Or do people just like alliteration? "Fight for Fifteen!" fits nicely on a protester's sign and sounds great as a chant. President Obama argued for a $10.10 minimum wage saying "It's easy to remember." Not, "It's the optimum value for this economic variable." Not, "We did our due diligence and arrived at the number $10.10." Not even, "We did a back-of-the-envelope and came up with the number $9.73, but then rounded it up a little bit." The proposal has changed dramatically over the past several years. Obama first called for a federal minimum wage of $9, then $10.10. James Galbraith called for a $12 minimum in 2013 (he describes his proposal here around the 50-minute mark), but now he's at $15. Has inflation changed the appropriate minimum wage in a few short years? No, inflation has only been at around 2% a year, not nearly large enough nor operating for long enough to turn a nominal 12 into a 15 (that would take about 15 years at 2% annual inflation). Has there been a recent flood of high-quality minimum wage studies that have killed off the downward-sloping demand curve for good? Not exactly. There's been some fine work on the minimum wage in recent years, but most of it in the same vein as the dueling "there's a disemployment effect/no there isn't" studies of the past 20 years. If anything, the new work on Seattle shows that all those disemployment effects are very real and in fact lead to low-wage workers earning less because of reduced work hours.
What's changing here is that political "entrepreneurs" are realizing they can get away with making bigger and more extravagant proposals, completely unanchored by underlying fundamentals. This is incredibly irresponsible. If these politicians and academics want to meddle with the labor market, they need to do a much better job of justifying their proposals. Politicians who propose higher minimum wages are riding a wave of populism. Irresponsible academics reassure them that minimum wages have no effect on unemployment and no other ill effects, contrary to what a straightforward reading of the literature actually says.
Illinois recently raised its minimum wage, which will be raised in steps to $15/hour. I contacted my representative to ask if there was some kind of policy whitepaper or report that justified the number 15 and the specific phase-in schedule. She gave me a stock "I always appreciate hearing from my constituents" answer, without actually answering my question. She referred me to Will Guzzardi, a representative from Chicago, who apparently sponsored the bill. In other words, she did not base her vote (which was in favor of the bill) on any kind of information that might have conceivably justified raising the minimum wage. If she'd had such a report, she would have pointed me to it.
My message to Guzzardi went unanswered. But I can find public statements by him here:
Again, this is unbelievably irresponsible. He is simply mistaken about the state of minimum wage research. Even if some academics have sold the "minimum wages don't reduce jobs" story, that is at best a highly selective reading of the literature. See, for example, what Arindrajit Dube says in his testimony to congress. Dube is an economist who does a lot of empirical work on the minimum wage and fairly consistently finds no effect on unemployment. Does he say, "The minimum wage has no effect whatsoever on unemployment, so we can raise it arbitrarily high."? No, of course not. Even though some careless people (like Guzzardi) give the impression that there's no effect on unemployment, no matter how high we jack up the minimum wage, almost everyone implicitly understands that raising it high enough will cause unemployment. (Go for the reductio ad absurdum and see how people demur: "Why don't we raise it to $100/hour then? Or $1,000, or $1 million, for that matter?") The economists who support raising the minimum wage understand that raising it high enough will cause unemployment. Even laypeople and politicians who have no grounding in economics understand at a gut level that raising the minimum wage high enough causes unemployment. The disagreement is about whether the effect is measurable, whether it's too small to matter, and whether the gains to the winners justify the losses to the losers. Dube says in his testimony,Guzzardi has repeatedly cited research showing no damaging economic effects where the minimum wage has increased. He said the only way to predict what will happen with an 82 percent wage hike in six years is to look at historical data.“Raising the minimum wage has no net effect on employment, it doesn’t drive jobs out of the state,” Guzzardi said. “All it does is put money in people’s pockets who need it.”
A natural target to set the minimum wage is to half of the median wage.Why not higher? Because Dube implicitly understands that the disemployment effects are real. These ill effects will rear their ugly heads if we raise the minimum wage too high. At some point it becomes hard to make the negative effects disappear using statistical tricks and inappropriate datasets.
[Dube said of an increase to California's minimum wage from $10 to $15/hour: "California's experiment is worth running and monitoring." Presumably he wouldn't refer to it as an "experiment" if he were certain that there were no negative consequences to raising the minimum wage. He's hinting at the very real possibility that such a large change in the minimum wage will cause unemployment.]
There is an unfortunate confluence at work here. Some minimum wage researchers are studying the effects of the minimum wage and finding no effect. They can't reject the null hypothesis, that minimum wages don't cause unemployment in low wage workers. They have done a poor job of differentiating between, "We've positively proven that minimum wages have no effect on unemployment" and "The effects are real, but our datasets and methods are too crude to measure them. Noise overwhelms signal, so we wouldn't see an effect even if there was one." The second story is a much better summary of the recent contrarian research on the minimum wage. But some economists are leading the lay public to believe that the former statement is true. People like Dube implicitly endorse the latter when they pick a finite minimum wage (half the prevailing median wage), rather than picking one that's arbitrarily high. Politicians and political commentators only absorb a very crude version of this. They don't understand that the economists are measuring the effects of small differences in minimum wages (say, the $1 or $2 difference between two neighboring state's minimum wages, or before and after a law change), not the total effect of the minimum wage (compared to what we'd see in a counterfactual world where it was $0). They don't understand that the economists are using crude proxies for "minimum wage workers", such as "employed teenagers" or "all restaurant workers." To the extent that some of these workers earn substantially more than the minimum wage, these kinds of studies will understate the effect of the minimum wage increase on their wages and on their subsequent employment/unemployment. The proxy approach effectively dilutes the true effect of the policy. They don't understand that most minimum wage studies are seriously under-powered, meaning they wouldn't find an effect even if there was one. Policy "entrepreneurs" take a crude version of the lessons from recent research and use it to sell bad medicine to the public. Economist who favor increasing the minimum wage (or even those who lend intellectual respectability to the idea) need to be more explicit about what they actually believe: that demand curves slope downward. If you make something more expensive (in this case, low-skilled labor), people naturally want to buy less of it.
That's my commentary on the academic side of the argument. There is a crude approach to justifying the $15/hour number, which I consider completely unserious even though it's fairly common. It goes something like, "We need to make sure everyone can earn a living wage" or "Everyone should be able to earn a wage that covers the cost of living." This is superficially compelling, but it falls apart under even a tiny amount of scrutiny. Terms like living wage and cost of living get thrown around without much thought. Do people who earn less than a living wage literally die? Do they wallow in Dickensian poverty for a few years, wasting away until starvation, disease, and exposure catch up to them? Not really. That's basically unheard of in developed countries like the United States. The term cost of living is wonderful propaganda. Taken literally it means, "The bare minimum necessary to ensure raw survival." That's not what it means in practice, though. Based on how the term is actually used, the "cost of living" usually means "the cost of the basket of goods and services, food, shelter, and medicine, that we've become accustomed to." It varies across regions in within the United States, not always because it's inherently more costly to live in some places but because the residents are used to a higher standard of living. Immigrants often come from relatively poor countries and live in conditions that the average American would regard as crushing austerity. These same immigrants are often trying to save their earnings (often "low-wage" earnings, but very high compared to what they could earn back home) to send remittances back home. For them, indulging an average American lifestyle would mean depriving their loved ones at home a substantial fortune. The literal "cost of living" is very low, a few thousand dollars a year maybe.
This argument about employers owing their employees a living wage is very strange in the sense that it seems to have bizarre implications that nobody actually believes. (I owe this argument to Jonathan Meer, see here and here.) Suppose a working mother has a baby. Her cost of living has increased. Does her employer owe her a raise? Suppose her working husband loses his job. Or suppose he dies. Is the employer supposed to be her all-purpose insurer for all these contingencies that potentially affect her cost of living? I think most people would say "No." Even forgetting for a moment that this would make every employer a giant insurance company and turn every single employee an enormous liability, it doesn't sound right that employers hold this moral obligation to fix their employees lives when their conditions deteriorate. Let's at least be upfront about this: the living wage argument is weird, because it has implications that most people would reject as logical absurdities. [If you're tempted to say, "No, I'm not saying any of that weird stuff about employers being obligated to insure their employees against all hits to their income. I just want a $15/hour minimum wage.", then you might have missed the point here. If your moral premise implies something logically absurd or utterly impractical, you don't get to quarantine off those inconvenient implications. It's not cricket to assert a premise, but then pick and choose which implications of that premise you like. It requires that you either reject the moral premise, bite the bullet, or legitimately explain away the implications you don't like.]
All that aside, this stuff about living wages assumes away whether employers can actually afford those higher minimum wages. This living wage framing of the argument simply side-steps all the arguments about minimum wages being counter-productive, throwing people out of work, making it harder to find work in the firs place, and reducing their work hours. If the Seattle results are reliable, then the minimum wage is the wrong policy tool for addressing grinding poverty. It's as if someone proposed bloodletting to cure cancer, and then a skeptic points out that bloodletting is not an effective cancer treatment. It's actually counter-productive to drain a sick person of some of his blood, one might reasonably protest. It would be irrational for the bloodletting advocate to accuse the skeptic of not caring, or to reiterate that cancer is a very serious problem indeed, so we have to do something about it. This kind of response would completely miss the point of the criticism: the treatment is not effective. Whether the underlying problem is real or whether there might be some other working solution are different questions entirely.
Another argument, which I also consider flippant, is that we should increase the minimum wage because it's popular. I guess under this framing, the "$15" doesn't have to come out of any kind of cost-benefit analysis. We should just do it because it's what popular opinion dictates. This argument doesn't make sense for a lot of reasons. Popular policies are often the wrong thing to do. Read Bryan Caplan's The Myth of the Rational Voter for a book-length treatment of this idea. Thank goodness there are checks and balances on "the will of the people," because the general public is often sadly misinformed about the effects of policy. The minimum wage is no exception to this pattern. If anything it's the archetypal example. There is no good reason that a number (such as "10.10" or "15") that the public happens to anchor on would tend to be the right number. Governments have an obligation not to simply impose democratically popular ideas on the population. They should use their bureaucracies to research such proposals and provide necessary information to the voting public and policy-makers. (Washington state's commissioning of the Seattle studies is a nice example of this process at work. Let's cost-benefit analyze more government programs, please.) Insofar as the "fight for 15" crowd has achieved policy success, governments have fallen down on the job here.
No comments:
Post a Comment