YET MORE ON UNIONS!….Tyler Cowen has been assigning homework. I read yesterday’s assignment, but today he’s got a new one: a paper by John DiNardo and David Lee suggesting, among other things, that unions don’t actually raise wages. So why bother with them, Tyler asks?

As it turns out, this paper is unavailable to the hoi polloi, so instead, Cowen-like, how about if I simply offer up a few reasons for thinking this result is probably wrong? Here are three:

  • In yesterday’s reading assignment (here) we learned that unions don’t negatively affect productivity but do negatively affect firm profitability. Now, it’s not impossible that there could be other reasons for lower profitability even as productivity stays constant, but higher wages sure seem to be the most likely. That paper also reports lower investment levels by unionized firms, and again, higher wages seem like the most likely culprit.

    (I should add that this also passes the “makes sense” test. Basically, you’d expect higher wages to get passed on in various ways depending on the nature of the union and the nature of the industry. Part of it would get passed through in the form of higher prices to consumers, part of it would come out of investment spending, part would come out of executive pay, part out of lower profit margins, etc. But none of these things should happen unless wages really are chewing up a bigger share of corporate costs.)

  • Other research has long shown that unions produce higher wages. Cross-country comparisons show that union wage differentials in the U.S. are bigger than in Britain, for example, and cross-industry comparisons have shown that unionized industries pay higher wages than non-unionized ones. Of course, all this other research could be wrong, but how likely is that?

  • Finally, at some point you have to assume that people aren’t completely and utterly irrational and misinformed. For over a century, employers have fought unions tooth and nail because, among other reasons, they think that higher wages will put them out of business or make them uncompetitive (or simply reduce their profits). Likewise, workers have been fighting for unions tooth and nail because, among other reasons, they’re convinced that unions provide higher wages and benefits. Maybe everyone has been wrong all these years. It’s possible. But is it likely?

Bottom line: An awful lot of other stuff, not to mention common sense, has to be wrong in order for DiNardo and Lee to be right. I’d take this with a big grain of salt unless other researchers corroborate it.

And while I’m at it, I’ll toss out one other thing about this whole field that makes a lot of research results less than compelling (including ones that favor my position): they mostly focus on unionization in manufacturing industries. There’s no other choice, of course, since that’s where most unionization has occurred in the past, but current union organizing activity is focused most heavily on service industries. The dynamics there are far different (there’s generally no overseas competition, just for starters), and it’s very hard to know if research results based on manufacturing industries, which are in decline for lots of reasons unrelated to unions, also apply to service industries. So caveat emptor.

And now I have an assignment of my own. It’s for Brad DeLong: Will you please walk down the hall and tell David Card that his life would be much richer if he’d start a blog? That would be very cool. Thanks.