Contractual solutions for employment law problems.

• Author: Epstein, Richard A.
• Position: The Federalist Society National Lawyer's Convention: 2014

It is my very great pleasure to speak at this general session of the Federalist Society on the vexing topic of employment law. It is a subject in which I have had a long-term interest, and on which I have done a fair bit of legal and theoretical work. In a previous talk, Commissioner Chai Feldblum of the Equal Employment Opportunity Commission emphasized the importance of empirical work in order to get a grip on how the system works. Her implicit assumption is that we should be skeptical of theoretical accounts of the determinants of success and failure in labor markets. No one should of course turn a blind eye to empirical information, but the emphasis that is attached to it all too often becomes a recipe for total and absolute stasis. The data is always incomplete. The studies are often too esoteric. The chains of inference are long and disputed. Rival interpretations are often inconsistent, and hence those who like the status quo ante use the claim for empirical research as a barrier against a revisiting of policies that I regard as far too interventionist in labor markets. No one did any empirical research to introduce the current melange of employment law. No one needs to do a great deal of empirical work to start down the road to dismantling the structure.

In making this claim I do rely on some gross empirical data, which is readily available from standard sources, and which makes sustained calls for more empirical research unwise and unnecessary--at least if its purpose is to guide reform efforts. The basic numbers of youth unemployment are very grim, and show that this situation is widespread with special pressure on members of minority groups. The available data does not paint a pretty picture. (1) Indeed, when one looks at the youth unemployment rate today (in this instance for sixteen- and seventeen-year-olds) we find that the numbers are virtually indistinguishable from the period between 1950 and 1955, when there were no protective laws on the books and systematic forms of segregation existed. (2) Starting in 1955 or so, the unemployment picture for both black and white males got worse, and the gap between black and white employment rates rose dramatically, even as the minimum wage compressed wages. What is striking about this early data is that the changes during this period of time, including an increase in minimum wage laws, had a differential effect between the groups, with a worse effect on black males. The simplest explanation is that earlier unemployment rates may have been roughly equal because the wage rates may have been lower for black workers, reflecting their weaker job qualifications at the time. This is a classic case in which institutional changes left both groups worse off.

The question then is what drives these changes, and here the theory tells a picture that is all too clear: regulation. The simplest graph of the effect of the minimum wage law shows that if the wage is constrained, supply will exceed demand at that level. This change means that the number of employment contracts formed will be fewer than would be the case if the restraint were removed, which would cause wages to fall and employment levels to rise. Labor markets are devilishly complex, but one thing that is sure today is that there are many forms of labor restrictions today not in place in the 1950s that help account for increased unemployment rates.

Most critically, it is important to be conscious of where the gaps in our empirical knowledge lie--it is with the uncertainty as to the relative harm that is done by different measures, acting alone or in tandem. For example, it is very difficult to understand the adverse impact of a minimum wage law. One complication is that the impact depends on the gap between the market wage and the minimum wage. If the former is above the latter, the effects will be small, because wages can reach equilibrium. But there may still be workers in a robust market who are only employable at wages below the minimum wage, even if they constitute a small fraction of the total. With some increments in minimum wages, other adaptations of the employment contract could offset the burden of the regulation: those workers could be required to work on split shifts, pay for their own meals, and the like in order to dull the effect of the high minimum wages. At this point the adverse effect of the minimum wage law will reflect itself in the reduction of joint employer-employee surplus, not in the unemployment rate. But social losses also matter, even if they are much harder to measure empirically. That said, none of these complications have to be resolved before making the simple but sensible recommendation that the removal of all these restraints, which often operate in tandem, should help move us toward fuller employment.

There is another way to put the point. A great deal of evidence suggests that in general, competitive markets outperform monopoly markets. Therefore, markets that have barriers to entry are those with monopoly-like characteristics, which impede efficiency. The simple theoretical point here is that competitive markets essentially exhaust all the gains from trade. Where there are many people on one side of the market buying or selling widgets, people will search for the ideal trading partner. The fewer the impediments to entry and exit, the more likely it is that people will start to reach ideal kinds of results. You will never get to perfection because transaction costs are always positive, but they can be reduced, and I like to think that the Coase Theorem stands for the proposition that the best way to improve social welfare is to minimize transaction costs. (3)

It is also important in this context to be careful in choosing the definition of a competitive market. Apart perhaps from a few commodity markets, there are no perfectly competitive markets anywhere. Labor markets are no exception to that rule. Workers are not perfect fungible in the way that shares of stock are fungible. Nor are they mobile in the way capital is mobile. As Professor Verkerke stressed in his remarks, ordinary people have other identities that interact with their status as employees. People have spousal issues, family attachments, personal disabilities, and all sorts of unique characteristics that impact their individual choices. But no matter how significant these personal features, they do not undermine the simpler point that for any given worker there are many job opportunities, just as for any given employer they are many different job applicants. The slippages that one observes in labor markets are of theoretical interest but of relatively little practical importance. Competitive labor markets are never, for example, the subject of an antitrust inquiry on the ground that some firm has dominant control in a...