Krugman Calls for Increasing the Minimum Wage. Here’s Why He’s Wrong.
And the evidence is overwhelmingly positive: hiking the minimum wage has little or no adverse effect on employment, while significantly increasing workers’ earnings.
It’s important to understand how good this evidence is. Normally, economic analysis is handicapped by the absence of controlled experiments. For example, we can look at what happened to the U.S. economy after the Obama stimulus went into effect, but we can’t observe an alternative universe in which there was no stimulus, and compare the results.
When it comes to the minimum wage, however, we have a number of cases in which a state raised its own minimum wage while a neighboring state did not. If there were anything to the notion that minimum wage increases have big negative effects on employment, that result should show up in state-to-state comparisons. It doesn’t.
So a minimum-wage increase would help low-paid workers, with few adverse side effects. And we’re talking about a lot of people. Early this year the Economic Policy Institute estimated that an increase in the national minimum wage to $10.10 from its current $7.25 would benefit 30 million workers.
He’s wrong and here’s why:
1) State vs. state cross-sections are still not controlled experiments, by a long shot. There are about a trillion kazillion million (if you were doubtful of my trustworthiness, how do you feel now?) ways one state is different from another. And yes, many of these ways will be relevant for the question at hand (whether the minimum wage affects unemployment maliciously or not).
2) Even if they were close enough to controlled experiments, there is no reason to assume that a historical fact, that a raise in the minimum wage had little or no effect on unemployment in one time period, should carry over to another time period. Human beings are not inanimate objects like the subjects of the physical sciences. Human beings have the ability to choose, and these choices can change over time. Economics deal with human beings, and as such, cannot use historical findings to create some sort of law set in stone. History by definition, has to do with specific dates and times and specific people and specific choices. They are not universals. We say “on April 12, 1950, Jimmy went to the store!” The date and the name and the action all can change as time goes on. The same applies to historical claim regarding the effect of an increase in the minimum wage.
3) Even if we let #1 and #2 slide, it is STILL a wrong statement. We cannot assume that the effect of one increase in the minimum wage, from some amount to $7.25, carries over to another amount, $7.25 to $10.10.
The empirical method applied to economics sucks, plain and simple. There are too many weaknesses with it. Economics is not a natural science, it is a social science. We can’t even point to the success of empiricism like scientists can in the natural sciences. Astronomers can point to correct predictions of objects such as comets, e.g. if we look in the sky at this time, we’ll see the comet right here. All that economists, meanwhile, have to point to, is blatant failures, from terrible predictions about what would happen to the economy (this is right before the Great Depression), to an inability to stop the financial crisis, to this piece of work (this one’s the funniest by the way).
The correct method of economics is not empiricism but praxeology. The social sciences can only lay claims to laws that do not have to deal with particular choices of humans, but the universal characteristics of all human action.
Just anticipating a possible response. Some might say “oh but there could be reasons to assume that something that occurred in history would happen again.” I’m not saying this is impossible, but I’m saying it’s bad reasoning for a science dealing with human action, particularly when other methods are available.
Secondly if you’re wondering “how do you think the minimum wage affects unemployment?,” see here.