Comparing the Time Horizon of Retrospective and Prospective Economic Evaluations

• In Event: Improving Decision-Making

Abstract

When evaluating the past, voters often appear myopic. In the context of retrospective economic assessments, people prioritize recent over earlier, more temporally distant economic conditions. This is consistent with findings in both behavioral economics and psychology about future decisions. People prefer sooner payments to payments days or weeks in the future unless the future payments are drastically larger. They put off unpleasant tasks, even if doing so only makes the tasks more arduous. These findings suggest a similarly compressed time horizon in both directions: People take a short view looking back, so they will also take a short view looking ahead.

There is a second possibility, however: Powerful heuristics operating in retrospection—a focus on the “end” of the time series or increasing trends in economic conditions—predict the opposite: The end is far away in prospection, and the trend points into the future. Research in psychology has shown that people are capable of looking into the future; for example, they like to save better experiences for later while getting less fun experiences over with. If people rely on these heuristics, we should see future-orientation and patience rather than impatience in prospection.

To discriminate between these two predictions, we conduct experiments where respondents were asked to evaluate economic growth over a four-year period. We manipulate whether they were evaluating past growth or future proposed growth. Our initial results, based on an MTurk sample, show that people have a preference for increasing trends of growth in both conditions. They prefer more growth sooner when looking to the past but are willing to delay economic growth when looking to the future. Together, these results suggest that existing characterizations of economic evaluations as short-sighted and myopic are at least in part due to only looking at retrospective evaluations.

Authors

• Markus Prior, Princeton University

• Talbot M Andrews, University of Connecticut