[vundercind]

There are multiple unions involved with teaching, depending on the state, not just one national one (the NEA or what have you). In some states teachers unions are effectively toothless and aren't even part of the contract negotiation process.

This should make it pretty easy to see how union strength affects firing rates (no, I don't happen to have the data on hand). IME schools tend to avoid firing teachers even when they easily could, in favor of pushing them out, because they don't want the bad press from a firing, so my guess is firing rates for teachers are low everywhere.

We might further hope to see whether union strength affects education quality, but there are too many confounders—the states with weak teachers unions tend to be red states and to have weak economies, either or both of which may have stronger effects on educational outcomes than union activity. But, on the specific question of the effect of teachers unions on teacher firing rates, we can maybe get something like a useful experiment out of these state-by-state differences.

[WalterBright]

What would you suggest is the reason for extremely low firing rates for union teachers?

[vundercind]

“Union teacher” isn’t the distinction, as unions also provide useful professional insurance even in states where they do practically nothing when it comes to employer/employee relations, so many teachers are still members. Do states with strong teachers unions have lower firing rates than those where the unions do almost nothing? I’m saying we may have to look elsewhere for the explanation, if the firing rate in states with nearly-useless teachers unions aren’t closer to where you think they should be.

I’d guess the rates remain low even with weak unions because schools are piss-pants scared of bad publicity, due to the public’s role in (indirectly) hiring and firing the top of their pyramid, and in allocating funding. But maybe I’m wrong and rates of firing are closer to whatever you consider a desirable rate, in states with weak unions. I did go looking, but couldn’t find datasets tackling that in particular. Frustrating, because with that we could get at least a strong hint of the actual effect of unions on this specific thing.