People who “think like economists” recognize that we have to trace the unintended consequences of both individual action and the policies of governments. Good intentions are not enough to ensure good outcomes. We have to understand how the action or the change in policy would interact with the knowledge and incentives facing choosers in order to understand the actual outcome.

This is particularly true with respect to economic policy and can be illustrated by the world of competitive sports, which often makes a nice analogy to markets given that both involve competition according to rules.

The current discussion about changing the rules for overtime in the NFL, at least for the playoffs, provides a good example. Right now, the rule for overtime is sudden death: The team that scores first wins. Of course this means the coin flip to determine who gets the ball first is crucial—for many people it’s too decisive: The first possessor wins too often. In addition, other changes to the game since overtime was introduced in 1974 have given the team with first possession an easier path to scoring, especially by a field goal. (For example, the kickoff line is five yards farther back and field-goal kickers are stronger and more accurate).

One proposal would allow the loser of the coin flip to have possession of the ball if the team that has it first only scores a field goal. The intent is to reduce the advantage of first possession by reducing the value of the field goal.

As with economic policies, however, we cannot ignore how this change might affect the teams’ choices and strategies.

Fooled by Inflation?

Consider the history of macroeconomics. Early Keynesians believed that inflation could fool workers into accepting lower real wages (thereby increasing employment) because they would not understand its effects. When Milton Friedman, Robert Lucas, and others later pointed out that people eventually learn, it changed how economists viewed loose monetary policy. If people know inflation will reduce their real wages, they will have an incentive to find out the inflation rate and do their best to incorporate it into their wage demands. As a result, inflationary policies will not have the nearly the effects on the average real wage that the early Keynesians thought.

Observe the analogous reasoning about the proposed NFL rule change. If the first possessor only gets a only field goal, the second team would know it has to score to keep the game alive and therefore would never punt on fourth down. It turns out that statistically the second possessor is more likely to score in such circumstances than on a typical NFL drive. So to that extent, it seems the rule change would have its intended effect.

But we don’t want to assume that the first possessor would do the same thing no matter the rules. After all, it would know that if it only scores a field goal, the other team would be more aggressive on fourth down, which would make it somewhat more likely that it would tie or win the game than under sudden-death rules when the first team doesn’t score. This knowledge may well induce the first team, on the margin, to risk a fourth-down play, passing up a field goal, to try for a game-winning touchdown.  Reducing the marginal value of a field goal to the first possessor would encourage more teams, again on the margin, to try to score a touchdown, and thereby counteract some of the intended consequences of the rule change. The change would be less effective than first thought when we consider how the first possessor might react.

In economic policy this point can matter quite a bit. As Friedman and Lucas put it, people’s expectations are not invariant to the policy regime. People form expectations that counter the intended effects of the policy. The sports version is: Game strategies will change when the rules change. In both cases a thorough analysis of the actual effects of a policy or rule change requires that we assume that people learn and react to the change itself. Just changing policies or rules without assuming people learn is a recipe for ineffectiveness and potentially damaging unintended consequences.