Consumer preferences for grocery purchasing during the COVID-19 pandemic: a quantile regression approach

Abstract

PurposeThe objective is to determine if consumer preferences for grocery purchasing are impacted by the severity of the COVID-19 pandemic and whether these preferences differ by demographic and psychographic characteristics.Design/methodology/approachThe authors conduct an online discrete choice experiment (DCE) with 900 U.S. consumers to assess grocery shopping preferences under various scenarios of the COVID-19 pandemic (i.e. decreasing, constant and increasing cases). The attribute of interest is the purchasing method (i.e. in-store purchase, in-store pickup, curbside pickup and home delivery) with minimum order requirements, time windows and fees as secondary attributes. Heterogeneity in individual-level willingness-to-pay (WTP) estimates for the main attribute is analyzed by means of mixed logit and quantile regression techniques.FindingsThe mixed logit model reveals heterogeneity in WTP estimates for grocery purchasing methods across participants. According to estimates from quantile regressions, the heterogeneity is partly explained by the severity of the COVID-19 pandemic. For example, the home delivery purchasing method is less preferred when the number of cases is decreasing. The results also show that consumer preferences for grocery shopping methods are affected more by psychographic characteristics than demographic characteristics. Consumers who comply with COVID-19 directives (e.g. wear face coverings) have stronger preferences for curbside pickup and home delivery, particularly at the tails of the WTP distributions.Originality/valueAlthough there is much data on food consumer behavior during the COVID-19 pandemic at the aggregate level, there are few analyses of grocery shopping preferences at the individual level. The study represents a first attempt to relate individuals' demographic and psychographic characteristics to their grocery shopping preferences during the COVID-19 pandemic, thus yielding numerous recommendations in terms of consumer segmentation.