With talk about increasing the minimum wage up to $10+/hour, the econ blogs have been alive with minimum wage chatter the last week or so, and purely coincidentally one of my students presented his senior research on the effects of minimum wage legislation last night.
And I no longer know where I stand.
I assume that labor is an ordinary good subject to the law of demand, just like shoes. That is, as the price of labor goes up, the demand for it goes down. Ceteris Paribus, of course. But the evident effects of minimum wage increases are generally shown to be rather small. Perhaps, as a grad school friend of mine used to say, ceteris ain’t paribus. That is, maybe other economic factors swamp the effects of a minimum wage increase.
Another factor, perhaps, is that the opposition to minimum wage increases generally manages to limit them. Proponents of increasing the minimum wage frequently argue that inflation has eroded the value of the minimum wage, so bringing it up to $X now simply restores it to what it was some number of years ago. If all we ever do is restore it to what it once was, rather than increasing it in real dollar terms, the effect should be fairly small.
But two things make me wonder about that. First, my student showed a graph showing job growth in Alaska before and after a minimum wage increase. In Jan 2003 Alaska increased the minimum wage, and there was an immediate dip in employment, but short-lived; by summer employment was back up to about its previous level. But the rate of increase was lower; the upward sloping curve was flatter than previously. Now, I don’t know what all was going on in Alaska in early 2003, so I can’t assert confidently that this was all due to the minimum wage increase. But I can point out that the raw employment numbers wouldn’t show much effect even if the rate of growth decreased. So just looking at the employment numbers won’t necessarily tell us the real story. Even the unemployment rate won’t necessarily tell us, because it doesn’t include people not looking for work, and if fewer jobs are made available due to higher labor costs, fewer people may be prompted to look for work even though the law of supply tells us that higher pay ought to lead more people to look for work (certainly the lack of obvious declines in unemployment rates following minimum wage increases indicates they don’t cause much influx into the labor market, and again, perhaps that’s because the increases are relatively small).
Second, Arnold Kling makes what seems to me a reasonable argument about the effects of phasing in minimum wage increases.
Giving people more time to adjust to incentives normally leads to larger adjustments, not smaller. If you suddenly raise the gas tax, for example, there is very little effect on gas consumption. But if people expect the gas tax to go up years before the higher tax kicks in, many will buy more fuel-efficient cars, leading to a large behavioral response. Minimum wage hikes should work the same way: Employers’ long-run response should exceed their short-run response…
If the minimum wage unexpectedly jumped to $12 today, the effect on employment, though relatively small, would be blatant. Employers would wake up with a bunch of unprofitable workers on their hands. Over the next month or two, we would blame virtually all low-skilled lay-offs on the minimum wage hike – and we’d probably be right to do so.
If everyone knew the minimum wage was going to be $12 in 2015, however, even a large effect on employment could be virtually invisible. Employers wouldn’t need to lay any workers off. They could get to their new optimum via reduced hiring and attrition. When the law finally kicked in, you might find zero extra layoffs, because employers saw the writing on the wall and quietly downsize their workforce in advance.
So maybe the employment effect of minimum wage increases is larger than I thought, but just less discernible?