Evaluating Precision, Privacy, and Representation with Cell Phone Data

Abstract

Mobility data are a promising tool for valuing recreation, but reliability depends on how data are processed and protected. This study evaluates how data practices affect recreation demand estimation. Using a random utility model of visits to Cape Cod beaches from 2019 to 2022, we estimate the marginal willingness to pay (MWTP) for avoiding fecal bacteria contamination. Under recommended practices, MWTP averages $8.92 per visit. Relaxing dwell-time thresholds and applying differential privacy reduce MWTP by 57% and 65%, while short-term choice sets inflate estimates by 10%. These results highlight the need for transparent mobility data practices.