to demolish any barriers - financial, cultural, political, whatever - that would prevent people from achieving as much as they possibly can. Otherwise, what you’re envisioning is not “meritocracy” in its ideal incarnation, but just a new form of aristocracy. (Or maybe this is meritocracy in its purest form, depending on how you define it: a system for preserving hierarchies that has just enough upward mobility to make people believe that everyone has a fair shot of getting to the top.)

I don’t know about you, but I don’t like the idea of a reward system that is fundamentally unfair. And maybe I overcorrect for fairness - maybe my threshold for what is considered “unfair” is too low, and moving toward a system I find fair would also mean sacrificing economic efficiency. It’s possible, I guess. I don’t necessarily see that as a problem, though, because economic efficiency shouldn’t be the most important thing. Plus, if you consider how the system works today, I’m fairly confident that we can design something that is both more fair and more efficient.

Tying this back to the topic of gatekeeping in tech: I think it’s valid to gatekeep for technical reasons (you want people to be competent at their job). I don’t think it’s valid to gatekeep for economic reasons, because you don’t think someone “deserves” to make that much money. Meritocratic capitalism functions by conflating a valuation of merit (however imperfect) with access to resources, which is incredibly inefficient, not to mention inhumane if it’s extreme enough. Should the most talented coders get to live in mansions while those with less aptitude for programming are homeless? This may be a caricature, but it’s also not that far off from what’s happening today, e.g. in the Bay Area.

I guess the gist of my argument is this: gatekeeping is only valid once the economic issues have been sufficiently separated from the technical ones. As the pool of technical workers continues to expand - largely due to tech being one of few industries with enough stable, well-paying jobs - the urge toward gatekeeping by workers in the industry will rise. If you’re worried that your wages will fall, or that competition will be stiffer, then it’s tempting to want to denigrate new entrants to the field, especially if they took nontraditional routes into the industry. And of course, you probably won’t admit the real reasons, because you don’t want to sound like an asshole; instead, you’ll express doubt that people without a 4-year CS degree could ever have a good enough grasp of the fundamentals. You’ll cloak your skepticism in technical reasons, not economic ones.

But look. How much do your technical qualms really matter, at the end of the day, compared to helping a fellow human being to navigate a deeply unfair system? Let’s say your company hires someone who you feel is slightly underqualified for an entry-level role, maybe because they graduated from a bootcamp and don’t have much experience. What’s the worst that happens? You have to spend a little more time coaching them, or some internal CRUD app is buggier than it should be? And in return, they’re able to pay the neverending bills dumped on them merely in exchange for existing under contemporary capitalism?

Prioritise your fellow workers, is what I’m saying. Don’t be the asshole who strives to keep other people out of technical roles by claiming that they’re just not good enough, especially if the roles are entry-level. Your company’s product is, in all likelihood, not that important anyway - a slightly slower development process will not be the end of the world - and hey, people need the money. Solidarity with other people should, where possible, come ahead of allegiance to a corporation, because the latter is soulless and does not care about you whereas the former are what makes life worth living.

So if you really want to be able to gatekeep without being an asshole, because you believe in the importance of well-designed technical systems, then you should fight for a world where the economic factors are removed from the equation. As in, either all jobs are paid similarly, or there are robust-enough public services (healthcare and education but also housing, transit, food, etc) that people can live decent lives even if they aren’t paid well. You’d still have to find a holistic way of gatekeeping without letting personal prejudices take over, but once you’ve done that, at least you wouldn’t have to feel guilty about shutting people out of roles that they aren’t suitable for. (Plus, maybe fewer unsuitable people would be interested in the first place, once the economic distortions have been removed.)

In other words, to solve gatekeeping, we need socialism. I realise that this is the same conclusion that I reach in most of my fragments so far. I’m sorry. It’s true.

« See the full list of fragments