Analyzing the Minimum Wage: Economics and Empericism

Yesterday, I posted my first piece on this series exploring the minimum wage and demonstrating why utilitarian arguments are so powerful. Today, I will delve into the utilitarian idea surrounding the minimum wage, and consequently an economic analysis of this policy.  Economic theory is an excellent way to understand the consequentialist impacts of a policy in marketplace, but this post will also cover the limitations of theory vs data (for the data analysis, see my subsequent post).

The Economics of the Minimum Wage

Economics allows us to understand how market actors and institutions impact the distribution and exchange of resources.  By making certain assumptions about how market actors behave, economics can predict the outcomes of policies, predictions that might be counterintuitive to someone not versed in the theory. But economics must also respond to empirical data; unfortunately, the science of economics has trouble conducting controlled experiments because they’re too expensive or would be unethical with human subjects (e.g. “Shock Therapy: An Inquiry into the Undergraduate Volunteer Test Subject Supply Curve”). This section will explore the many economic perspectives surrounding the the minimum wage.

The general layman intuition of the minimum wage runs as follows: employers generally have greater market power than poor employees (employers can find different workers easier than employees can find a different job) and thus a minimum wage is a fair rebalancing of the power in this relationship, helping employees gain more leverage.

Basic microeconomic theory, however, turns this concept on its head. Economists assume that suppliers will sell more at higher prices, while buyers will buy more at lower prices, creating Demand Curves which slope down and Supply Curves which slope up. The price at which the buyers and sellers transact occurs where the two curves intersect:


Public domain image, Wikipedia.

A minimum wage is a type of price floor–it sets, by statute, the lowest possible price one can put on hourly labor. To have an effect, the price floor must be placed above where the market price would be otherwise, thus in the diagram, the minimum wage is a horizontal line above the equilibrium. Theory dictates that with such a policy, more workers would be willing to work at that higher wage.  If you’ve ever heard the argument that a minimum wage hike increases worker morale, this chart illustrates that idea mathematically, as an upward sloping labor supply curve implies more workers would be willing to work at that higher wage level.  But problems occur when we look at the demand curve, as at this new higher wage, employers are less willing to employ as many workers than in the equilibrium situation, and far fewer than the amount that want work at this new wage level.  Microeconomic theory thus predicts that a minimum wage or a raise in the minimum wage would result in unemployment.  It also predicts that at any given minimum wage, if that wage was above equilibrium, there would be an excess of workers, meaning that the market power of employers has actually been enhanced by the minimum wage.  With so many extra workers, employers would have little incentive to provide any benefits for their employees, since they could be replaced so easily. This creates political incentive to raise the minimum further to fix the balance of worker vs employer market power, which theory predicts would only make things worse.

Since the introduction of the wage, this was pretty much the debate: economists said you should not raise the minimum wage, while political reality meant the minimum wage was continually increased in nominal terms by Congress. In real terms, the minimum wage has seen a 20% decline since its peak in the 60s.  This is often used as a talking point by wage hike supporters that the minimum wage should be higher since it was higher in the past. I will address this below when discussing the empirical data.

Source: Dept. of Labor

Lots of opponents of the minimum wage stop the story here. I’ve even heard Professor Walter Block, someone who has been a great voice for libertarianism, simply state that we’ve done the math, the minimum wage is bad, and anyone who says otherwise doesn’t know their econ and probably has some sort of political motivation. I argue that this is not the case, and this position makes minimum wage opposition seem much weaker than it should be.

Empirical Data

In 1992, David Card and Alan Krueger studied the employment effect on fast food restaurants during a minimum wage increase in New Jersey, using neighboring Pennsylvania as a control group.  Their paper is linked here.  They found no evidence for an increase in unemployment in New Jersey over Pennsylvania.

This is significant because it was one of the first times that minimum wage theory was tested.  The puzzling result inspired many more studies on the subject to better understand the forces acting on employment and the minimum wage.

Here it is worth noting that Economics is not a science in the same vein as chemistry or biology no matter how mathematically rigorous it appears to be; a core part of the scientific method is controlled experimentation, something that is fairly difficult to do in a real economy.  As a result, economic studies are observational: they do not test hypotheses, and often data is viewed through a lens of pre-conceived theory.

Preview of the Data

Since 1993, there have been a lot of studies on the minimum wage. The fact that virtually all studies are observational (due to the limitations of economic science) means these studies simply try to provide possible explanations for the data, and definitive answers can be tricky to nail down.  Moreover, a common argument (I believe first espoused by Elizabeth Warren in 2013) points out that had the minimum wage kept up with where it was in the past, it would be much higher now.  Of course, since virtually no empirical data was gathered before 1993, there seems little reason to trust policymakers from the 1960s who had no access to the vast trove of data we have today.

Turning to an example of the murky art of drawing policy conclusions from the data, try taking a look at these two posts, one from the leftist Roosevelt Institute and the other from the conservative Daily Signal (published by the Heritage Foundation).  Oddly enough, they both refer to the same studies: David Card and Alan Krueger’s studies are the most commonly referred to by the left, while the right prefers David Neumark and William Wascher.  Despite talking about the exact same studies, these posts are almost intractable.  And this is not because some of these academics are just crazy partisan hacks: they hail from respected institutions including, UC Berkeley, Princeton, UC Irvine and the NBER, and the Federal Reserve. And they are in significant disagreement over the employment effects of the minimum wage.

A quick sidetrack to see if popular political websites might have anything insightful to add in terms of economic papers cited reveals that mainstream debate on this topic (ThinkProgress, HotAir) really couldn’t be bothered with peer reviewed studies.  Keep it up guys, I’m sure the other side will give in if you just post a few more opinion poll results!

What is needed is a harder look at the empirical findings of economists, which can reveal what actors in the market really do when confronted by a minimum wage policy. Unfortunately that data is far too large for this post. Next time, I will walk through an in-depth analysis of the data published in the last 20 years.