A great deal of the debate surrounding public sector unions focus on how much public employees earn versus private workers. Every credible analysis – those that account for huge differences between public and private workers in terms of characteristics like profession, education, and experience – find that public compensation is competitive or lower than that of private-sector workers (for recent examples, see here, here, and here, or a review here).
I have, however, heard a few thoughtful observers make the point that virtually all these analyses include education workers, and that this might be a little misleading. It’s a fair point. Roughly one in five state/local government employees are in fact K-12 teachers, while another five percent are professors at public colleges and universities. This is important because analyses of public/private sector compensation essentially compare public employees with workers with similar characteristics (education being the most important one) in the private sector. The research above indicates that workers with more education pay a larger “price” for working in the public sector, whereas many lesser credentialed, lower-skilled government jobs actually pay more. Since many teachers have master’s degrees (and professors Ph.D.’s), and they are such a huge group, it’s reasonable to wonder if they might be skewing the overall estimates.
So, I decided to see if a comparison of public/private compensation that does not include teachers and professors would yield very different results. Let’s take a look.
In a previous post, I used U.S. Census data (American Community Survey, available from IPUMS.org) to break down the public/private gap in wage and salary earnings, controlling for state, work schedules, experience, and education. Here, I replicate this analysis, but exclude K-12 teachers and college/university professors (both public and private).
Let’s first see how average public/private earnings compare before and after educators are excluded (note: none of these analyses includes benefits).
The exclusion clearly makes a difference. It increases the average private sector salary (probably because private school teachers are poorly paid), while decreasing the public sector average over $1,000.
But this comparison, of course, does not account for differences among workers in terms of key wage-determining factors such as education and experience. When those differences are controlled for, it is almost certain that the earnings gap found in my previous post (about 20 percent) will be lower, since the public sector workers included in the analysis will be less educated, on average, than they were when teachers were included.
In the table below, I present the results from the previous post, along with the public/private gap when educators are left out. The models control for work schedules (hours and weeks), state, gender, experience, and education. The details are in the table stub, if you’re interested.
The results show that excluding teachers and professors does mitigate the public/private earnings gap – reducing the "advantage" for private sector workers from 20.2 percent to 14.2 percent – but the difference remains large.
This illustrates the fact that more educated, highly-skilled workers - such as teachers and professors - suffer a more substantial earnings penalty for government work. When a large group of the most educated workers is excluded – i.e., educators – the gap is narrower. But it is still huge – around 14 cents on the dollar.
So, it is certainly true that educators are “responsible” for part of the public/private earnings gap. They are the dominant public employee group, and would therefore exert considerable influence on the results, no matter what.
Nevertheless, with or without teachers and professors, public employees’ wage and salary earnings are substantially lower than those of comparable private sector workers.