We evaluate secure survey methods designed for the ongoing monitoring of harassment in organizations. We use the resulting data to answer policy relevant questions about the nature of harassment: How prevalent is it? What share of managers is responsible for the misbehavior? How isolated are its victims? To do so, we partner with a large Bangladeshi garment manufacturer to experiment with different designs of phone-based worker surveys. Garbling responses to sensitive questions by automatically recording a random subset as complaints increases reporting of physical harassment by 288%, sexual harassment by 269%, and threatening behavior by 46%. A rapport-building treatment has an insignificant aggregate effect, but may affect men and women differently. Removing team identifiers from survey responses does not significantly increase reporting and prevents the computation of policy-relevant team-level statistics. The resulting data shows that harassment is widespread, that the problem is not restricted to a minority of managers, and that victims are often isolated in teams.
