E.COMBINED LIKE Cracking numbers and building models to guide policy makers. But who leads them in turn? Isaiah Andrews from Harvard University tried to help. On April 20, the American Economic Association awarded him the John Bates Clark Medal, an award for leading economists under 40 years of age, for his efforts.
Mr. Andrews did not expect to become an econometrician (an economist specializing in statistics). However, during his studies he was drawn to questions about teaching and the limitations of data. With Matthew Gentzkow from Stanford University and Jesse Shapiro from Brown University, he proposed a way to evaluate the quality of economists’ models. He also showed economists how to figure out how much their results might change under alternative assumptions, and helped them stave off allegations that their results were a feature of their assumptions rather than the data.
Working with Maximilian Kasy, now at Oxford University, Andrews shed light on the fact that economists are too eager to get remarkable results and that journal editors are too inclined to approve them. For example, studies that find that minimum wages have no significant influence on employment are only published a third of the time than studies that find a stronger negative effect. The researchers developed a way to adjust the reported results for this publication bias – and found that the average effect of a higher minimum wage on employment decreased by half.
Andrews, along with others, has also explored what is called the winner’s curse when it comes to choosing between guidelines: the guideline that performs best in a process may owe its highest rank to chance and later become a disappointment be condemned. To illustrate this, the researchers turn to a study that evaluates the most effective ways to encourage people to donate to charity by combining requests for specific donations with promises that match the original contribution. The researchers find that the charity always overestimates its donations when choosing the method that works best in an experiment. They suggest ways to make a more realistic estimate that takes into account the role of chance.
Overall, according to the award committee, Andrews contributed to “redirecting econometrics to the investigation of the most important problems of empirical research”.
This article appeared in the Finance & Economics section of the print edition under the heading “Model behavior”.