Willingness-to-pay (WTP) experiments have been widely used to assess demand for a variety of products. Do they also generate persistent treatment effects? We answer this question using a randomized controlled trial of a baseline WTP experiment, combined with in-person and phone survey data over a four-year period. We find that a simple experiment leads to positive and persistent effects on adoption and usage of an improved storage technology, as well as disadoption of traditional technologies. These results are primarily driven by households who experienced the product, rather than information or salience. Failing to account for demand elicitation experiments conducted at baseline may affect the external validity of the broader experiments in which they are embedded.
