Modelling pandemic-induced shifts in residential location preferences: Insights from a large-scale discrete choice experiment in Chinese cities

Key findings from this study

• The study found that preferences for commute duration, public transport access, and school quality strengthened persistently across all three decision frames, whereas concerns about rent sensitivity and shopping proximity elevated immediately post-lockdown before subsiding.
• The authors report that middle-income respondents and those experiencing asymptomatic or mild infections exhibited the broadest preference adjustments, while households with children or older adults registered the largest magnitude changes.
• The researchers demonstrate that urban residential preferences diverged between local-convenience logic (proximity-dependent preferences strengthening in family-centred households) and spatial-flexibility logic (reduced distance and cost sensitivity among higher-income, vehicle-equipped residents).

Overview

The pandemic disrupted residential location preferences among urban residents in China, yet longitudinal individual-level evidence documenting preference evolution across pandemic stages remained scarce. A discrete choice experiment with 7318 urban residents estimated within-person changes in residential trade-offs at three decision frames: pre-pandemic baseline (September-November 2019), post-lockdown present, and envisioned post-pandemic future. Analysis identified four temporal trajectories of preference change—rising, falling, V-shaped, and inverted-V patterns—demonstrating that pandemic impacts operated non-uniformly across the population.

Methods and approach

The study employed a hypothetical, scenario-based discrete choice experiment design with scenario randomisation across three temporal frames. Respondents evaluated residential location trade-offs by making sequential choices among alternatives that varied on attributes including rent, commute distance, public transport access, shopping proximity, and school quality. Within-person preference coefficients were estimated from stated choices across all three frames, enabling detection of preference trajectories and heterogeneous response patterns by respondent characteristics including income, infection severity, household composition, and satisfaction.

Results

Preferences for monthly rent sensitivity and shopping proximity surged immediately after lockdowns but subsequently declined, indicating transient economic and provisioning concerns. Conversely, preferences for commute duration, public transport accessibility, and school quality strengthened continuously across all three frames, signalling durable reordering of priorities. Preference shifts exhibited substantial heterogeneity: middle-income respondents and those with asymptomatic or mild infections adjusted preferences across the broadest attribute set. Households with older adults or school-age children and highly satisfied residents registered the largest magnitudes of change. Family-centred households became progressively more sensitive to commuting burdens and facility distances, while higher-income residents, licenced drivers, service workers, and satisfied residents progressively reduced sensitivity to distance and cost.

Implications

The research documents divergence between two competing logics in residential location preference. The emerging local-convenience logic reflects heightened demand for proximity to daily services and short commutes. Simultaneously, spatial-flexibility logic gains traction among certain segments, particularly higher-income and vehicle-equipped households less constrained by geographic proximity. This divergence carries implications for urban planning and real estate markets, as different household segments increasingly optimise locations according to distinct priorities and constraints. Policy responses must balance these competing logics to ensure equitable access across income and family-type segments.

For urban governance, findings underscore necessity of citywide investment in commuting reliability, educational access, and affordability rather than neighbourhood-specific improvements alone. The persistent strengthening of public transport and school-quality preferences across all temporal frames suggests that pandemic-induced preference shifts reflect durable household prioritisation changes rather than temporary disruption effects. Service provision and neighbourhood investment tailored to heterogeneous household needs—family-oriented services in residential areas, mobility infrastructure for commute-sensitive households, and cost-control measures for income-constrained groups—align with emerging preference patterns. Failure to address these differentiated demands risks spatially segregating population segments by household type and income.

Scope and limitations

This summary is based on the study abstract and available metadata. It does not include a full analysis of the complete paper, supplementary materials, or underlying datasets unless explicitly stated. Findings should be interpreted in the context of the original publication.