Yes, worker skills matter

It's not just "low-wage" and "high-wage"

The other day, NYC Mayor Eric Adams caused a stir when he said that “My low-skill workers, my cooks, my dishwashers, my messengers, my shoe-shine people, those who work at Dunkin’ Donuts — they don’t have the academic skills to sit in a corner office.”

Achmat X @AchmatX

Eric Adams just said “Low skill workers like cooks, messengers and Dunkin’ Donuts employees don’t have the academic skills to sit in a corner office”

This prompted a backlash online, because it was seen as insulting to cooks, messengers, and Dunkin’ Donuts employees.

In fact, I do think that this language is both insulting and inappropriately reductive. I personally use the terms “high-skilled” and “low-skilled” a lot when writing about immigration, because everyone else uses these terms, but I’ve never liked them. As Sam Hammond rightly notes, skills are not arranged on a linear scale from low to high, and the people we call “low-skilled immigrants” actually have plenty of skills — just not ones that happen to command a high wage premium in the labor market. I hope that at some point we can switch to describing workers in less insulting terms, and terms that more accurately reflect the fact that there are different kinds of skills.

But that doesn’t mean that skills don’t matter! On Twitter, lots of people called for simply replacing the terms “low-skilled” and “high-skilled” with “low-wage” and “high-wage” — as if the labor market outcome is the only thing that matters, and the factors that produce that outcome don’t matter at all. In an article over at Vox, Jerusalem Demsas and Matt Darling — two very skilled writers whom I deeply respect — make a mistake when they endorse this position:

When people refer to low-skill workers, the more precise term would be low-wage. There is nothing inherently unskilled about standing in a hot kitchen for hours cooking or picking countless pieces of fruit every day in the blistering heat…

There is no absolute inherent measure of skills that determines whether a worker has what it takes to sit in a corner office. Skill is dependent on the context of supply and demand for specific tasks and roles…

“As the labor market starts getting tighter then you might sit down and actually read someone’s resume and realize that ‘Hey this person was in the Army for five years doing a complicated mechanical task’ and that’s really relevant to the work that we do,” explained Matt Darling, employment policy fellow at think tank Niskanen Center.

Wages aren’t based on skills, they’re based on scarcity and want. You could be the best Ping-Pong champion in the world, and if nobody wants to watch Ping-Pong you probably won’t make a living playing the sport. Alternatively, you could be a terrible construction worker, but if there’s a severe enough shortage of people needed to build houses, you could get by without ever improving your “skills.”

Of course it’s true that supply and demand determine which skills are rewarded by how much. But supply and demand are not ethereal, inexplicable forces that are merely to be assumed instead of examined. There are reasons that some skills are in high demand in the labor market, and others in low demand, and some of these reasons — not all, but some — have to do with productivity.

Suppose that what people in our society mostly wanted was to have their leaky pipes fixed, and were willing to pay a whole lot of money for that. Well, plumbers have the skill to fix your pipes, and economists mostly do not. If we were to send all the skilled plumbers off to a gulag in Montana, and hand their tools to the economists, and say “Here, economists, go fix the pipes,” it would not go well. No matter how many dollars we paid the economists to fix they pipes, they wouldn’t know how to do it. And the pipes wouldn’t get fixed, and people would have leaky pipes, and our society would be poorer as a result. This disaster would happen whether we paid the economists $1 an hour or $1,000,000 an hour to fix the pipes. “High wage” is not the same as “high-skilled”, no matter what kind of skills the labor market demands.

That distinction is especially important when it comes to immigration. An average software engineer in India, by some estimates, makes less than US$10,000/year — a poverty wage in the U.S. Adjusting for local purchasing power, that’s still only around $25,000 a year, which would be low-wage in America. But that average Indian engineer is still what we would call a “high-skilled” employee, and could undoubtedly command a much higher salary if they moved to America (indeed, the most important determinant of your wage is just where you were born). Not all immigrants could move to the U.S. and become software engineers, though. That’s why we use the awkward terminology “low-skilled” and “high-skilled” to refer to immigrants — because their skills offer a sign of their potential productivity, and their potential earnings, that isn’t reflected in their current wages.

Now, let’s be clear — I am not trying to say that productivity is the only thing that affects wages, or potential wages. Indeed, studies of worker deaths generally conclude that workers are paid somewhat less than their marginal product (though athletes may be an exception). Nor is it to say that skills are the only thing that determines either productivity are wages — aggregate demand is very important, as are human networks and sheer luck.

But nevertheless, productivity is one important link between skills and wages. To say that should only talk about wages without talking about skills is going too far, because it demands that we don’t talk about productivity. It’s appealing to think of the world as one in which the only thing that determines your wage is the job society decides to slot you into, because that would imply that the wage distribution is arbitrary — meaning that we could create a much more economically equal labor market at no cost. But appealing as it is, that worldview is wrong. Re-slotting everyone into jobs that they don’t have the skills for would come at a substantial cost in terms of society’s overall productivity.

This is why it makes sense to use other tools, like income and wealth redistribution, to make our economy more equal, rather than just trying to rearrange people’s work tasks. And this is also why we spend so much money and time on education. If you really believe that society can just reassign wages with no regard to skills, then you should ask yourself why we have redistribution and education in the first place. Personally, I think those things are both there for a good reason.

Education and skills

I especially want to talk about the skills that come from education. This kind of skill is precisely what Eric Adams mentioned, and it’s important to this discussion for at least two reasons:

1. Education is a fairly general skill, meaning that it improves your ability at a variety of valuable jobs instead of just one kind of job,

2. Education is an expensive skill, meaning that it takes a lot of time, effort, and resources to acquire.

The NYC mayor claimed that academic skills are necessary for doing a corner office job. Obviously that’s not true in all cases — Steve Jobs, Mark Zuckerberg, Oprah Winfrey and Bill Gates all dropped out of school, and I’d say they all did pretty well in the corner office. But on average, there’s a reason our country works hard to send lots of people to college, and there’s a reason employers pay more to hire college graduates, and it’s not just about signaling as some would have you believe.

There’s plenty of research to show this. A 2018 paper by Ost, Pan & Webber compared people who just barely made the GPA cutoff to stay in school to those who just barely missed the cutoff, and found that getting to stay in school was well worth it in terms of earnings. A 2016 paper by Arteaga found that a reform in Colombia which reduced the amount of coursework necessary to earn a degree resulted in a substantial drop in future wages. Indeed, the college earnings premium rises over a worker’s career, which is the opposite of what you’d expect if it were just a signal. In fact, a 2016 paper by Campaniello, Gray, and Mastrobruoni found that this earnings premium is even stronger for mafiosos than for other workers; one would assume that the mafia cares relatively little about the prestige value of hiring someone with a college degree, meaning that college was teaching the mobsters useful skills. Detailed longitudinal studies have also found that college improves a variety of skills for the long term. And psychologists have even found that education has a huge positive effect on IQ.

This doesn’t mean signaling doesn’t exist, or that everyone is prepared to complete a degree, or that college is always worth the price. But it does show that on average, college is teaching people stuff that is valuable in the working world. That “stuff” is what Eric Adams is calling “academic skills”. That doesn’t mean his statement is correct; maybe we could stick cooks in all the corner offices, and after a period of adjustment they’d be doing well enough to keep their jobs. But it means he’s right that educational skill differences do seem to matter. That’s why we have stuff like Pell grants!

In other words, when it comes to education, it really does make sense to differentiate between those workers who have more and those who have less, at least when speaking in statistical generalities. So Eric Adams isn’t talking nonsense.

On-the-job learning and America’s wasted potential

At the same time, I think it’s easy to put too much importance on education when it comes to skills. Eric Adams is discounting a very important way that people acquire skills — learning on the job. This can happen through several methods — on-the-job training, learning by doing, or learning to work within a specific organization.

Plenty of research finds that the skills people acquire at their jobs are a big deal. Specific on-the-job training programs have been found to have positive returns in India, Chile, and elsewhere. Heining (2019) finds that much of workers’ productivity is company-specific; people learn how to work within an organization, and that increases their value by a lot. Heckman, Lochner, and Cossa find evidence for substantial learning-by-doing. And so on.

This is important because we don’t want to make the mistake of forcing workers onto permanent job tracks based on their education alone. Yes, college is useful, but that doesn’t mean a cook would necessarily do a terrible job in the corner office. Some people without much education would do a great job in management, and there need to be more ways of helping them find out if they’re cut out for that role.

In countries like Japan, where people spend much of their careers within a single organization, this is easier to do. Bosses can observe their workers’ aptitude over the long run, and put them onto a management track if they deserve to be there (whether they actually do this, of course, depends on corporate culture; Japan tends to overvalue seniority). Companies can invest in on-the-job training programs, secure in the knowledge that their workers are unlikely to take that investment and hop to another employer.

In a more fluid labor market like the U.S., where moving to another company is often the best way to get a promotion or a raise, it makes less sense for companies to invest a lot in building — or even monitoring — their workers’ skills. This might be why we rely on education so much as a measure of skill; we just don’t have access to the alternative measures. In fact, our penchant for job-hopping may have trapped us in a bad equilibrium where companies socialize the cost of worker training by depending on universities, and where tons of talent remains undiscovered down in the kitchen or the mail room.

So even though Eric Adams is not completely off base with his comments about workers’ skills, I think the U.S. has gone too far down the road of assuming that education = skill level. We risk becoming an ossified class society based on college pedigree. The people yelling at Eric Adams have a point — cooks and dishwashers have a ton of unused potential, and we need to think about how to fix that.

---