Building Hierarchies of Retail Centers Using Bayesian Multilevel Models
Sam Comber; Daniel Arribas-Bel; Alex Singleton; Guanpeng Dong; Les Dolega (2020). Annals of the American Association of Geographers, 110(4), 1150-1173. DOI: 10.1080/24694452.2019.1667219
Abstract
The perceived quality of urban environments is intrinsically tied to the availability of desirable leisure and retail opportunities. In this article, we explore methodological approaches for deriving indicators that estimate the willingness to pay for retail and leisure services offered by retail centers. Most often, because the quality of urban environments cannot be qualified by a natural unit, the willingness to pay for an urban environment is explored through the lens of the residential housing market. Traditional approaches control for individual characteristics of houses, meaning that the remaining variation in the price can be unpacked and related to the availability of local amenities or, equivalently, the willingness to pay. In this article, we use similar motivations but exchange housing prices for residential properties with property taxes paid by nondomestic properties to glean hierarchies of retail centers. We outline the applied methodological steps that include very recent, nontrivial contributions from the literature to estimate these hierarchies and provide clear instructions for reproducing the methodology. Using the case study of England and Wales, we undertake a series of econometric experiments to rigorously assess retail center willingness to pay (RWTP) as a test of the methods reviewed. We build intuition toward our preferred specification, a Bayesian multilevel model, that accounts for the possibility of a spatial autoregressive process. Overall, the applied methodology describes a blueprint for building hierarchies of retail spaces and addresses the limited availability of spatial data that measure the economic and social value of retail centers.
Extended Summary
This research develops a methodology to rank retail centres across England and Wales based on their economic attractiveness and willingness to pay for retail and leisure services. The study addresses a significant gap in quantitative evidence about retail centre performance, which has undermined effective policy formulation in urban planning. Instead of using traditional residential housing market data to measure urban environment quality, this research employs business rates (property taxes paid by commercial properties) as an alternative lens to explore retail centre hierarchies. The methodology combines hedonic pricing models with advanced spatial statistics to control for property-specific characteristics such as floor area, parking spaces, and store type. The remaining variation in business rates is then attributed to the retail centre willingness to pay (RWTP) - essentially how much consumers value the retail and leisure opportunities in that location. Using data from 355,076 individual high street stores across 2,951 retail centres, the research compares several econometric approaches including spatial fixed effects, multilevel models, and Bayesian hierarchical spatial models. The preferred methodology employs a Bayesian multilevel model that accounts for spatial autocorrelation between neighbouring retail centres. This approach provides more precise estimates than traditional methods by incorporating ‘shrinkage’ effects that pull estimates toward more reliable values based on local context. The validation exercise demonstrates that higher RWTP values correlate with areas having better health outcomes, lower vacancy rates, more retail variety, and higher employment levels - confirming the methodology’s effectiveness. The research reveals significant regional inequalities, with retail centres in the East of England and London showing the highest willingness to pay values, while areas in the North West and South West England show lower values. The methodology accounts for the complex geography of modern retail consumption, including the impact of online shopping and changing consumer behaviours toward local, convenience-oriented shopping patterns. This work provides retail practitioners with quantifiable measures of centre performance that could inform investment decisions, store portfolio rationalisation, and footfall generation strategies. For policymakers, the research offers evidence-based insights into retail centre economic health and urban development patterns. The methodology is replicable and generalisable to other national contexts, providing a blueprint for building retail space hierarchies that addresses the limited availability of spatial data measuring retail centres’ economic and social value.
Key Findings
- Bayesian multilevel models with spatial autocorrelation provide more precise retail centre willingness to pay estimates than traditional fixed effects approaches
- Significant regional inequalities exist with East England and London retail centres showing highest willingness to pay values
- Business rates data successfully captures local market conditions and consumer preferences for retail centre locations and amenities
- Validation confirms higher retail centre values correlate with better health outcomes, lower vacancy rates, and greater employment levels
- The methodology provides first national-scale quantitative ranking system for retail centre economic performance and attractiveness
Citation
@article{comber2020building,
author = {Sam Comber; Daniel Arribas-Bel; Alex Singleton; Guanpeng Dong; Les Dolega},
title = {Building Hierarchies of Retail Centers Using Bayesian Multilevel Models},
journal = {Annals of the American Association of Geographers},
year = {2020},
volume = {110(4)},
pages = {1150-1173},
doi = {10.1080/24694452.2019.1667219}
}