Are teachers overpaid or underpaid?

www.slowboring.com

Are teachers overpaid or underpaid?

Matthew Yglesias

13 - 16 minutes

There was some discourse last week on whether American teachers are underpaid or overpaid, to which my basic reaction was that these are very ambiguous terms.

I think the most normal thing that people mean when they express the view that teachers are underpaid is probably not a policy analysis at all, but an observation along the lines of the diamond-water paradox — I like watching the NBA and I also like my kid’s elementary school, but if there were no NBA games my life would be fine, whereas there being no elementary schools would be a catastrophe. It seems perverse on the level of intuitive folk morality that you can become dramatically richer by being a star entertainer than by contributing to something as socially valuable as education. That’s particularly true because even though obviously top athletes, actors, and musicians work incredibly hard to hone their craft, they are on some level doing things that most people do as hobbies as jobs.

And though I’m no LeBron James, I also acknowledge that I personally am part of this slightly perverse moral economy. What I do for work is also something that lots of people do for fun in their spare time. And while I’m sure people would miss my takes if I were hit by a bus tomorrow, plenty more take-slingers are out there ready to step into the breach. Substacking is more lucrative than teaching not because I’m providing more value to my readers than teachers provide to their kids, but because we have more than 13,000 paying members, while teachers max out at 20 to 30 students in a classroom.

That basic moral economy point is completely correct and it’s why we, as a society, both do and should hold teachers in high esteem as a group. You’ll never see someone just generically offering praise for political columnists and the hard work we do, and there’s no “sports star appreciation week.” If someone wanted to start a campaign to get people to fly pro-teacher flags or post “teachers are good” yard signs, I’m 100 percent on board with devoting more attention to the cause of supporting our Educator Troops.

But then there’s a question of public policy. The NBA’s collective bargaining agreement guarantees a minimum salary of $925,000 for a player with no experience, rising to a veteran minimum of $2.6 million for players with 10 or more years in the league. We could say that since teaching is more important than basketball, teachers should be compensated on the same scale. But since there are way more teachers than basketball players, the cost of this would be enormous and require huge tax increases. That’s obviously an extreme proposal, but it speaks to the basic question: should we raise taxes and pay teachers more?

You often see the trope that American teachers are working in conditions of dire poverty. Part of the setup of “Breaking Bad” was the idea that as a high school chemistry teacher, Walter White was so poorly compensated that he worked part-time at a car wash in order to make ends meet and also couldn’t afford cancer treatment.

There is a lot of spatial variation in teacher compensation, and it is a documented fact that some teachers have to work second jobs, but on the whole, teachers are not mired in poverty.

Like most government employees, teachers also generally have above-average benefit packages in terms of health care and pensions. This is not to say that anyone is earning royal sums of money doing this. But the United States of America is a high-income country, and teachers on average enjoy decent living standards.

The Economic Policy Institute, a progressive group with pretty close ties to teachers unions, argues that we should compare teachers’ weekly pay to the pay of non-teacher college graduates, and while teachers generally earn a bit more than the average American, they earn less than the average college graduate.

This is an interesting and relevant fact about teacher compensation, and I think it’s also part of a universe of stylized facts that most people are broadly aware of. A converse fact would be something like “plumbers earn more than you see in a typical working-class occupation,” which I think we broadly understand to be because plumbers have specialized skills that distinguish them from most retail and food service workers. Andrew Biggs from the conservative American Enterprise Institute offers a lot of methodological quibbles with EPI’s math, but far and away his best criticism is the broad point that this is a category error. Plumbers aren’t “overpaid” relative to their educational credentials; they have valuable skills (they know how to fix the plumbing) that don’t happen to be taught in BA programs.

    In fact, about four in every 10 occupations we analyzed show an alleged wage premium or penalty greater than the one EPI claims for teachers. We could write dozens of reports with titles such as “The Electrician Pay Premium” or “The Massage Therapist Penalty,” all modeled directly on EPI’s analysis of teachers. No one would publish these articles, however, because alleging significant pay penalties or premiums in nearly every occupation makes no sense. Such results say much more about the method of analysis than they do about America’s labor market.

    The control variables in EPI’s statistical model obviously cannot explain the significant salary differences across occupations — otherwise there would not be so many occupations with large premiums or penalties. The model fails because salaries are determined by the supply and demand for specific skills that vary across occupations even after controlling for education.

I think Biggs arguably overcorrects in the opposite direction there at the end. The pay for electricians and massage therapists is set by the relative supply and demand of certain skills. But teacher pay is set by a mix of bureaucratic processes and collective bargaining. In a certain tautological sense, we have exactly the teacher workforce that we are paying for, so the individuals are neither overpaid nor underpaid. But there’s still a question of whether those bureaucratic processes have landed on a good solution.

Precisely because teacher salaries are set in a quasi-marketplace, one possibility is that inadequate pay would result in shortages of teachers — large-scale quitting, inability to fill vacancies, or soaring class sizes.

The Biggs piece I linked to is from September 2019, and he looks at quits and concludes there’s no evident problem here. But that was a lifetime ago, macroeconomically. One plausible story would be that school districts got accustomed to a certain pay scale during the long, slow recovery from the Great Recession, but those pay scales are inadequate for the much hotter labor market of 2022. And certainly there are plenty of stories floating around about teacher burnout, shortages, and recruiting problems. At the same time, there are over 100,000 schools in America so there could be thousands of individual schools struggling with retention and recruiting without a systemic problem.

    Matt Barnum, one of the best education journalists, looked at this and says he can’t find evidence of a systemic national problem.

    Derek Thompson, one of the best trend-debunkers, also looked at it and says he can’t find evidence of a systemic national problem. 

And as Georgetown professor Marguerite Roza notes, one proximate issue in many districts is declining enrollment rather than understaffing.

But characterizing national averages, while useful, only gets you so far. Public school enrollment is declining nationally and declining specifically in D.C., but it is growing at my kid’s school. If your particular school district is struggling with recruiting and retention, offering higher pay seems like an obvious solution. But that’s not the national situation.

On average, though, district pay scales seem adequate to fill slots with licensed teachers. This leaves us with the bigger-picture question of whether we are recruiting the teachers we want to be recruiting.

Vox writers earn less money than corporate lawyers. But a lot of my friends from college are corporate lawyers. And a healthy share of people who I met as a young journalist left the field to go to law school. And generally speaking, young Vox writers tend to have similar resumes as young lawyers — they had good SAT scores, went to flagship public universities or well-known private schools, majored in something non-STEM, and graduated in a timely manner. The relevant skill sets are also broadly similar, involving a lot of research and writing and some public speaking.

Basically, highbrow websites and good law firms are competing for the same kind of people. And the reason the lawyers end up making more comes down to what is called “compensating differentials” — i.e., journalism is kinda fun while lawyering is a pain in the ass.

Teaching is not really like that. After over a decade of improvement, it’s still the case that most new teachers come from the bottom half of the SAT/ACT distribution. That’s really the synthesis of the EPI and AEI points made above — if you pay teachers less than the average college graduate, then you end up with a population of teachers who mostly couldn’t get the jobs in the high end of the salary range.

And you can see this in international comparisons. Because America is a rich country, American teachers earn more than most teachers in other countries. We’re looking here just at primary school teachers to maximize comparability because different countries handle high school differently, but the same broad point holds.

But, again because America is a rich country, those salaries are not necessarily super-competitive with the salaries people could earn elsewhere. I wouldn’t take these figures as gospel truth because there are a lot of nebulous questions about how to factor benefits into this. But the big point is that teachers in Portugal and Greece are very well-paid compared to other Greek or Portuguese people in a way that isn’t true of American teachers, and that’s true even though American teachers earn more than Greek or Portuguese ones.

Germany is a lot richer than Portugal. Consequently, to make primary teacher pay as competitive in Germany as it is in Portugal, they need to pay salaries that are very high in absolute terms. Since America is a bit richer than Germany, we’d have to go even higher. The goal of that wouldn’t be to give a windfall to the existing crop of teachers — it would be to raise the bar over time for who goes into teaching.

Looping back to where we started, I think there’s a good case that it would be worth investing more money in recruiting and retaining teachers.

But the union-backed idea that teachers are underpaid so we should do an across-the-board pay increase seems like the least cost-effective way you could possibly do this. It actually seems like a pretty serious problem that teacher compensation is so heavily backloaded. Because your salary rises with time spent on the job, because you can get bonus pay for obtaining a mid-career master’s degree, and because teachers generally get defined-benefit pensions if they stay on the job for a long time, new teachers often earn dramatically less than the average teacher.

A paper Jesse Bruhn, Scott Imberman, and Marcus Winters published last year showed an interesting complementarity in this regard between traditional public schools and charter schools in Massachusetts. The charters have more flexible hiring and firing rules — a recent college grad who got good grades and now thinks they want to teach can do so in charter schools without the time and expense of obtaining a teaching license. And what they find is that the least effective charter school teachers tend to leave the profession (by getting fired, I think), but the most effective charter school teachers tend to get licensed and go work in the traditional public school system, which pays better. That’s a neat story for the Bay State, but it also illustrates that left to their own devices, Massachusetts’ public schools would likely be screening out a bunch of potentially excellent teachers because they’re not sure they want to commit to a teaching career.

So it seems to me that it would be good to offer public school teachers charter-like levels of job security for their first three to four years on the job, but in exchange provide significantly higher entry-level salaries. The goal should be to make it so that doing the job of an entry-level teacher is reasonably remunerative on its own terms, even if you don’t end up getting mid-career pay bumps and a pension. That way you’d get a larger flow of applicants for the entry-level positions and quality would rise over time.

It would also be good to offer bonus pay to teachers with 5+ years of experience and demonstrated above-average efficacy who are willing to work in high-need schools as a compensating differential. High-poverty schools often have difficulty retaining teachers, which leaves principals struggling to fill vacancies with inexperienced staff or problematic veterans who the more desirable schools don’t want. We should aim for a situation where new teachers are generally getting their reps in with assignments that have a low degree of difficulty, and then once they’re more experienced, have the opportunity to further enhance their earnings by taking on tougher jobs.

Long story short, I think there’s a decently strong argument that we are underinvesting in hiring teachers. But that’s different from saying that we need to give large blanket raises to every teacher. Some districts probably do need to do that, but what we need to do on a national basis is something much narrower, though absolutely worth spending money on.