Does Price Sensitivity Vary with Sleep Duration? A Cross-category Analysis

Abstract

Lack of sleep, a threat to both physical and mental health, is increasingly prevalent. Yet, the influence of sleep insufficiency on consumer behavior has rarely been studied. To the best of our knowledge, price sensitivity, one of the most studied phenomena in marketing, has not been associated with physiological factors such as sleep. Consequently, we ask does sleep duration affect price sensitivity? We answer this question by measuring price elasticity using household-level data across 18 grocery categories from 2006-2016. As a later sunset time is negatively correlated with sleep duration, we use the sunset time (by week) for each household (by zip code) as a proxy for sleep duration. For each product category, we estimate a random coefficients nested logit model with category incidence at the upper level and brand choice at the lower level, using control functions to address price endogeneity. Our estimation models also control for brand specific effects, state dependence, features/displays, and seasonality. The identification assumption is that short-term local sunset time variation only affects price sensitivity through its impact on sleep duration after conditioning on the control variables in the model. Our preliminary key result, which holds across most of the product categories we study, is that a reduction of sleep tends to increase price sensitivity. This result largely is related to changes in brand choice. Research on other factors underlying this relationship is currently being pursued. Beyond sleep, our results suggest that a greater understanding of the relationship between price sensitivity and physiological factors is worthy of future research.