Firing teachers: the evidence

In most of the anti-union South, it’s easy to fire teachers. That must be why Southern students outperform Northern students.
Oh, wait …

In the discussion among Matt Yglesias, Kevin Drum, Mickey Kaus, and me over the importance (vel non) of making it easier for principals to fire teachers, all of us have been respectful of the rules of blogging. That is, we have refrained from offering any actual evidence (as opposed to raw assertion, plausible argument and anecdote) for our respective positions. But a reader who is a teacher in Gwinnett County, Georgia, reminds me that the experiment has been done: in most of the anti-union south, teachers have little or no protection against arbitrary dismissal.

I am a teacher in a very large school district in metro Atlanta where the superintendent and school board threaten any teacher that tries to move out of her obedient servile position under their feet. We have formed a teachers’ group, The Teachers’ Alliance of Gwinnett, just to have a seat at the table. It is non-union, but still teachers are being harrassed if they are suspected of being part of the group.

I have worked in the public school system in Georgia for 15 years and do plan to leave now. Still, in the South, teachers have no protection already. Our principals are political appointees and yes-men.

Ignore the rhetoric for the moment and concentrate on the fact: The tyranny of the teachers’ unions is not universal! There are places where it’s as easy for a principal to fire a bad teacher (or, of course, a good one) as it is for a Wal-Mart manager to dump a union organizer.

No coddling teachers: that must be the reason the South leads the country in educational attainment, and in particular why Georgia’s students so outperform students from union-ridden Massachusetts and New York.

Oh, wait ….

Footnote I don’t have any problem saying that I would prefer a solution that wouldn’t terminally piss off the teachers’ unions, because the unions help Democrats win elections and I like it when Democrats win elections. That doesn’t mean I’m not willing to support programs the unions dislike if they’re necessary to the program of improving public education; only that, other things equal, I’d rather find a modus vivendi than start a civil war.

How about Mickey? Will he ‘fess up to the fact that, like Bill Bennett, he’d much rather smash the unions than improve the schools? And that he demands that Democratic candidates diss the unions for the same reason I demand that Republican candidates diss the TV preachers: because it’s a good way to break up what could otherwise be a winning electoral coalition?

Once he’s done that, he can start thinking about why, in the face of the past six years, he still wants Democrats to lose elections, even if he doesn’t really want Republicans to win them.

Update In case it wasn’t entirely clear: the chain of reasoning “Massachusetts has teachers’ unions, and its students score better than students from Georgia, which mostly doesn’t” is snark, not social science. Too much else is different.

I checked with my colleague Meredith Phillips, who reports that to her knowledge (and Meredith sees all, knows all) the right sort of statistical study, controlling for the relevant background variables, has never been done. That seems odd.

Second Update Tom Sgouros of the Rhode Island Policy Reporter writes:

It might seem odd that there has been no study done, but that’s really because there have been many. I can’t vouch for your friend’s standards, but it’s wrong to say that this research hasn’t been done, where researchers tried to control for other effects.

Eberts and Stone (1988) were the first of the decent studies I found,

where they tried to control for other variables, but there’s also

Milkman (1998) and Argys and Rees (1995). There are a bunch of bad

studies, too, that include comparisons of entire states. These show

high union performance, but they’re ridiculous.

The decent studies show that the average performance of students is

pretty clearly improved by unions. There are equivocal, or slightly

negative effects at the top and bottom of the student achievement

scales, but the overall positive effect is repeatable, significant and

measurable, even after controlling for everything education researchers know how to control for.

You can find citations to the studies I mention, and a bit more detail

in an annotated bibliography I put together for a Rhode Island teacher union last year.

I’ve passed Sgouros’s study along to Meredith Phillips for review.

If anyone who believes &#8212 as I’m inclined to &#8212 that the difficulty of getting rid of bad teachers is a contributor to poor school performance has any actual evidence for that view, I’d be happy to share it.

As long as we’re trying for social science, as opposed to snark, let’s note that “unionization” probably isn’t the right independent variable; what we want is some measure of the difficulty of firing a teacher, which might be measured in hours of effort by the principal or months of elaspsed time between the beginning and the end of the firing process.

In a quick conversation today, Meredith suggested that a convincing study would have to look at districts where the independent variable had changed for some exogenous reason; otherwise you’d be left with the suspicion that easy-fire districts and hard-fire districts varied on some unmeasured dimension that also correlated with school performance. For example, if districts where the teachers hate the principals, and vice versa, tend to evolve union contracts or administrative procedures to protect teachers from arbitrary firing, and if such districts also tend to have badly-performing schools, then protections will turn out to correlate with poor performance, even if there is no actual causal link.

All this illustrates an under-appreciated point: good policy-relevant social science is really, really hard, and ought to be left to really, really smart people such as Meredith. That’s why I mostly confine myself to policy analysis and snark, which are much easier.

Third update The social science has been done on a related question: the effect of unionization, as opposed to barriers to firing, on student performance. The answers are clear: unionization hurts. Or doesn’t hurt.

Author: Mark Kleiman

Professor of Public Policy at the NYU Marron Institute for Urban Management and editor of the Journal of Drug Policy Analysis. Teaches about the methods of policy analysis about drug abuse control and crime control policy, working out the implications of two principles: that swift and certain sanctions don't have to be severe to be effective, and that well-designed threats usually don't have to be carried out. Books: Drugs and Drug Policy: What Everyone Needs to Know (with Jonathan Caulkins and Angela Hawken) When Brute Force Fails: How to Have Less Crime and Less Punishment (Princeton, 2009; named one of the "books of the year" by The Economist Against Excess: Drug Policy for Results (Basic, 1993) Marijuana: Costs of Abuse, Costs of Control (Greenwood, 1989) UCLA Homepage Curriculum Vitae Contact: