The teacher pay gap is a myth? Maybe. This one has a different spin than the typical unthinking “teachers are underpaid” articles (it argues they’re not, and yes there is data to support the assertion), which is why I point to it specifically, but I think both sides are underestimating an important point about flexibility and possibility in any given career; teaching has a relatively low pay ceiling and even if a given teacher wants to work twice as hard and make twice as much money, that’s really difficult and potentially impossible. High IQ and high conscientiousness persons in other professions, meanwhile, can maximize incomes to a much greater extent than teachers. For example, as a consultant there are various avenues I can explore to further raise my income, but if I taught high school I couldn’t. In big metro areas, too, there are also many more employers to choose from than most teachers can choose from, since there’s usually one district, or a handful in nearby areas; at the same time, high rents in many superstar cities make teaching less attractive, so the “teacher pay” story is also a story about land-use regulations that raise the cost of housing—it’s just never discussed in those terms.

Teaching has other challenges: bureaucracy is a real problem and so is low morale, perhaps related to bureaucracy. In most sectors, if you don’t like the bureaucracy at one company you can switch to another, but this is considerably harder for teachers, since most districts are local monopolies. Financial returns to IQ also seem to be going up and at the same time we’re making housing in many cities and first-tier suburbs out-of-control expensive—so teachers are really taking it financially, from both the cost-of-living side and from the payment side. I have heard through the grapevine, for example, that pretty much every new public school teacher in Seattle says the same thing about how she got there: “My husband got a job at Amazon and…” That’s no doubt an exaggeration, but how much of one?

To use myself as an example, my parents moved from California to a boring suburb when I was a kid, and they left California because the cost of housing was so high. I checked how much they paid for the house in the boring suburb; in inflation-adjusted terms, they paid about $350,000 in today’s dollars. The same house is estimated to be worth about $700,000 today, so the cost has just about doubled in real terms, but that suburb forbids townhouses in the vast majority of its land, so no one can buy the house, knock it down, and put two houses on the same lot. If that were legal, we’d not have the housing crisis we do in many high-productivity cities and their suburbs. If we could lower the overall cost of living, struggles around teacher pay would seem less dire. Oregon, for example, has legalized duplexes and fourplexes statewide, so it will offer a natural experiment in lowering housing costs.

Let’s return to the original link:

shortages exist precisely where expected in a nationwide labor market that pays an increasing premium for STEM and other specialized skills. When researchers have examined the teaching vacancies that districts say they have trouble filling, they find that elementary, English, and social-studies teachers are not the problem. In fact, the Department of Education found that 20 states and the District of Columbia produced over twice as many elementary-education graduates as they had elementary-teaching positions to fill. At the same time, some districts struggle to fill STEM and special-education positions.

It’s hard to measure teacher “productivity” and yet almost no one bothers trying. So the most productive teachers are incentivized to move somewhere they can be paid in line with their productivity, which can’t be at most schools. Talk about an adverse selection problem! We have problems in data and definitions in both health care and teaching; rarely have two fields absorbed so much concentrated thought and produced so little change. The data and definition problems make it much harder to evaluate quality, inputs, and outputs than in other fields, especially those related to manufacturing. In most businesses, a major goal is simple (profit) and measurement is comparatively simple. The output of high-quality teaching may not manifest itself for decades, and very little of the value improvement is captured by the educator. By contrast, if you build a better widget, there’s a decent chance you’ll be able to capture more of the widget value beyond the cost of production.

Because we’re not measuring, or able to measure, teacher quality effectively, most schools also seem to pay based on years of experience. Anecdotally, I’ve been told that it can be very hard to get hired by another public school after 10 – 15 years of experience, because then you’re too expensive and most schools would rather hire cheaper teachers. So this is another way of ossifying the labor market.

There’s also a narrative about teachers working a lot of hours per week, but Bureau of Labor Statistics data consistently show that teachers work 40 hours a week. One example, although I have seen many other similar ones. Anecdotes consistently run one way, and data consistently runs another.

Overall, however, it seems that the number of people trying to have a completely honest conversation around this topic is not high, and many who argue that we should pay teachers “more” don’t know what “more” looks like. We may also see lots of composition effect problems, in which people who can get higher wages or better working conditions leave and get them, while those who can’t stay at school districts and take what they can get, which is at least consistent with some of the teachers I had in high school, but that may not be desirable at the societal level.