Friday, September 5, 2014

My minimum wage paper is out at EPI

You can find it here. The idea is to communicate to a broad audience the fact that the "credibility revolution" in econometrics matters a great deal for the minimum wage debate, and that if we divide studies according to whether they employ quasi-experimental methods or fixed effects models the quasi-experimental methods (which are strongly preferred) tend to produce the "no disemployment effect" results. That should matter for the debate.

I also include some cautions about how to apply this research to policy. These are every bit as important. I've had one discussion with a journalist so far about the paper and the fast-food strike for a $15 minimum. I did the math for him on how big an increase that $15 would entail and compared it to Figure A in my paper and flatly told him that you can't justify this with the existing research - our strong priors ought to be that it will reduce employment, with perhaps a few high wage metropolitan areas as the exception.

An excerpt from the report I like ("matching methods" is a more user-friendly term that I use to talk about quasi-experimental methods - that is all well-defined up front and defined in even more detail in the endnotes - the idea is that in all quasi-experimental methods you are matching some treatment case to some comparison case):
"It is difficult to overstate how uncontroversial it is in the field of labor market policy evaluation to assert the superiority of matching methods to the nonmatching approaches described above.9 The seminal evaluations of the effects of job training programs, work-sharing arrangements, employment tax credits, educational interventions, and housing vouchers all use at least some sort of matching method, if not an actual randomized experiment. In their widely cited survey article on non-experimental evaluation, Blundell and Costa Dias (2000) do not even mention state-level fixed-effects models when they list the five major categories of evaluation methods. In a similar article, Imbens and Wooldridge (2009) do mention fixed-effects models as a tool for policy evaluation, but clarify that these were used before more advanced methods were developed, noting that the modern use of fixed-effects models is typically in combination with other more sophisticated techniques. For example, Dube, Lester, and Reich (2010) also use a fixed-effects model, but more importantly it is a fixed-effects model that utilizes rigorous matching strategy to identify the effect of the minimum wage. Sometimes fixed-effects models are the best available option if no natural experiment or other matching opportunity emerges to provide a more rigorous approach. Well specified fixed-effects models can still be informative. But faced with the choice between a well matched comparison group and a fixed-effects model, the former is unambiguously the stronger study design."


  1. I like this paper a lot. It does a very good job of laying out both the methodological issues and the methodological choices made by researchers. And I think the conclusions are spot on.

  2. Congrats on the new paper! It was a very interesting read.

    However, I do have one question that doesn't seem to be answered in your literature review. Specifically, how do we know that the minimum wage increases that are considered in these matching studies actually established a higher wage floor (i.e. increased worker compensation to levels above market equilibrium)?

    This seems important to me, because if a minimum wage hike did not result in companies having to pay their workers more than they would in equilibrium, then every economist would agree the law would have no impact on employment. Yet, it seems like there was no effort to reject this alternative hypothesis.

    You could point out that some workers did see an increase in take-home pay after the law was enacted, but I don't think that is enough. After all, compensation comes in more forms than just your pay check. When I washed dishes a truck stop growing up, I made minimum wage, but also got free drinks and half off any food that I bought. I could easily see that company responding to a minimum wage hike by increasing my wage, but taking the other benefits to keep my actual compensation the same.

    Labor econ is not my field, but this doesn't seem like a crazy idea to me. I mean, fast food restaurants have powerful lobby groups like the National Restaurant Association that they pay to protect their interests from legislation. If the NRA did that job very well, you would expect that the only minimum wage laws that actually get passed would be those that don't significantly impact fast food profits--meaning those laws that don't increases real worker compensation much if at all.

    Any thoughts you have on this subject would be much appreciated.

    1. Dallas Wood -
      So I didn't spend a ton of time on this, but this is part of many potential alternative margins that I was referring to on page 4, and the fact is we don't know a lot about how those margins are affected. I know Michael Reich (a co-author on the famous RESTAT paper and a big researcher in the field) is interested in looking into this more. He is doing a lot with local minimum wages out in California so it's a real question of how this all works through.

      He typically mentions price changes rather than changes in alternative compensation, and I think that's reasonable. My guess is (although I could be wrong) that workers don't drink enough soda to make up the difference. I also bet employers know that workers are going to keep drinking the soda anyway. Generally I doubt these workers have enough non-wage compensation to make up the difference (and this is something I speculated on in the paper too).

      So yes, I think it's important. It is not ignored - it's regularly cited as a possible explanation, but there could be a lot more empirical work done on it. The tough thing is there's not very good data on it (which is another clue-in that the cost savings involved are probably not that substantial).

      Two other points:

      1. I think it's important to distinguish between equilibrium and market clearing prices in your model. Ignore that non-wage compensation issue for a second and imagine that workers are only paid in their wages. It's possible to set the minimum wage above the prior equilibrium but below the market clearing equilibrium, and that would explain the lack of decline in employment. This is the monopsony power point. I just want to make that clear because your comment seems to use equilibrium and market clearing interchangeably. There are several potential non-market clearing equilibria (in the perfect competition market clearing sense of the term). I know you are thinking of something a little different with your non-wage compensation point (essentially you're suggesting maybe everything's in equilibrium and we're just altering the composition of compensation).

      2. So on NRA - maybe. But the NRA hasn't been happy about any of the minimum wage increases including the ones that seem to show no negative employment effect. One natural explanation for why is that the minimum wage is not just altering the composition of compensation - it's cutting into profits.

    2. Thanks for the feedback, Daniel!

      #1. I'm glad to hear that the point has not been ignored. I honestly didn't think it had been, I just am not familiar enough with the literature to track down what has been said on the topic. This is one reason I enjoyed your lit review, it taught my a lot about what has been done and the methodological issues involved. And in an easy to digest fashion! Thanks for taking the time to answer my question. I will definitely check out Reich's work.

      #2. I take your point on there being a difference between market clearing and equilibrium prices. Like you point out, this distinction is particularly important when we are dealing with markets that may not be perfectly competitive.

      Thanks again for answering my question.

  3. I've been seeing a lot of e-mails about this post, I'm guessing from spammers posting bs that gets deleted. Anyway, I thought I would use this as an excuse to revisit the question I asked.

    Specifically, I just want to say that my question is not rhetorical. I really would like to find out how we know that previous increases in the minimum wage actually established higher wage floors. Have we just assumed that this is the case? If so, why? Couldn't it be just as likely that the lobbying groups like the National Restaurant Association would block any legislation that actually increased the compensation workers received in order to protect their client's profits?

    This seems like a question we could try to answer empirically. After all, establishing a higher wage floor would affect more things than just employment. For example, wouldn't an increase in the minimum wage increase the opportunity cost of working a job that is not covered by the minimum wage such some harvesting and ranching work? If so, wouldn't you expect the wage of field hands to rise after a min wage is introduced as more field workers start trying to find jobs covered by the higher minimum wage?

    That is just one example I could think. It isn't a conclusive test that the minimum wage actually established a higher wage floor, but it helps build the case. Maybe this study has already been done? If so, I can't seem to find it.

    1. Sorry I missed this - can't answer right now, but bug me in a day or two if I don't get to it.

      Yes, I imagine notifications get sent when a spam comment comes through. I get a ton of them and they're automatically filtered.

    2. The field hands point is interesting. I don't know of any work on that, and I don't even know how good data is in the non-covered sector. A better option might be tipped workers although technically they should still get whatever the statutory minimum wage is when tips are all accounted for.

      The trouble with the field workers is also that this has to be a genuine opportunity cost. I think it's plausible that some of them would have a hard time getting minimum wage jobs if there are immigration issues involved, etc. I don't know. But it sounds like an interesting idea.


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