The Minimum Wage - A Perpetual Motion Machine?

I recently tweeted a link to a story about a bookstore in San Francisco that announced it would be closing because the owner couldn’t afford to pay his employees the new, higher minimum wage ($15 an hour after being full phased in over a few years). The irony in the story is that the owner, a devout liberal, thinks the minimum wage hike is a good idea even though it killed his business. (At least he has the courage of his convictions, so kudos to the guy for consistency.)

That link precipitated a brief back-and-forth with a twitter friend regarding the effectiveness of minimum wages and the effect of such laws on employment. Because a limit of 140 characters isn’t really a help in parsing complicated subjects, I thought I’d expand on those points here.

First, a great deal of research has gone into understanding the relationship between increases in the minimum wage and changes in employment. It’s a tough problem for a number of reasons:

• The vast majority of workers earn more than the minimum wage, so small increases in the minimum don’t directly affect those workers
• Changes in employment because of national events (e.g., a recession or recovery) or local events (e.g., jobs flowing to Silicon Valley or away from the Rust Belt) often swamp small changes in the minimum wage
• To the extent that changes in the minimum wage lag changes in actual wages, and thus the law isn’t a binding constraint on employers (that is, they’re already paying more than the minimum), as has often been the case, one wouldn’t expect much, if any, change in employment.

Most studies have concluded either that small changes in the minimum wage have no statistically significant effect on employment or that such an effect, while negative, is small. Few studies show that employment actually increases, a result that would be sharply at odds with basic economic theory. Almost invariably, making something cost more reduces the demand for it. The question is not whether higher (binding) minimum wages reduce employment, but how much such laws reduce employment. And my emphasis on small changes in the minimum wage is important: I presume that everyone accepts that sufficiently large increases will have a substantial impact on employment; otherwise, we’d just impose a $100/hour wage and everyone would be rich.

Another thing to keep in mind is the composition of which workers earn the minimum. In general, they are young, often still in school and in their first jobs, and often working part time. Full-time workers generally move quickly from a minimum-wage position to something paying above the minimum. This has important implications for public policy, as I’ll discuss below.

The logical argument in favor of increasing the minimum wage is that the small decrease in employment is more than offset by the greater purchasing power of the vast majority who keep their jobs, and that the social welfare net will backstop any job losses. To put it in economists’ terms, demand for (unskilled) labor is sufficiently inelastic that making such labor cost more doesn’t change the quantity demanded much.

Does this make sense? Keep in mind who earns the minimum wage: high school kids and adults with few marketable skills. Businesses don’t hire those people because they bring a lot to the business; businesses hire them because they’re cheap. What happens when they’re not as cheap? Businesses cut back. The summer hire who cleans out the stockroom might not be affordable any longer. When thinking about the value the business gets from having a cleaner stockroom, the owner may decide it’s not a good deal any more. The young woman who handles the cash register may no longer be affordable; customers will just have to wait in slightly longer lines. Business owners may increase their own hours instead of hiring now-more-expensive labor. Bigger businesses may decide to increase their amount of automation, such as having customers order food from an iPad rather than from a waitress.

And here’s the thing that few people keep in mind: most businesses aren’t going to fire an employee if the minimum wage increases. Instead, they just won’t hire the next employee. That job will simply never exist. But how can anyone keep track of that? It’s easy to keep track of existing jobs that are lost, such as those six employees of the San Francisco independent bookstore, but it’s impossible to collect anecdotes of jobs that would have been profitable to create under the old wage but are no longer profitable under the new wage.

I often hear a “fairness” argument in favor of higher minimum wages. It costs a lot to live in the city, and it’s not “fair” to employ someone at, say, $7.25 an hour. I have to reactions to that. First, if people are willing to offer their labor for $7.25 an hour, they clearly view the compensation as enough to take the job. Although people get emotional about labor, this decision is no different than saying a Mercedes is overpriced because at a lower price more people would drive one. I don’t have much patience for fairness arguments, because what’s fair to you might not be fair to me. If the government wants to insert itself into any market, including the labor market, it should have a sound basis for doing so. (Other government actions, such as how progressive the income tax schedule should be, or how much food stamps should be worth, are clearly and unavoidably subjective.) Second, while the people who remain employed are clearly better off, the people who fail to get jobs are not only worse off today, but may well remain worse off forever. The high school graduate who can’t get her first job can’t use that as a stepping stone to a better job. Thus the long-term implications of lower employment, which are even harder to measure than the immediate effect, may be much larger than the immediate implications.

One last thought: it’s one thing for the voters in, say, Seattle or San Francisco to vote for higher minimum wages in those cities, but it’s an entirely different thing to impose higher minimum wages everywhere, i.e., in a national law. In Seattle, for example, a high cost-of-living city, it may be the case that almost everyone already earns at least $15 an hour, so the law is not terribly binding, so the effect of the law is minimal. However, employers in low cost-of-living areas, where the prevailing minimum wage is, say, $7.50 an hour, would face a doubling of labor costs, so workers in those areas would be disproportionately hit by the law.

If increasing the minimum wage made almost everyone better off, it would be the economic version of a perpetual motion machine. Back in the real world, though, it’s just another price control, and works just as well as other efforts to monkey with market forces.