Sunday, February 04, 2007

Minimum wage laws

Steven Landsburg, a smart economist, wrote an interesting article in Slate, “The Sin Of Wages: The Real Reason To Oppose The Minimum Wage”.

Landsburg’s article makes two valid points. First, that increasing the minimum wage has little impact on employment. Second, that minimum wage laws are unfair because they constitute a tax on a small segment of society that should be paid, if paid at all, by all of society.

According to Landsburg:

Ordinarily, when we decide to transfer income to some group or another—whether it be the working poor, the unemployed, the victims of a flood, or the stockholders of American Airlines—we pay for the transfer out of general tax revenue. That has two advantages: It spreads the burden across all taxpayers, and it makes politicians accountable for their actions. It's easy to look up exactly how much the government gave American, and it's easy to look up exactly which senators voted for it.

By contrast, the minimum wage places the entire burden on one small group: the employers of low-wage workers and, to some extent, their customers. Suppose you're a small entrepreneur with, say, 10 full-time minimum-wage workers. Then a 50 cent increase in the minimum wage is going to cost you about $10,000 a year. That's no different from a $10,000 tax increase. But the politicians who imposed the burden get to claim they never raised anybody's taxes.

If you want to transfer income to the working poor, there are fairer and more honest ways to do it. The Earned Income Tax Credit, for example, accomplishes pretty much the same goals as the minimum wage but without concentrating the burden on a tiny minority. For that matter, the EITC also does a better job of helping the people you'd really want to help, as opposed to, say, middle-class teenagers working summer jobs. It's pretty hard to argue that a minimum-wage increase beats an EITC increase by any criterion.

Landsburg is right that this is reason enough to oppose minimum wage laws. However, it is worth noting why minimum wage laws may not have a material impact on unemployment.

It is clear that a large increase in minimum wages, say to $100/hour would cause widespread unemployment, other things equal. This establishes that, other things equal, any increase in minimum wages causes some unemployment. If other things really are equal, the fact that increases in minimum wages have not caused material unemployment only goes to show that an immaterial percentage of workers have minimum wage jobs.

But are other things equal? Over the years, the cumulative increase in the minimum wage has been dramatic, with no apparent impact on unemployment. One possible explanation is that the real impact of increasing the minimum wage by X% is a dynamic reaction leading over time to inflation of X%. In this case, workers earning below the minimum wage have no ultimate benefit from increasing it. Another possibility is that productivity increases reach X% over time offsetting the adverse unemployment impact of the minimum wage increase. In this case, workers end up with the same, higher, wage they would have gotten anyway. Or, there could be a combination of the two.

In either case, eventually, the increased minimum wage is detrimental (if you think inflation is detrimental) and has no impact on the people it is designed to help.

Saturday, February 03, 2007

Racial profiling or good policing?

The New York Post article “NYPD frisked blacks at 5 times the rate of whites” notes that of total frisks in 2006, 52% were Black, 29% Hispanic, and 10% White.

Does this indicate racial profiling?

Does this indicate racially biased policing?

Consider that the arrests resulting from these frisks were 51% Black, 30% Hispanic, and 12% White.

The fact that the arrest and frisk percentage breakdowns by race were almost identical implies that the probability of an arrest, given a frisk, was the same for all races. Assuming that the arrest decisions were made without regard to race (something that remains to be determined), then the implication is that the frisk decisions were racially unbiased and not due to adverse racial profiling. This is not to say that there was no racial profiling, only that, if there was, it was statistically relevant.