Friday, February 21, 2014

Two quick minimum wage links

The minimum wage is becoming what high skill immigration was for me last spring - my new pet interest. That's probably good... maybe I should make a New Year's resolution every year to dig into a new labor topic for a couple months, just to round myself out. Anyway... two quick links:

1. Jonathan Catalan suggests we take a page from Kahneman's book and have collaborations between researchers that disagree on the minimum wage. In theory this should work fine. I don't think any of the participants in the discussion go in to cook the books, after all. I do think there is a clear difference of opinion on the best way to model this (I'm trying to narrow in on this issue in something I'm writing right now) That could prevent some collaboration and of course that is what's highly correlated with certain results. But if people could hash differences on that point out together, sure that could be constructive. It's a good idea on any issue, actually.

2. A couple days ago the New York Times reported on people crossing state borders to take advantage of higher minimum wages. This has been in the back of my mind throughout this debate, as this sort of displacement that screws up quasi-experimental studies is a major focus of one of my dissertation chapters. It's not entirely clear how this should affect the estimate. Think of a price floor in a standard competitive model. Shift the supply curve to the right. What happens to employment? Nothing if you're assuming the competitive model. Now you would have employment reduced in the sending state, so in a difference in differences set-up you'd bias the estimate up. But that just begs the question - why would they cross the border if employment opportunities contract rather than expand as a result of the minimum wage? For that matter, why wouldn't you have the opposite effect under the competitive model - people without a job crossing the border to the low-minimum wage state where wages may be lower but at least they can get a job? That would bias it down. So the net impact gets very unclear very quickly. That does not mean, of course, that it's not important to consider.

We should have the data to look at this in the ACS (ACS I believe has both county of work and county of residence... I could be misremembering). Somebody should to get a sense of whether it's negligible or needs attending to in future work. An alternative is to do the same analysis with employer based surveys and employee based surveys, which would give you an upper and lower bound. (Actually, Dube has already done this latter suggestion and it doesn't seem to qualitatively change the results).

UPDATE: I forgot:

3. Jared Bernstein on the CBO on the minimum wage.
4. Arin Dube on time series econometrics and the minimum wage (if you look back at the literature there's a lot of old papers on this but not a lot of new ones... there's a reason for that).


  1. #2 is very interesting. If the minimum wage does change both worker and employer strategy, such that both prefer and expect longer job tenure (as the Brochu and Green paper seems to suggest), then the minimum wage might make out-of-town/state workers better hires. What information does the employer get from knowing that a worker has to make a long commute? On average, due to the Alchian-Allen effect, the long-distance applicants will be of higher quality than local applicants: the fixed commuting cost filters out workers who would value the job less, and possibly select for positive attributes such as conscientiousness. Long commutes would also negatively select for workers who can't find a job near home, but that's true before the minimum wage and shouldn't affect changes in employment. The minimum wage allows employers to find higher quality workers, which shifts labor demand right, weakening static disemployment effects.

    On the other hand, if paying higher wages gives employers a way to hire more productive workers (from this mechanism(, there needs to be a reason that employers did not offer higher wages before the minimum wage. Two thoughts that come to mind: first, minimum wage employers tend to filter workers on the job. Give good workers raises, push poor workers out. Raising the minimum wage raises the cost of discovering the quality of a worker on-the-job, making substitute sources of information more attractive. Second, a minimum wage lowers the cost to job-searchers of finding above-reservation-wage (or above-reservation for a long commute wage) salary data.

    1. These are great thoughts. Yes - it's precisely the Alchian-Allen effect.

      Funny story - I was talking about Thomas Mun's thoughts on the carry trade during our discussion of mercantilism in the HET class last semester. I was grasping at the term "Alchian Allen Effect" but couldn't quite put my finger on it (he was hitting a lot of the same points when he was talking about shipping spices from India). Anyway, I was floundering, just explained the point, and never could recall what the effect was called on the spot there.

  2. Daniel Kuehn: I have a comment on the first point - wouldn't it be more appropriate for you and Jonathan Catalan to refer to the research papers written by Daniel Kahneman and his other colleagues in behavioural economics, economic psychology, and other similar domains (e.g. Truman Bewley's book)? IIRC, Amos Tversky published some papers dealing with nominal wage rigidity and the money illusion a few years before he passed away.

    1. I think at Jonathan discusses is fine - why is it inappropriate? I don't see how nominal wage rigidity or money illusion are relevant to the point he is making.

    2. Actually, I realised too late that I should have read Jonathan Catalan's actual post. I had jumped to conclusions that he was citing Daniel Kahneman's recent book for a general audience (Thinking, Fast and Slow). I also thought that he was referring to a statement by Daniel Kahneman on the nominal wage rigidity. Sorry about that!


All anonymous comments will be deleted. Consistent pseudonyms are fine.