In his 2013 State of the Union, President Obama proposed a $9 federal minimum wage, indexed to inflation. Here to discuss the minimum wage as a policy is Arindrajit Dube.
Dube is a professor of economics at the University of Massachusetts Amherst and a widely respected scholar of labor markets and the minimum wage. Along with T. William Lester and Michael Reich, he is the author of Minimum Wage Effects Across State Borders: Estimates Using Contiguous Counties (2010), a major study that found no adverse employment effects of minimum wages increases by studying counties that cross state lines.
Dube has written a summary back in 2011 on the state of this research here. Narrow, technical issues have dominated so much of the debate on the minimum wage, so I wanted to step back and get a better understanding of the minimum wage as a policy mechanism. These remarks are lightly edited.
What does an introductory-level Economics 101 textbook tell us about the minimum wage, and how does that contrast with how advanced scholars approach the topic?
In the Economics 101 textbook model, there are a couple of effects we can expect from an increase in the minimum wage in a competitive labor market. It's going to increase the cost of production and, as a result, employers may be expected to cut back on employment. But they'll also be able to pass on some of the costs to consumers. So to what extent, even in a competitive market, the minimum wage causes a loss in employment, as opposed to an increase in price, is theoretically ambiguous.
However nowadays when economists—labor or otherwise—study the labor market, they not only think about employment levels but also about flows. There are a lot of things about the labor market that don't really fit really well with the simplest competitive model of the labor market. In reality there are good jobs and bad jobs, and workers try to get to a better job whenever they can. So there's a lot of turnover and churning in the low-end of the labor market.
To the extent that the minimum wage makes the lowest paid jobs better, it tends to reduce turnover and reduce vacancies. So an increase in the minimum wage may not kill jobs but kill vacancies in a low-end labor market. This is consistent with the more realistic models of the labor market. Our new work shows this for the U.S., but there is evidence from other countries as well. So minimum-wage laws may make jobs more stable while raising wage.
Your research argues that there aren’t employment effects from raising the minimum wage. But why bother increasing the minimum wage when we could just increase the Earned Income Tax Credit (EITC), which also supports the wages of low-wage workers?
There are some opponents of minimum wages who point to the EITC as a better alternative. But here's the thing. Research by Berkeley economist Jesse Rothstein shows that roughly 27 cents on the dollar from the EITC is passed on to employers. So there's some leakage there. And for some people, the presence of EITC acts as a multiplier for a hike in the minimum wage.
So as a result, when thinking in terms of efficacy, these two policies may complement each other. They may go together.
Right. David Lee and Emmanuel Saez have also theorized that, when employers capture part of an EITC, a minimum wage is a good compliment. And there’s also leakage from the EITC to “fast cash” refund loans too. But does the minimum wage really help the poor? I've heard it isn't well targeted as a program.
Now they both have leakages of various sorts. The point on the minimum wage may be somewhat overplayed. But certainly it is the case that there are a lot of poor people who don't have jobs, and so the minimum wage is not going to be able to help them.
Food stamps are a lot better at aiding the very needy who aren’t even in the labor force than a minimum wage, because it makes sure they have basic sustenance. But that doesn't necessarily mean that we should have food stamps only and not a minimum wage. It's a false choice.
So the first part of the answer is thinking in terms of a portfolio of policies that policymakers should consider.
Another point goes back to the nature of low-wage labor markets. We understand that the low-wage labor market has a lot of turnover and frictions. A minimum wage can reduce churning while not reducing employment. So those together point to the idea that a higher minimum wage not only increases wages but may also improve the functioning of the labor market.
There are reasons to believe that, for instance, in a high turnover environment, firms may be less willing to invest in training. There are good theoretical reasons why that would the case. With a higher minimum wage reducing turnover and increasing job stability, they might be more incentivized to do that which may increase productivity.
So what is the right way to think of how the minimum wage is targeted in regards to workers?
There are two issues here. How much of the gains go to people who are poor? But there’s also a second issue of how much of the money goes to people who are near poor. From a certain perspective, even if a relatively small but not negligible gain go to the poor, we aren't doing very badly given how few of the poor are in the labor market.
If the minimum wage is going to be good for working-class families, but not necessarily poor families, I'm not entirely sure why we should think of that as being a leakage of the policy. We have different polices designed for different distributional goals. We need to think not in terms of a single policy, but instead think in terms of what is the right portfolio of policies given the range of objectives you have.
So if going to a $9 minimum wage is such a good idea, why not go to a $40 minimum wage?
Is there ever too much of a good thing? I think for most things the answer is yes. The range of minimum-wage increases we have seen in the United States, the evidence speaks pretty clearly to there not being job losses, and also gains to workers.
So what is that range? Let's go back to 1968. In 1968 the minimum wage was around 50 percent of the average production-worker wage. If that were true today, that minimum wage would be around $10 an hour. So from that perspective, raising it to $9 is well within, indeed under, the historical norm between the minimum and average wage of production workers.
Another way to think about this is looking at other countries. There are many countries in the developed world that have a minimum wage that is as, or even more, generous than 50 percent of average wages.
Another way to think about this is to consider what the purchasing power of the minimum wage. The 1968 minimum wage, adjusted for inflation, today would be somewhere between $9.25 and $10.50 depending on what CPI index you use. Again what the Obama administration is proposing is safely below this.
We can't say, based on historical evidence in the United States, what a $40 minimum wage would do, but we can say a lot about what a $9 or $10 minimum wage would look like.
In addition to a $9 minimum wage, President Obama proposed indexing it so that it goes up with the cost of living. How does that complicate a minimum wage?
First of all, nine states already have an indexed cost-of-living increase for the minimum wage.
One thing that I would do is makes wage changes more predictable. At the federal level, we have gone through extended periods when the nominal minimum wage has been stagnant, then there's a big fight about it followed by a sizable increase, and then it stays there for a while. I don’t know of any economic model which says this is a good way to go. This creates uncertainty about the timing and extent of the wage increases, and it generally not a great way of setting the minimum wage, no matter what level you think it should be.
We could disagree about the level of the minimum wage, but it seems like a no-brainer to me that no matter what level we pick, we should have a relatively smooth adjustment process. A natural adjustment process is tying the rate of cost of living increases, like many items including Social Security payments. It makes a lot of sense. Daniel Hammermesh—a well-known labor economist who in general is not in favor of high minimum wages—has nonetheless come out supporting indexation.
Let’s say that we increase the minimum wage to $9, and it turns out that your research showing no job losses wasn’t correct or didn’t hold. What’s a bad case scenario for workers who would be affected?
Here's an important thing to keep in mind. What does the minimum wage do in a competitive labor market? If the law of demand holds without any frictions then employment falls. But by how much? And does it mean that if it falls, we shouldn't do it? Absolutely not. Before the debate that started on minimum wages in the mid 1990s, when the consensus of the discipline was that there were detectible effects of minimum wages on jobs, many economists still supported the minimum wage!
And their idea was that the costs were there, in terms of jobs, but the benefits were nevertheless higher. Now the debate has shifted to whether there are any costs in terms of employment, while there is still the benefit.
So even if some jobs were lost, a person who might have a smaller probability of employment is also going to have a higher wage if he or she were to be employed. As a result, it is not at all clear that they'd be worse off. When you balance these things, including less churning in the labor market, it is easy to imagine workers being better off even if there are job losses.
As a result, it is interesting that surveys of workers show a clear preference for an increase in the policy.
Anything else you’d like to share on Obama’s proposal?
Full disclosure: if this passes, it would directly hurt my career as an applied micro-econometrician! We'd have little variation in minimum wages with which to run more regressions. But the variations we've seen, the kind I've done my work on, don't make much economic sense. They are driven by political cycles, and business cycles, in ways that are counterproductive. So while we can disagree about what the level of the minimum wage should be, putting it on a predictable adjustment process is something we should all be able to agree on.