53 Comments
Jan 12, 2022Liked by Noah Smith

I think it's also important to recognize here the degree to which we have willfully moved towards "de-skilling" the lower half of the workforce.

Fifty years ago, a loan officer in your local bank needed to _actually evaluate customers' trustworthiness to repay loans_. And that method of handing out credit had serious problems in terms of discrimination, but it also meant that successful loan officers had to have some real skills at evaluating people's finances. Now, instead, the "loan officer" is just filling out a web form, and getting the applicant's sign off to access their credit report, and the whole evaluation happens behind the scenes. So the programmers who make that system are "high productivity workers", and the loan officers are "low productivity workers", and in theory this probably lets us process loans more easily and cheaply -- not that the end consumer ends up paying any less. Mostly, private investors pocket that expanded margin.

And of course, all of that "highly productive" mortgage processing code, without the oversight of people who were both skilled at judging whether customers would actually pay back, and motivated to make that judgment correctly, got us the crisis of 2008.

Expand full comment

It seems to me that people are leaving out a key concept in this discussion. Aptitude.

And I want to be clear aptitude is not synonymous with IQ. When Eric Adam talks about the service workers and the corner office, what he is really saying (and assuming) is that the workers don't have the aptitude for whatever corner office job there is.

On the aggregate this might actually be true, but for all we know the hypothetical messenger is a brilliant student who is working his way through school.

Now let's go back to aptitude, and let me attempt to explain it. Aptitude is whether a person has the potential to successfully learn and perform a job or skill.

Now... let me be clear here. Aptitude is not just IQ. Let me give some examples:

Computer programmer: not only have the ability to think in code and logic, but also needs the patience and drive to want to do it for long periods.

Special Forces Soldier: must be able to handle stress. have quick reflexes. hand eye coordination. etc...

CEO: have the general ability to understand finance and business. Also has to have people skills (theoretically)

Cowboy: (yes I just watched Yellowstone). Be able to deal with animals. Hand eye coordination. Tolerance for discomfort. Etc...

Finally... my job in inspecting gas turbines. Must know NDT (a subset, which involves eddy current, ultrasonics, etc), hand eye coordination (we use borescope deep in unit), some computer skills (we have to keep track of 100s to 1000s of photos, label them, and write a report).... but the reason I get paid as well as I do is... must have the aptitude to enjoy unpredictable and long periods of travel. Approximately 20% of people we hire quit because of the travel.

Now getting back to productivity and high skilled, low skilled, etc... A certain jobs pay is based on the skills or education needed to perform it... some might require a longer lead time.. such as doctor, and then how many people have the aptitude to perform that job, and then how valuable that job is.

What's any job is worth is a function of. (how many people have the aptitude to do it) x (how long it takes to learn in) x (how many people exist currently that had the aptitude and training) x (how important that job is to the company... multiplier effect)

Another anecdote. My brother is a digital logic engineer for Boeing Space Systems. When he was hired by Boeing after graduation, his boss straight up told him that the University only had taught him 10% of what he needed to know. Boeing hired college graduates because completing a degree in electronic engineering was a sign that they had the aptitude to do the specific job.

What we are seeing now in some industries... especially programming, is that Google and Apple aren't solely relying on Stanford to signal they have the aptitude. They are giving chances to non-college graduates who display the aptitude via tests or past performance.

Now on to Bill Gates and Zuckerberg and being drop out. They are drop outs out of elite Universities. I think it's safe to say that when they entered Harvard or Stanford or whatever, they had a better education that some college graduates.

Expand full comment

The Left's go-to move (and I most certainly exclude Noah from this) is to look at a problem, say that the names are wrong, propose changes, then declare victory. And when the underlying problem stubbornly persists, well, then it is time for another round of changes to the names.

A low-skill job is, in my mind, one that any healthy adult can perform. One must have led a gilded life to have never dirtied one's hands with such work. Nearly all of us start here. And plenty stay here, for a myriad of reasons. But very few of those working low-skilled jobs could simply jump into a high-skilled position. And, dear reader, there's almost certainly a shinier job for which your current skills are insufficient. This is reality, and not an insult. And as difficult as it may be to swallow, there are countless skills for which you simply lack the raw ability. I'm not getting on an NBA court unless it be with a mop! And I don't have the mind needed to work as a doctor. Or an artist. And I've seen people in my own domain who are frightfully intelligent and combine it with a drive I've never possessed.

If you want to change the situation, well, on a personal level you could treat people laboring in lower-skill positions with basic human respect. The way some of us treat waiters and store employees is a disgrace. On a political level, you could support a decent minimum wage, and various other measures that would allow people on the lower rungs of the economic ladder to lead a comfortable life. Take a look at Switzerland to see what sort of life (and country!) one can have when liveable wages are paid.

As for providing better education and restructuring society so that each of us can better realize his potential... these are fine thoughts but, sadly, far beyond our current skills to act upon. Our democracy is crumbling and the world is aflame, and we are not up to the task. When you cannot recognize the fire in front of your nose, there's little use in complaining that you never got your chance at the corner office in the Department of Public Safety.

Expand full comment
Jan 12, 2022Liked by Noah Smith

Reading Noah improves my vocabulary “ ossified” is a banger

Expand full comment

I’ve always thought of “skilled labor” as a technical economic term that specifies training as a barrier to entry for a labor market. Think this is captured in a dictionary definition of the term [1]

> labor that requires special training for its satisfactory performance.

I think this barrier for entry is an important distinction when considering different labor markets. In this context, “unskilled labor” implies that anyone can be hired to perform the job regardless of their previous training/education. All firms and workers compete in a single large “unskilled labor” market. E.g., Amazon and Walmart increasing starting wages can lead to siphoning of employees from other firms in this market or force these firms competing for this labor to raise wages.

In contrast, there is not a single “skilled labor” market, but instead numerous disjoint markets, each requiring different skill sets. At times, the training requirement to enter a specific skilled labor market can benefit workers by limiting supply of workers. Although it is still possible for supply to exceed demand within a specific “skilled labor” market and drives wages down. E.g., the lawyer glut that started appearing in the 90s.

In this context acquiring formal training to enter a specific “skilled labor” market is viewed as an investment of time and money with the aim of increasing earnings. We can analyze the financial returns to an individual for purchasing their own training and incurring the opportunity cost of education versus working. We can also look at government-funded training programs through an economic lens and compute the expected return on investment to society for a specific program.

It can be argued that “skilled” doesn’t necessarily need to refer exclusively to formal training, but can instead also encompass on the job training. E.g., a line cook with years of experience is more productive than someone hired without previous experience. Yet I think it useful to separate “skills” and “experience” and consider them as two orthogonal dimensions. For example, a junior accountant role only requires the formal “skills” of specific education/certification, whereas a senior role additionally requires sufficient on the job experience.

[1] https://www.dictionary.com/browse/skilled-labor

Expand full comment

I note that you left "power" out of your explanation for wages.

And, sure, power is partly a reflection of supply and demand (your replacement cost to your employer would be an even better approximation)

But power can be organised. Labor unions are obvious but union busting is just as much of a power move. Ditto getting to define what amounts to abusive anti-compete clauses in workers' contracts. Etc.

Expand full comment

The obvious reply to Adam's comment is : "so what, those skills are all vital (except the Dunkin Donuts business which should be shut down because its a purveyor of diabetes and obesity).

In fact those skills are more valuable than the overpaid predators at Goldman Sachs - "doing God's work" - who caused the GFC.

Expand full comment

The problem is three fold: First, Managers do not like promoting hard working employees as production will decrease without them there. Second, Managers promote or hire people who are similar to themselves which results in an echo chamber of rectal-nasal integration. The third point being that Managers feel they are entitled to their salaries even when they are utterly incompetent but because the next higher up has the same mentality none of them understand why the business is going down the drain and moral is abysmal, this usually happens after the founder dies and the business is left to the bean counters.

You can talk about skilled and unskilled all you want, but when merit or experience have no bearing on pay raises or promotions you have a situation where the incompetent boot lickers are the ones that rise to top and only promote those who act the same as them.

Expand full comment

I think of low skill vs. high skill jobs differently. I think that picking fruit or being a dishwasher in a restaurant are low skill jobs because no matter how much you train or how long you hold the job, your efficiency (measured as throughput) will increase very little. Accordingly, there is very little motivation for employers to screen for aptitude.

High skill jobs may or may not be trainable through education but they are trainable. Often this training is best achieved by doing the job.

Expand full comment

Good article, but minor nit... never ever use Payscale, or for that matter, Glassdoor, to get recent salary data for software engineers. Those sources use self-reported data sometimes from years past, which doesn't reflect current wage reality.

I assure you, as a software startup founder, if there are engineers getting 10K to do any work of any level of quality, I will a.) immediately double their salary b.) hire 20 of them right now.

SWE salaries in India for decent engineers are up to around 50K US, far higher if they are working in FAANG companies, where the salary + equity packages are well over 1 Crore (around 250K US last time I checked) for people at the middle/upper engineer ranks.

Expand full comment

Or it could be that there is something entirely different going on that is just being obfuscated by all the measures of wages, education and so called "productivity"- that's a joke term surely?

Like the CCP spends one third of what the US spends on Defense and churns out more naval ships, equipment ,space rockets etc in a few years than the US can in a decade.

Its like we have hypnotized ourselves to believe our own "BS" as if it explains the world and affords us a strategy.

In those immortal words - “Everyone has a plan until they get punched in the mouth.” <searches for mouth guard :) >

Expand full comment

Darrell Owens once differentiated the groups as "credentialed" and "uncredentialed" workers and that stuck with me.

Expand full comment

Wasn't the context an argument for returning "highly skilled" workers to the office? That seems to have gotten lost in the discussion. Everyone knows that "low skilled" is a euphemism for "poorly paid, hourly, lacking benefits and generally easy to jerk around". Throw in COVID and that list includes "essential" meaning that they had to show up for work to get paid unlike "more skilled", i.e. "non-essential", workers.

Adams was arguing in favor of getting everyone back to the office. Personally, I think that's a dumb idea, but one can make a class equity arguments for it. It's the old "rich man's money, poor man's blood" argument that ravaged New York City back in the 1860s. Did I mention class? Adams won the working class vote, so he'll appeal to the working class. Maybe he feels a lot of working class people are tired of being the only ones risking their lives and health to keep the city going.

Needless to say, once again, Twitter, once again, demonstrates something important about our elite discourse. It's clearly not about "skills" in the sense of being able to do things, like think before yammering, that we're talking about.

Expand full comment

I would be interested to see a list of jobs where skills/productivity don't match the salary (like overpaid BS jobs). Real estate agent comes to mind.

Expand full comment

I'd agree that the terms low-skill, high-skill are inaccurate shorthand. If I was to try and fix that shorthand, I would say gradual-scale-skill, steep-scale-skill. I'm not a great cook. If you hired me to work in a restaurant's kitchen, I'd work more slowly, make more mistakes, and produce less quality. But the degree to which I'd be more slow, more error prone, and lower quality, when compares to a cook with 30 years of experience would be less dramatic than if the average cook tried to do my job (operating large scale IT systems).

The effect translates into the value of each incremental hour of experience. Swap someone with 30 years of IT experience out for someone with 10 years of experience, and productivity drops by a factor of 10, for someone with 5, by a factor of 100, and for someone with 1 year, by a factor of 1,000.

Swap a cook with 30 years, for one with 10, and productivity drops by 2, for 5 years, 4, 1 year, 8. (Numbers are examples.. I'm sure there's some good studies on various professions here.. and if you disagree with the general view.. we can find them)

This does two things.. it means that if your forced to choose between making these two trade-offs, you'll make the first. So if you have a person with 15 years experience cooking, and 15 years experience in IT, he'll rationally be offered more to use his IT skills than his cooking skills. 15 years of cooking skills and 15 years of IT skills might be considered the same cost to acquire (some real differences.. but I'd hold the primary investment is the time of the person doing the learning), but the cost to sacrificing the IT experience is greater and so the value in acquiring them is higher.

Expand full comment
Jan 12, 2022·edited Jan 12, 2022

I think Mr. Smith has a particular mental image of "skills", how they are acquired, how they are evaluated, and what this means for employment and income. I don't agree with all of this.

Specifically, I think he believes that education provides useful skills for a variety of jobs, that these skills are reflected in workers' earnings, and that employers find these skills to be useful in setting requirements for hiring, promotion, and compensation. Earnings data shows that workers with more education tend to earn more, and Mr. Smith finds this as confirmation for his view.

I propose an alternative view, based on my own experience toiling in the corporate world.

Education, if pursued enthusiastically, can develop many skills: understanding the political background of the civil war, conjugation of irregular German verbs, derivation of mathematical theorems, analysis of Shakespearean tragedies, performing least-squares linear regression, and many other things. Any of these are potentially "high skills", at least in an academic environment, but most are not directly relevant to any employment outside the academy. It is also highly questionable how much of such material is retained by students beyond the course completion.

Supporters of college education might argue that the important skills are not the ones I listed above, but more general skills such as critical reasoning, research, logic, and persuasive writing. These could certainly be useful skills, but there have been widely-reported tests showing that graduating seniors have not improved their skills in these areas during four years of classes.

I think that the college wage premium is largely driven by the fact that many large employers require a college degree for hiring into "professional" jobs. In most cases, any college degree will qualify, so employers aren't expecting specific technical skills. In many cases, college degrees weren't required for the same positions 50 years ago. Paralegals, manufacturing engineers, buyers, news reporters, and bookkeepers are all positions that did not require college degrees, but now generally do. The change is not because the fundamental job requirements have changed.

Large employers now routinely require college degrees, I believe, because of a constellation of factors: increase in the number of people with college degrees, professionalization and specialization of the HR function, and prohibition on use of tests for skill and aptitude.

In 1970, only 11% of the US population had a college degree; in 2020, 37% did. Since so many more people have college degrees, it is easier to use "lack of degree" as a screening factor for "unambitious, incapable employee". A hundred years ago, most adults did not have a high school diploma, yet were still able to function in demanding jobs. As high school completion became the norm, "high school dropouts" were considered unsuitable for hiring in most jobs. The same is happening with respect to college. When I was hired at a Fortune 50 industrial company in 1993, most manufacturing engineers did not have college degrees, and many who did earned them while working as manufacturing engineers. Today, that company requires a college degree to apply for the job.

In 1970, "human resources" was not a career field, and did not have dedicated college programs or certifications. Now, that has changed. Human Resource managers mostly have no experience outside the Human Resources field, and have control of the hiring and promotion process. They are themselves college graduates, and have been taught that a college education is a key employment credential. So, a college degree is a requirement for more and more positions.

In 1970, many companies used aptitude tests to evaluate potential hires. Those who didn't develop their own focused tests could use IQ tests to identify promising hires. Now, such tests are subject to legal challenge for discrimination against protected classes, so companies use them much less. However, companies can safely use a college degree as a requirement. Since protected minorities are underrepresented among college graduates, I'd think this requirement could be challenged for disparate impact, but I haven't heard of any such challenges, successful or not.

Some of the arguments Mr. Smith marshals to support the value of education and skills seem to me to have opposite interpretations that are just as compelling. For instance, he cites a study that students in academic difficulty who nevertheless completed a degree earned better than students forced out by academic difficulty. I would interpret this study to show that students who had a diploma earned more than students with essentially identical learning who did not have a diploma. It seems the diploma, not the learning, is the important thing.

He cites a study showing that mafiosos with college degrees earn more than those without. My interpretation of this study would be that the high-ranking mafiosos send their sons to college to earn respectability, and these graduates are put in high-ranking positions due to family connections.

If skills and knowledge from higher education are the keys to higher earning, it is very easy for anyone to acquire them - classrooms in most schools are open to anyone who walks in off the street. Smaller classes may take attendance and identify interlopers, but large lectures are uncontrolled - or at least were before current trends in campus security. Class syllabi are available to anyone, or at least were before the current electronic classroom technology. A high school dropout from New Haven could sit in on Yale classes and pick up all the knowledge and skills offered. But he couldn't get a Yale diploma.

It is certainly true that worker skills matter - they influence the success of their employers, and of the overall economy. To some extent, they also influence worker success in their jobs, but current trends toward bureaucratizing HR processes, and measuring diversity goals rather than performance goals make an individual's skills and performance less important than they were 50 years ago.

Expand full comment