Statistical Problem of Minimum Wage and Poverty

Original Article:

  • Author: Michael Saltsman

  • Publication Date: March 2013

  • Newspaper: Politico

  • Topics: Minimum Wage

On Thursday, Sen. Tom Harkin (D-Iowa) will pick up where President Barack Obama left off in his State of the Union address, hearing testimony on a newly introduced bill to raise the federal minimum wage.

Obama and Harkin differ on the specifics — the president had asked for a $9 minimum wage, while Harkin has joined with Rep. George Miller (D-Calif.) to call for $10.10. But both are united in their belief that a higher minimum wage can reduce poverty without reducing employment.

They’re wrong on both counts.

Let’s start with the State of the Union. In his prime-time pitch for a higher minimum wage, the president referenced 19 states that currently have a wage floor that’s set above the federal level. He left out the 28 states that raised theirs between 2003 and 2007, and with good reason: Economists from Cornell and American University found no trace of lower poverty rates associated with these wage hikes.

It’s a statistical problem that continues to haunt even the staunchest minimum-wage advocates, mainly because the facts just aren’t on their side. A majority of our country’s poor don’t have employment and thus won’t benefit from a raise, while large numbers of minimum-wage earners don’t live in poor households. It’s the same with the Harkin/Miller proposal: The average family income of someone receiving a raise under this bill would be $52,396. More than 35 percent are living at home with family or relatives; by contrast, just 9 percent represent single parents supporting children.

In other words, the minimum wage is poorly targeted at the poor. If that were the only problem, the policy might still make sense as part of a larger poverty-reduction package. But empirical evidence published in the Journal of Human Resources suggests that a minimum wage hike can actually move more people into poverty than it moves out of it.

Here’s why: An increase in someone’s hourly wage won’t translate to an increase in their annual take-home pay if they lose hours or employment as a result. That’s exactly what’s happening: Employers who keep just 2 cents to 3 cents in profit from each sales dollar (think: restaurants or grocery stores) can’t just absorb a 39 percent hike in labor costs. They either have to raise prices, or — more likely, given their cost-conscious customer base — find a way to provide the same product with less service.

This means fewer hours of work and fewer opportunities for less skilled groups like teens, who already face a 25 percent unemployment rate. That’s why a new Employment Policies Institute analysis of Census Bureau data finds that roughly 988,000 jobs would be lost because of the Harkin/Miller proposal, with 30 percent of the lost jobs occurring in the retail industry and 29 percent occurring in accommodations and food service.

Harkin, Miller and Obama have directed legislators concerned about such consequences to a comforting study authored by economists affiliated with UC Berkeley. (One of them, Dr. Arindrajit Dube, is scheduled to testify at Thursday’s hearing.) Their study claims that, contrary to the vast majority of published economic literature on this issue, wage mandates do not reduce employment.

As the saying goes, if it sounds too good to be true, it probably is. The work of Dube and his co-authors received a devastating reply from a team of economists at the University of California-Irvine and the Federal Reserve Board. (It’s available now via the National Bureau of Economic Research.) Having analyzed and rechecked all the analysis in the original Berkeley studies, these economists determined that “neither the conclusions … nor the methods they use are supported by the data.”

Of course, empirical data have always taken a back seat to rhetoric in the campaign to raise the minimum wage. Witness the claim that a higher minimum wage is necessary to catch us up (adjusting for inflation) to the 1968 minimum wage. A quick check of Bureau of Labor Statistics data shows that 1968 was an inflation anomaly — the minimum wage would be only about $4 an hour today if properly adjusted since its inception in 1938.

But whether or not the facts are on wage hike advocates’ side, their campaign will continue. Democrats have said they plan to use it as an election issue in the 2014 midterms, and polling data suggest that — absent an intervention — their strategy might be successful. That means it’s time for those who appreciate economic common sense to counter the “Time for $10.10” campaign with the message that $10.10’s a dead end.

Michael Saltsman is research director at the Employment Policies Institute.