What does macroeconomic policy look like when people aren’t perfectly rational?

One of the leading criticisms of economics is the field’s reliance on perfectly rational agents in many of its models. Read the assumptions behind a macroeconomic model and the rules under which households and businesses are presumed to make decisions. Chances are high that they will seem fanciful if not outright ridiculous. Of course, simplifying assumptions are always going to be needed to make models workable, but “perfect rationality” seems like a jump too far—not least because behavioral economics research shows how individuals very often fail to act in rational ways.

Is there a way to integrate the insights of behavioral economics into the assumptions that inform a macroeconomics model? Well, there is! A new paper does just that.

The paper, updated this month, by economist Xavier Gabaix of Harvard University, integrates behavioral economics findings into what’s known as a New Keynesian macroeconomic model—one that includes microeconomic decisionmaking and markets that aren’t perfectly competitive and that is the kind of model often used by central banks. The specific insight that Gabaix adds to the model is “bounded rationality,” where a person’s ability to make a decision is bounded by cognitive limits. One way to think of this is an individual who isn’t trying to make the best decisions but rather find the decision that is satisfactory. Part of these cognitive limits is that individuals or businesses will be myopic, or care a lot more about the present than the future. This is in contrast to perfectly rational agents who knows exactly how events in the future will impact them.

What does the inclusion of bounded rationality mean for a macroeconomic model? Gabaix’s model leads to a lot of interesting results, including fiscal policy that is much stronger and a resolution of a debate about the effects of interest rates on inflation. (See Bloomberg View columnist Noah Smith’s piece on an earlier iteration of the paper for a fuller explanation of this second debate.) Two other findings, however, are of particular interest to monetary policymakers.

Consider how central banks have recently made a turn toward emphasizing their future plans for monetary policy, reflecting a similar movement in academic macroeconomics. These ideas about communication include forward guidance, or “communication about the likely future course of monetary policy,” as well as policy tools that create commitments for the central bank, such as targeting nominal gross domestic product. But in a world where businesses or individuals aren’t that concerned about the future because they are myopic, these policies won’t have as much power as they would in a world of perfectly rational agents. Gabaix’s model explains a puzzle other economists have noted—that forward guidance isn’t as effective as other models would have us believe, and that nominal GDP targeting isn’t the optimal commitment policy for monetary policy.

Gabaix’s model is a powerful sign of how changing the assumptions behind models can change our understanding of what’s the best possible route forward. Even if these two monetary policy tools —targeting nominal gross domestic product and communicating the likely future course of policy —aren’t optimal or as powerful as other models indicate, they still could be quite effective. Further research that brings the insights of behavioral economics into macroeconomic modeling would be very valuable to policymakers.