CBO estimate on additional unemployment from raising minimum wage is probably wrong

By Marc Jampole

The announcement by the bipartisan Congressional Budget Office (CBO) that raising the minimum wage to $10.10 an hour will lead to 500,000 jobs disappearing is just wrong. Now I’m not saying that the report’s authors are lying or stupid, just that they are making the wrong assumptions or looking at the numbers in the wrong way.

The CBO admits that its numbers are a guess at best, but that admission is buried in the fine print.  In Appendix A to its report, CBO says that it reviewed a large body of research on what it calls “employment elasticity.”  In economics, elasticity is how sensitive one economic variable is to another. A simple example of elasticity is price: If we double the price of a product, how many fewer people will buy it? Clearly, doubling the price of something that people need or want very badly such as milk or a college education will have less impact on demand than doubling the price of a luxury, such as a meal at a fancy restaurant or jet skis. Or consider what would happen if twice as many people suddenly wanted something of which there was a finite amount, like gold in times of economic turmoil. But twice as many people wanting Doritos might not lead to such a dramatic rise in the price of each box, because, as the man says, “We’ll make more.” Many factors go into creating a relationship between two economic variables.

One truism of economics is that the higher the price the lower the demand. Economists accept it as a given, much as mathematicians accept as a given that the shortest distance between two points is a straight line. But in some cases, demand is more elastic (stays the same or close to the same unless there is a steep price increase), as college education has proven to be.

“Employment elasticity” measures how the market will respond when the minimum wage is raised (or lowered).  But all the measurements use methodologies, each of which employs a different series of variables and makes a different set of assumptions. Appendix B of The CBO’s report lists dozens of methodologies and studies of methodologies that it considered. Some concluded that raising the minimum wage would have a negligible effect on unemployment; some said as many as a million jobs would be lost.  Instead of weighing the relative merits of each, the CBO took an average. That’s what the 500,000 is—the midpoint in a bunch of generated by a  bunch of different methodologies. And the methodologies only measure teen unemployment! CBO uses another set of methodologies to infer the effect of raising the minimum wage for the entire economy based on what happens to teens, our most inexperienced and unskilled workers.

I’m inclined to believe the studies that show a minimal impact on unemployment by raising the minimum wage to $10.10 and seven Nobel prize-winning economists, four former presidents of the American Economic Association and more than 600 other economists agree with me.  The Nobel laureates and professors are going to give you their mathematically based models. Let me tell you what happens in the real world.

In the real world, a business increases its profit by either growing its market or lowering costs. Every successful company is constantly looking at both. Part of cost-cutting is to make sure your labor costs are as low as possible. So in theory—let’s call it the efficient company theory—no business ever hires someone they don’t need and can’t make money from because it would unduly raise costs.  In a like manner, no business ever lets someone go just because they cost too much without replacing that person because they need someone to do the job (having not hired too many to begin with).

Now the real world is a little messier, as the following examples suggest: A large business may reevaluate labor needs every six months (or two years) and within the six month period have too many (or too few) employees but hasn’t gotten around to making the routine adjustment. A small business owner may have lost a large share of her business, but is reluctant to lay off people in case business turns around in a month or two. A new labor-saving technology is on the market and a business is evaluating it.  A company will not refill a position once someone retires. Or how about the real estate profession during the bubble that was destined to end? In all these cases, companies are carrying excess labor, and a rise in the cost of labor may make them change their minds quickly.  These job losses aren’t caused by the increase in wages. The job loss was activated before the increase in wages.

I would consider these job losses to be noise or leakage in the economic system, similar to the concept of the natural rate of unemployment, which is the unemployment rate that occurs with full employment, stemming from the fact that there are always people looking for jobs and employers looking for workers. I’ve read that the natural rate of unemployment used to be 4% but is now higher.

What I’m saying is that any increase in unemployment because of an increase in the minimum wage is nothing more than noise (or friction as Milton Friedman and others called it). I’m not saying that the noise or friction that prevents companies from being one hundred percent labor efficient is that same 4+% as the friction that explains why unemployment never falls below a certain floor. But even if it were only one tenth as much, that would still compute to hundreds of thousands of jobs that these methodologies may be counting as losses because of a rise in the minimum wage to $10.10. The noise factor almost certainly explains why some estimates are so much lower than others.  

The 500,000 who CBO is predicting will lose their jobs if the minimum wage reaches $10.10 would raise unemployment by three tenths of a percent. That’s 500,000 out of more than 166 million jobs. So even if CBO is right, it sounds like they are measuring noise for the most part, which in this case means the number of jobs that would have been lost no matter what.