Thursday, January 30, 2014

What is a "modest" increase in the minimum wage?

Bob Murphy asks me whether I think the proposed 39% nominal increase in the federal minimum wage counts as "modest". As readers know, I'm not a supporter of increases in the federal minimum wage. I don't think it spells disaster the way some people do, largely because of my read of the empirical literature. But I think it could hurt especially vulnerable labor markets and there are simply better policy options (and better ways of doing the minimum wage!) than raising the federal minimum wage.

Still, the question of what a "modest" minimum wage increase consists of is important.

We care about a couple things when answering this question, first and foremost being when the minimum wage is binding. If it's not binding it obviously doesn't matter.* The literature suggests that on average, past increases have not been large enough to be binding, at least in a disemployment sense (they are clearly binding on earnings, because we see increases in earnings as a result of the increases, again on average). There are at least two things to think about when it comes to whether the minimum wage is binding: erosion of the previous minimum wage level by inflation and productivity growth. Here's a very crude extension of an EPI chart to try to capture these things for the proposed increase:


I've done two things here. First I've crudely extended the erosion of the real value of the current minimum wage out to approximately 2016, which is when the $10.10 minimum wage is supposed to be set. Then I've added a horizontal red line indicating a crude estimate of the value of $10.10 in 2016, measured in 2011 dollars. We don't know what inflation will be going forward. I could be really fancy and use inflation expectations or TIPS spreads. Instead I just used an inflation calculator to see the 2008 dollar value of $10.10 in 2013 and I'm calling that the 2011 dollar value of $10.10 in 2016. So it's NOT perfect (2009 was a rough year), but it's ballpark. If you want to figure out something sturdier, please be my guest.

As it stands, without thinking about what is binding, this is certainly a huge jump relative to past increases.

I don't think that's the end of the story, though, because if you want to know whether the minimum wage is binding you really do have to think about trends in productivity, because that's what ought to determine labor demand. EPI has also done excellent work tracking what the minimum wage would look like if it had grown since the late sixties at the rate of productivity growth. Now productivity growth as it is measured is NOT the same thing as marginal productivity that we use in theory, but I think it's a fair proxy to consider when we're ballparking this sort of thing. And of course we don't have productivity information for minimum wage workers (although BLS now breaks it out by industry - perhaps someone should take a look at productivity growth in typical minimum wage industries). This is what EPI has on that:



Now we might not want to index the minimum wage today to the 1968 minimum wage. But even if you want to take the mid-nineties as your starting point - a period that a lot of research has covered and where the real value of the minimum wage was lower - productivity is outpacing real minimum wage growth by a substantial margin. Even if we halve the gap between the productivity line and the real minimum wage line, because we think that a 1968 baseline is too unrealistic a goal and minimum wage workers might not have as fast growing productivity anyway, and it still seems like the productivity-growth-consistent binding wage level is above $10.00 in 2011 dollars.

So Bob might not be satisfied because this is not a single clean answer, but I think I can narrow it down to four points!:

1. Yes, the jump in the real wage seems to be very big and in that sense is out-of-sample relative to the empirical tests we've been tossing around, but
2. The empirical tests don't tell us where the binding minimum wage is, and
3. When we look at productivity data - which should indicate something about where the binding point is, it seems like we still have a fair amount of running room before we hit this point, nevertheless,
4. Every labor market is different. DLR's average null effects have a distribution associated with them, so we can expect that many low wage labor markets (for example, in the South) would experience a negative effect from the minimum wage - even more so from an increase to $10.10.

So what is to be done? Well I think Bob Lerman (American University and the Urban Institute) has some good points to make in this interview:
"Lerman believes that wages will increase for low-income workers if you make them more skilled and valuable. “You can try to increase wages by mandate, but if you do that without doing anything on the productivity side, there will be fewer workers,” he says. To build a stronger labor force, Lerman has promoted apprenticeships and founded the American Institute for Innovative Apprenticeship. 
“You can get maybe six weeks of training to become a simple welder. Or you can have a serious apprenticeship program that gradually moves you into robotic welding and programmable stuff that makes people much more valuable,” he explains. “Just as you can be a cook at McDonald’s or move into a higher value-added culinary activity.”"
Acting on worker productivity, as Lerman suggests, will increase labor demand. Of course we can work on labor demand directly as well, through fiscal policy and direct employment measures (Jon Wisman, also of American Unviersity, talks about this later in the article). As this post suggests I don't know what to say about whether the minimum wage increase will be "modest" or not. But I really think that's the wrong question - at least it's the wrong policy question. If we have to ask that question maybe there are better ways to help low income workers than raising the minimum wage.

Point #4 about differences between local labor markets has bothered me a lot over the last year. I'd love to work more with this question, although I don't have the time now. State legislatures and local governments are in a much better position to assess what a modest minimum wage increase would be for their labor market, and if they make a mistake, the costs of the mistake will be less broadly distributed. So to the extent that there is action on the minimum wage I think it should happen at the state and local level. It just seems wiser than federal minimum wage increases.

* - You might still think there is an acceptable trade-off between wage increases and disemployment effects. This will be determined by your own policy preferences, labor demand elasticity, and what other policies you want to see acting in concert with the minimum wage. Let's put this prospect aside for the moment, and just say that when the minimum wage is binding in a disemployment sense we can legitimately start to worry about the immodesty of the increase.

1 comment:

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