Consumption as a social signal: upward/downward visibility and income elasticities

I’m hoping to eventually help say something about the origin of preferences. In past work on conspicuous consumption I showed that a survey-based measure of expenditure visibility—how quickly survey respondents state they would notice if someone around them spent more than average on different consumption categories—helps explain cross-commodity variation in total-expenditure elasticities in U.S. household data; I interpreted this as suggesting that social signaling can provide some guidance regarding a theory of Engel curves. But many questions remained open, including questions of causality and of signal symmetry—to what extent does spending more than others on a visible commodity sends a different signal than not spending less than others on it? In this project, I report and analyze results from a series of new visibility surveys. I randomly assign 1,223 U.S. respondents (from the 18+ population) into four treatments, each designed to measure a different aspect of expenditure visibility, and I show that the different treatments uncover rather different aspects of visibility in the sense of incrementally and jointly predicting the cross-expenditure variation in elasticities (again estimated from U.S. expenditure data). I discuss implications and potential modeling.