Gender Differences in Accurately Estimating Business Metrics

Abstract

Self-reported data are the most common way to measure the performance of small businesses, especially in developing countries where businesses tend to be informal and few owners keep records. However, we do not know how men and women differ in evaluating business measures that are notoriously difficult to measure objectively. How do men and women differ in their accuracy to estimate business measures? In this paper, we conduct a pre-registered field study using a representative sample of small-business owners in India. We ask business owners to report the amount of cash they have on them to surveyors, which corresponds to about three-quarter of daily revenues. We benchmark this cash report to an objective measure that was obtained by physically counting how much cash the respondent had. We find women are 14 to 20% more likely than men to report accurate guesses. This result holds after controlling for the amount of cash on hand and various business characteristics. We use behavioral games to test for mechanisms but these explain at best a small part of the gap. We additionally show that the magnitude of the self-reported gender profit gap varies considerably based on the time horizon used in the survey, from 11% to 29%. This gender reporting gap raises concerns about the use of self-reported data for business measures. These differences could also considerably underestimate the gender gap in business performance, especially in developing countries.