E-shoppers and multimodal accessibility to in-store retail: An analysis of spatial and social effects

Abstract

Amidst the growing interest in enhancing the academic understanding of the relationships between e-shopping and transport, a key element remains underexplored – the impact of e-shopping on spatial accessibility to in-store retail. The paper studies variations in multimodal accessibility to in-store retail between e-shopper groups and the associated spatial effects. The research is based on a face-to-face questionnaire, administered in the city of Alcalá de Henares (Madrid Metropolitan Area, Spain), which provides data on socio-economic characteristics, e-shopping habits, and travel time preferences to reach in-store retail. Clustering techniques serve to identify three e-shopper groups: occasional e-shoppers with a car, infrequent e-shoppers with a car, and frequent e-shoppers without a car. A comparison of e-shopper distance-decay functions to reach in-store retail is made, revealing significant differences between the three e-shopper groups for car and public transport for any time interval. However, for walking such differences are limited to time intervals between 10 and 40 min. Distance-decay functions are processed through a gravity-based model, identifying five categories of multimodal accessibility places that provide information on how in-store retail locations may be affected by e-shopping habits. The paper closes with concluding remarks on policy-making and a few pathways for future research.

Introduction

These are challenging times for in-store retail activity at the city level. The unprecedented growth of e-shopping rates is generating both competing and complementing dynamics between e-shopping and in-store retail, with an impact on person-time travelled (Farag et al., 2006, Farag et al., 2007) and the in-store landscape (e.g., in-store retail closures) (Maat and Konings, 2018; Zhen et al., 2018). For example, frequent e-shoppers are likely to spend less time to travel to in-store retail, being able to access a limited number of such locations (Ferrell, 2005). That could originate a closure and re-location of certain in-store retail, negatively affecting population groups less-familiar with e-shopping (Ariza-Álvarez et al., 2019; Olsson et al., 2019). Furthermore, infrequent e-shoppers could be forced to travel longer distances to reach in-store retail. The COVID-19 pandemic has further intensified this trend (OECD, 2020), with a feasible increase of spatial accessibility variations in the long-term (van Wee and Witlox, 2021). Given the spatial and social effects of the described phenomenon, enhancing academic understanding of how multimodal accessibility to in-store retail varies between e-shopper groups is particularly pertinent.

An e-shopper group is conceived as a set of individuals with similar socio-economic characteristics and e-shopping habits. The particularities of each e-shopper group depend on their individual priorities, needs, and circumstances (Jaller and Pahwa, 2020; Rotem-Mindali and Weltevreden, 2013). The scientific approach of multimodal relative accessibility seems appropriate for gaining insight into how spatial accessibility to in-store retail varies between e-shopper groups as well as the associated effects (Chang and Liao, 2011; Kelobonye et al., 2019; Páez et al., 2010b, Páez et al., 2010a). This approach recognizes multimodal accessibility to major destinations as relative rather than universal for everyone, since people usually have different transport preferences, needs, and abilities, relying on their socio-economic realities, time constraints, priorities, and cultural norms (Bantis and Haworth, 2020; Dixit and Sivakumar, 2020). Multimodal relative accessibility is specifically defined as “a set of opportunities available to an individual with defined characteristics at a selected location, relative to an individual from a reference group at the same location” (Páez et al., 2010a, p.3). This notion of multimodal relative accessibility pays strong attention on identifying whether some population segments may experience accessibility-related disadvantages according to their e-shopping habits, transport preferences, and land-use configuration. Therefore, approaching accessibility from a relative perspective facilitates the identification of both specific population groups and places that are highly dependent on in-store retail (Arranz-López et al., 2019; Brand et al., 2020; Páez et al., 2010a; Rotem-Mindali, 2010).

The academic literature has paid limited attention to research the interface between e-shopping and multimodal spatial accessibility. While a majority of existing studies focus on the impact of e-shopping on the number of shopping trips (Edrisi and Ganjipour, 2016; Lee et al., 2017; Rotem-Mindali and Weltevreden, 2013; Shi et al., 2019), little is known about how variations in trip duration affect spatial accessibility and the subsequent link with types of e-shoppers (Van Wee, 2016). Gaining insight into the e-shoppers' accessibility patterns to retail may provide guidelines for addressing potential conflicts between e-shopping and in-store retail. There are also a persistent interest in academic studies in knowing how e-shoppers' behaviour impacts on motorized modes and trips, rather than on addressing research designs that follow a multimodal approach (Arranz-López et al., 2019; Etminani-Ghasrodashti and Hamidi, 2020; Lee et al., 2017). Using a multimodal framework to explore e-shopper accessibility (car, public transport, walking) to in-store retail can have significant implications for policy-making, particularly in overcoming potential accessibility disadvantages among vulnerable groups with limited e-shopping habits (e.g., seniors).

Building on these important issues, this paper explores the following research questions: Which are the time intervals that indicate shifts in spatial multimodal accessibility to reach in-store retail between e-shopper groups, and which are the spatial and social effects of such multimodal accessibility levels? The city of Alcalá de Henares (Madrid Metropolitan Area, Spain) serves as the case study. First, a face-to-face questionnaire was disseminated during January and February 2020, capturing data on e-shopping habits, socio-economic characteristics, and travel time spent to reach in-store retail by multiple modes (car, public transport, and walking). Second, three groups of e-shoppers were identified using the k-modes clustering technique: Group #1, occasional e-shoppers with a car; Group #2, infrequent e-shoppers with a car; and Group #3, frequent e-shoppers without a car. Third, variations in distance-decay functions to reach in-store retail were analysed between the three e-shopper groups, using the Kruskal-Wallis and Mann-Whitney U tests. Finally, a gravity-based model analysed and mapped the spatial effects of the e-shoppers' accessibility patterns, identifying multimodal accessibility places (MAPs) with different ranges of vulnerability/resistance to e-shopping. The obtained findings can serve as guidance to better incorporate e-shopping into land use and transport policy-making, with the aim to achieve more inclusive, equitable, and fair cities.

The remainder of the paper is structured as follows. Section 2 reviews the literature on the impacts of e-shopping on transport and relative accessibility. Section 3 presents the study area and Section 4 details the research design. The key results are presented in Section 5, while Section 6 shares concluding remarks and highlights issues for further research.