According to the economists' analysis of externalities, a substance should be taxed, and in extreme cases banned, only if it raises the earnings and other benefits to users at the expense of harm imposed on others. For this reason, excessive drinking is fined and punished also in other ways, since heavy drinkers are more likely to get into accidents that harm others. Similarly, a substance should be subsidized if it benefits others while benefiting users. Posner's example of substances that encourage greater innovation fits this category. Substances that raise benefits to users by the same amount as it raises their overall productivity should be neither banned nor encouraged.
In economies with competitive labor and products markets, individuals tend to receive earnings and other benefits that equal their "marginal product"; i.e., equal their contribution to total output. Substances or anything else that raise the productivity of individuals who receive their marginal product should be neither taxed nor subsidized. For example, sleeping pills enable users to get a better night's sleep, which allows them to get to work on time and have a clearer mind while at work, and thus become more productive while at work. There is no reason to tax or subsidize this use of sleeping pills because individuals taking the pills would generally be paid the full increase in their productivity, and they would bear the full costs. The vast majority of substances that people take to affect their cognition fit into this category of meriting neither ban nor encouragement since they primarily affect the productivity of users without having significant external effects on others, either positive or negative.
Exceptions are substances that raise the benefits of users in situations that have a "zero sum" aspect, so that the user's gain corresponds to about an equal lose to others. Sports competitions have important zero sum aspects since fans care a lot about who wins and loses in addition to the quality of the play, so that increases in the performance of every player without changing outcomes brings relatively little benefit to fans. This importance of relative performance to fans provides the justification put forward in our discussion on sports doping on August 27, 2006 for major league baseball and other organized sports leagues to place limits and bans on the use of certain drugs among their players. For these drugs may have long run negative consequences for the health of users without raising fan welfare by much.
However, I do not see any reason why governments should be involved in enforcement of any bans imposed by sports leagues, and it was absurd for Congress to hold hearings on whether either Roger Clemons or his former trainer were lying. Governments may be helpful to sports leagues in enforcing their ban by punishing violators, but governments might also be helpful to companies in getting their workers to reduce absenteeism. Still, that is no reason for governments to be punishing workers who do not show up for work because there are no externalities outside the workplace of the individual company or league that warrant government intervention.
Other examples where relative performance helps to determine benefits include entrance to college based on test scores, such as the SAT test, or patent races to see who comes up first with the technique or process that several competitors are seeking to discover. South Korea and other countries have tried to use laws to cut down on private tutoring and other investments that increase the likelihood that a student may succeed in gaining entrance to top universities, where the number of acceptances remains constant. Presumably, these countries would want to ban students from taking various stimulants that improve their performance, perhaps at a risk to their health, but such bans are difficult to enforce.
Some economists have claimed that superstar situations involving singers and other entertainers, novelists, money managers, and lawyers have this zero sum character since a person can make a huge income by being only slightly better at what he does than others. Since fans and customers prefer to listen to the best singers or have their money managed by funds that seem to be the best, someone who is only a little better than the competition can attract a very large fan base, or a lot of money to be managed. They may earn only a little from each fan or on each dollar invested, but make it up through the millions of fans they have and the billions of dollars they manage.
The superstar phenomena discovered by my late colleague Sherwin Rosen is real and important in modern societies with huge economies of scale in communication, but it is not a zero sum situation. A superstar still only collects the value of his net contribution to output. The value is large not because of externalities, but because a superstar may only add a little utility to each fan, but he adds thislittle to each member of a huge fan base.
All in all, even aside from enforcement issues, I see little reason for governments to ban the use of Provigil and other stimulants that improve cognitive performance. There are some situations where this improvement mainly benefits users at the expense of harm imposed on their competitors. For the most part, however, potential users are the best judge of whether they should use stimulants since they bear the lion's share of the costs as well as receive the benefits.