Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
We introduce a novel method to elicit belief distributions and apply it to elicit inflation expectations in a representative US sample through a pre-registered survey experiment. Our approach elicits beta belief distributions directly in a two-step process. First, participants specify their minimum and maximum inflation. They then use a graphical interface with two sliders to adjust the mean and variance of their inflation belief distribution. We benchmark our method against the “Bins” method, popularized by the New York Fed’s Survey of Consumer Expectations (SCE). Our findings reveal significant variations in elicited belief distributions depending on the method used. Specifically, our approach yields higher mean inflation estimates and substantially reduces the standard deviations of the distributions. Respondents report that our method is easier to use and more engaging. Furthermore, the resulting distributions more accurately reflect participants’ beliefs across several dimensions and show stronger correlations with their point predictions.