Are Teachers Overpaid? It Depends on Your Political Party

Jordan Weissmann produced an article in the Atlantic that provides a perfect example of the adage that goes something like this: "There’s lies, there’s damned lies, and then there’s statistics." The subject is teachers’ salaries, and the issue is how their salaries compare with non-teacher peers. Not surprisingly, there is a lot of leeway in deciding what a "peer" is; equally non-surprisingly the results tend to follow the political biases of the people performing the analysis.

Teachers have become an integral part of our political scene. They enter this realm because they play such a large role in education—a contentious subject at all times. They also tend to be unionized and supporters of the Democratic Party, which automatically makes them targets of the Republican Party. In addition, as mostly public employees, they are categorized as lazy, overpaid, and coddled by a significant fraction of the populace.

Weismann provides us with the results of four attempts to address the pay issue: two from conservative organizations, one from a progressive-leaning outfit, and one, supposedly politically neutral, from an international study.

Let’s begin with the Economic Policy Institute (EPI) which tends to lean left on economic issues. It runs a continuing study which is periodically updated (2010).

"....there are two key stats--pay versus other college graduates and pay versus ‘comparable’ occupations. How do they pick ‘comparable’ fields? Using skill level data compiled by the BLS [Bureau of Labor Statistics], which rates jobs based on factors such as complexity and the knowledge base required to perform it. When they last ran the analysis in 2008, the researchers settled on 16 different other professions, including accountants, reporters, registered nurses, computer programmers, and clergy."

"Overall, the EPI finds that teachers make about 12% less than other similarly educated workers. The picture only changes a little when you factor in benefits. Extrapolating from trends they found in 2006 data, the researchers estimate that total compensation for public school teachers lags by 9%. Compared to professions requiring similar skill levels, the wage gap was 14.3%."

Weissmann suggests these results could be criticized on the basis of weighing formal educational attainment too heavily, because it does not necessarily translate into skill or productivity. That is a reasonable concern, but then he questions the validity of comparing educational attainment when it is known that the most intellectually promising students tend to enter fields other than teaching. That presumption implies that the higher a person scores on SAT tests, the better able they would be to teach addition to first graders. The skill set required for that task is much different than that required to design and build microchips for computers for example. Consequently, I would reject that as a concern unless there was some form of evidence that teachers were too cognitively impaired to perform their duties.

The Manhattan Institute for Policy Research, a conservative outfit, produced a study in 2007 that tried to look at hours worked as a means of comparing teachers with other occupations. They used BLS data for urban areas around the country to arrive at their conclusions.

"Public school teachers did pretty well by the hour. In 2005, they worked an average of 36.5 hours per week at an average wage of $34.06 an hour. That was better than 61% of the other occupations the researchers examined, including architects, psychologists, chemists, mechanical engineers, economists, and journalists (!)."

The criticisms of this approach tend to focus on the BLS data base itself. The fact that the analysis arrived at a 36.5 hour workweek should make one a bit suspicious. All the professions mentioned could also consist of a significant fraction of public service workers. While one might conjure up an image of a chemist who is developing anti-cancer drugs (and highly reimbursed for the deed), most chemists with a bachelors or masters degree are probably in public service in one way or another, and certainly making a lot less money.

The Heritage Foundation and the American Enterprise Institute came up with a recent study that claimed to correct flaws in other such studies. Their intent is to demonstrate that teachers should be compared to other professions on the basis of cognitive ability. In other words, the higher the IQ, the higher one deserves to be paid. Quit giggling—these guys are serious. This "enlightened" approach provides the result that:

"....public school teachers take home total compensation that's 52% higher than ‘fair market levels’ for professionals with similar cognitive abilities."

"When education is taken into account, teachers salaries are more than 12% lower than their peers. But when measured based on cognitive skills, the salary gap evaporates. Once you factor in benefits such as retiree healthcare and pensions, total teacher compensation starts to eclipse what others in their cohort make. To top it all off, teachers tend to take a pay cut when they move to other professions."

Anyone who spends a day listening to the discourse of our elected officials immediately realizes that cognitive abilities have no correlation with success—in fact one would be justified in assuming an inverse relationship. This study allows that teachers have above average scores when it comes to measurements of cognitive ability. That should be sufficient in a profession that is as much an art as a skill-driven occupation. To be a good teacher, is a high IQ of more value than attributes like patience and empathy? This whole approach is rather silly.

Finally, Weissmann provides us with some data from the OECD that attempts to compare salaries for teachers in different countries. The OECD’s approach was again to try and compare the salaries of educational peers.

This study probably tells us something about the prestige associated with teaching in the various countries. On that basis the US falls in the bottom quartile—not a surprising result. Weissmann is justifiably concerned that the absolute value of the results might be inaccurate given the exceptionally low implied compensation rates. If the OECD’s intention was to exhort nations to spend more on education, this data would be supportive.

The studies also seem to imply that all occupations are created equal. Is it true that a teacher has the same value to society as a computer programmer or an accountant? I suspect the answer to that question will also depend on political orientation.

This discussion indicates that one can manipulate data in many ways to arrive at the result one desires. What is the casual reader to assume when the headline reads "Teachers are overpaid by 52%"? What is more frightening is that the media controls which of these studies one gets to hear about.