Towards a sustainable urban mobility: comparing online and in-store shopping choices

In recent years, e-shopping has gained increasing popularity, with more people gradually shifting from traditional shopping channels to online platforms causing significant impacts on city sustainability due to small, frequent, sprawled, and failed deliveries. In fact, due to the necessity of using sometimes-inefficient delivery trips to deliver products to consumers (such as at their residences), this can have a substantial influence on freight traffic in metropolitan regions. Using data from interviews with 509 respondents carried out in Sardinia (Italy) in 2022, the current study investigates how end consumers’ choices between online and physical (in-store) shopping are related. In doing this, two different econometrics models for simulating online and in-store shopping were constructed: a multivariate ordered probit model to understand which covariates influence the propensity to purchase different kinds of products online and in-store; a binary probit model to identify who is more likely to reduce the number of trips due to e-shopping. From the descriptive statistical analysis, it emerged that a majority of individuals in the sample (62.3 %) reduced their number of physical shopping trips due to e-shopping (substitution effect). The multivariate ordered probit model shows that socio-demographic characteristics, land-use attributes, and psychological variables significantly influence shopping behavior. Specifically, the perception of online shopping accessibility and quality positively correlates with the likelihood of purchasing certain product categories online. Conversely, the perceived importance of touching products and in-store safety positively affects in-store shopping preferences. Additionally, positive correlation terms among online and in-store shopping tendencies for the same product categories suggest that consumers inclined to buy certain items online are also more likely to purchase them in-store. The binary probit model highlights substantial heterogeneity in the likelihood of reducing physical shopping trips. Individuals with more experience shopping online, higher perceptions of online quality, and lower importance placed on touching products are more likely to reduce in-store visits. From a policy perspective, this study emphasizes the need for urban planners and policymakers to integrate consumer shopping behavior into strategies aimed at managing urban mobility, logistics, and last-mile delivery systems.