Observed associations between the residential built environment and behavior (travel mode choice, car ownership, walking, and adjacent health and social outcomes) conflate two processes: causal effects of the environment on residents, and residential self-selection, the sorting of households into locations that match pre-existing attitudes, preferences, and constraints. The problem remains open in a precise, well-defined form. Later literature has substantially answered the oldest sub-question — whether travel attitudes remain stable across residential moves: they generally do not. Multi-wave and retrospective studies in Belgium, the Netherlands, Finland, and China find that attitudes adapt toward the modes the new environment affords, so both self-selection and its reverse operate. Quasi-random residential assignment (affordable-housing lotteries in San Francisco, randomized replacement housing in Shanghai, resettlement housing in Beijing, public housing in Hong Kong) now supplies causal estimates of built-environment effects, but only for travel and auto outcomes and for constrained populations, and the resulting estimates diverge sharply: the implied self-selection share of the built environment–behavior association ranges from under 2% to roughly 38% depending on context and estimation method. The surviving open core is a consistent quantitative account of (i) the magnitude, pace, and persistence of attitude adaptation around moves and the bias it induces in each identification strategy in use; (ii) whether quasi-experimental estimates generalize to general populations, other institutional and geographic contexts, non-marginal built-environment changes, and outcomes beyond travel, including physical activity, health, and sociological outcomes such as education and income; and (iii) the mechanisms behind outcome-domain heterogeneity, such as attenuated effects on commute relative to non-work trips.

Background

People who live in walkable, transit-rich neighborhoods walk more and drive less than people in car-dependent suburbs, and residents of disadvantaged neighborhoods fare worse on many health and social outcomes. The standing question is how much of such differences the residential environment itself causes. Two families of explanation compete. The first is a causal effect of the built environment — the street layout, land-use mix, density, transit access, and parking supply that characterize a place — on the behavior of its residents. The second is residential self-selection: households choose where to live partly on the basis of preferences, attitudes, and constraints (housing cost, job location, school quality), so the people who end up in walkable neighborhoods are disproportionately people who already liked walking. In this literature, travel attitudes are latent, survey-measured evaluations of travel modes (liking or disliking travel by car, public transport, bicycle, or on foot). Because attitudes influence both the chosen location and the observed behavior, a cross-sectional regression of behavior on the built environment conflates causation with selection, and the direction of influence cannot be established from such data (Frank, Saelens, Powell and Chapman 2007).

Work on this impasse has taken three routes. First, preference control and dissonance: Schwanen and Mokhtarian (2005) studied neighborhood type dissonance — residents living in a neighborhood of a different type than the one they prefer — and found that dissonant residents' travel patterns fall between those of consonant residents of the two neighborhood types, evidence that both environment and preference matter; Frank et al. (2007) controlled for measured neighborhood preference in data from the SMARTRAQ survey (Strategies for Metropolitan Atlanta's Regional Transportation and Air Quality, a regional household travel survey) and found built-environment effects on walking and driving that survived the controls, but the design remained cross-sectional and left open whether preferences themselves respond to where people live. Second, structural equation modeling, i.e., path models decomposing associations into direct and mediated effects: Lin, Wang and Guan (2017) found that attitudes substantially mediate the association between the built environment and travel behavior, but could not determine whether attitudes drove residential choice (self-selection) or the environment reshaped attitudes (residential determination), because their data were cross-sectional. Third, quasi-experiments — designs in which residential location is determined by a process close to random rather than by household choice: Millard-Ball, West, Rezaei and Desai (2022) exploited San Francisco's affordable-housing lotteries, in which eligible households are randomly allotted specific housing units and hence specific neighborhood attributes and parking supplies; they verified that winners' baseline preferences were uncorrelated with the attributes they were allotted, and estimated sizable causal effects of randomly assigned parking and transit access on car ownership and mode choice, with more muted effects on commute trips than on non-work trips (an attenuation the paper itself flags) and null effects on employment and job mobility. Parallel quasi-random schemes elsewhere include government-assigned replacement housing in Shanghai (Chen and Costa 2024), public housing in Hong Kong (Zang et al. 2019), and resettlement housing in Beijing (Liu and Ding 2024).

A parallel neighborhood-effects literature in sociology and economics shows that randomized mobility designs can identify effects on sociological outcomes: in the Moving to Opportunity (MTO) experiment, low-income US families received vouchers that randomized relocation to lower-poverty neighborhoods, and Chetty, Hendren and Katz (2016) found substantial earnings and college-attendance gains for children who moved young. But MTO varied neighborhood poverty and quality rather than specific built-form attributes, and the two literatures have not been connected.

The hinge of the whole problem is attitude dynamics. Panel studies following households across relocations — in Ghent (De Vos, Ettema and Witlox 2018), the Netherlands (van de Coevering, Maat and van Wee 2018; Tao 2024, using random-intercept cross-lagged panel models, which estimate reciprocal lagged effects between person-centered deviations of preferences and behavior across three waves), Beijing (Wang and Lin 2019), and Helsinki (Ramezani et al. 2021) — converge on the finding that travel attitudes are not fixed traits: they adapt toward the modes the new environment affords, over months to years, while simultaneously influencing where households move in the first place. This means both self-selection and its reverse operate at once, and any estimator must say which share of an association each accounts for. Current estimates of the self-selection share diverge — from under 2% in Beijing resettlement housing (Liu and Ding 2024) to roughly 38% on average in a Dutch application (van Herick and Mokhtarian 2020) — and the estimate depends on the method itself. The unresolved objective is a consistent quantitative account of how much the residential environment causes behavior, how attitudes co-evolve with location, and how far quasi-experimental estimates travel across populations, contexts, and outcome domains.

Problem Statement

Central question: in the relationship between residential context (the built environment of the neighborhood and the dwelling) and behavioral outcomes — travel behavior, physical activity, and adjacent sociological outcomes such as education and income — how much is causal influence of the environment, how much is residential self-selection (the sorting of households into locations according to pre-existing attitudes, preferences, and constraints), and how do attitudes and preferences themselves co-evolve with residential location?

In the form the literature has left open, the problem comprises three jointly necessary components under this single objective:

  1. Attitude dynamics and identification bias. Quantify how travel and residential preferences and attitudes change before, during, and after residential relocation — their magnitude, direction, pace, and persistence — and determine how much this adaptation biases each identification strategy in use (cross-sectional statistical control, i.e., regression adjustment for measured preferences and selection factors; propensity-score and sample-selection approaches, i.e., matching or re-weighting households on the observed predictors of their location choice and models correcting for non-random selection into observed location types; and mover or panel designs that follow households across relocations), including cases where self-selection operates not only on residence but on correlated long-term choices such as workplace and car ownership.

  2. External validity of quasi-experimental estimates. Establish whether causal estimates of built-environment effects obtained from quasi-random residential assignment (housing lotteries, replacement or resettlement housing, public-housing allocation) extend to general populations beyond the eligible or constrained households studied, to other geographic and institutional contexts, to non-marginal changes in the built environment (such as large-scale parking removal or transit network expansion), and to outcome domains beyond travel — including physical activity and health, and sociological outcomes such as education and income.

  3. Mechanisms of outcome-domain heterogeneity. Identify the mechanisms through which causal effects differ across outcome domains and trip types — for example, why estimated effects on commute trips are attenuated relative to non-work trips — and how constraints such as workplace location, workplace parking, income, and housing institutions moderate the causal effect of the residential environment.

The problem retains its natural generality: it is not restricted to a specific behavior, geographic context, population, or method class. An adequate answer resolves the shared objective across all three components above: a consistent, empirically grounded division of the residential-context association with behavior into causal-environmental and self-selection components, together with an account of the attitude dynamics that connect them and of the identification bias those dynamics induce in each identification strategy in use; a determination of how far quasi-experimental built-environment estimates generalize across populations, contexts, non-marginal changes, and outcome domains; and an identification of the mechanisms behind outcome-domain heterogeneity — valid across the domains where the question is asked.

References

  1. Frank, L. D., Saelens, B. E., Powell, K. E., & Chapman, J. E. (2007). Stepping towards causation: Do built environments or neighborhood and travel preferences explain physical activity, driving, and obesity? Social Science & Medicine, 65(9), 1898-1914. DOI: 10.1016/j.socscimed.2007.05.053. https://doi.org/10.1016/j.socscimed.2007.05.053
  2. Lin, T., Wang, D., & Guan, X. (2017). The built environment, travel attitude, and travel behavior: Residential self-selection or residential determination? Journal of Transport Geography, 65, 111-122. DOI: 10.1016/j.jtrangeo.2017.10.004. https://doi.org/10.1016/j.jtrangeo.2017.10.004
  3. Millard-Ball, A., West, J., Rezaei, N., & Desai, G. (2022). What do residential lotteries show us about transportation choices? Urban Studies, 59(2), 434-452. DOI: 10.1177/0042098021995139. https://doi.org/10.1177/0042098021995139
  4. Schwanen, T., & Mokhtarian, P. L. (2005). What if you live in the wrong neighborhood? The impact of residential neighborhood type dissonance on distance traveled. Transportation Research Part D: Transport and Environment, 10(2), 127-151. DOI: 10.1016/j.trd.2004.11.002. https://doi.org/10.1016/j.trd.2004.11.002
  5. Scheiner, J., & Holz-Rau, C. (2013). Changes in travel mode use after residential relocation: A contribution to mobility biographies. Transportation, 40(2), 431-458. First published online 2012. DOI: 10.1007/s11116-012-9417-6. https://doi.org/10.1007/s11116-012-9417-6
  6. Giles-Corti, B., Bull, F., Knuiman, M., McCormack, G., Van Niel, K., Timperio, A., Christian, H., Foster, S., Divitini, M., Middleton, N., & Boruff, B. (2013). The influence of urban design on neighbourhood walking following residential relocation: Longitudinal results from the RESIDE study. Social Science & Medicine, 77, 20-30. DOI: 10.1016/j.socscimed.2012.10.016. https://doi.org/10.1016/j.socscimed.2012.10.016
  7. Ettema, D., & Nieuwenhuis, R. (2017). Residential self-selection and travel behaviour: What are the effects of attitudes, reasons for location choice and the built environment? Journal of Transport Geography, 59, 146-155. DOI: 10.1016/j.jtrangeo.2017.01.009. https://doi.org/10.1016/j.jtrangeo.2017.01.009
  8. Wang, D., & Lin, T. (2019). Built environment, travel behavior, and residential self-selection: A study based on panel data from Beijing, China. Transportation, 46(1), 51-74. First published online 2017. DOI: 10.1007/s11116-017-9783-1. https://doi.org/10.1007/s11116-017-9783-1
  9. van de Coevering, P., Maat, K., & van Wee, B. (2018). Residential self-selection, reverse causality and residential dissonance: A latent class transition model of interactions between the built environment, travel attitudes and travel behavior. Transportation Research Part A: Policy and Practice, 118, 466-479. DOI: 10.1016/j.tra.2018.08.035. https://doi.org/10.1016/j.tra.2018.08.035
  10. De Vos, J., Ettema, D., & Witlox, F. (2018). Changing travel behaviour and attitudes following a residential relocation. Journal of Transport Geography, 73, 131-147. DOI: 10.1016/j.jtrangeo.2018.10.013. https://doi.org/10.1016/j.jtrangeo.2018.10.013
  11. van Herick, D., & Mokhtarian, P. L. (2020). How much does the method matter? An empirical comparison of ways to quantify the influence of residential self-selection. Travel Behaviour and Society, 18, 68-82. DOI: 10.1016/j.tbs.2019.09.002. https://doi.org/10.1016/j.tbs.2019.09.002
  12. Guan, X., & Wang, D. (2020). The multiplicity of self-selection: What do travel attitudes influence first, residential location or work place? Journal of Transport Geography, 87, 102809. DOI: 10.1016/j.jtrangeo.2020.102809. https://doi.org/10.1016/j.jtrangeo.2020.102809
  13. Zang, P., Lu, Y., Ma, J., Xie, B., Wang, R., & Liu, Y. (2019). Disentangling residential self-selection from impacts of built environment characteristics on travel behaviors for older adults. Social Science & Medicine, 238, 112515. DOI: 10.1016/j.socscimed.2019.112515. https://doi.org/10.1016/j.socscimed.2019.112515
  14. Ramezani, S., Hasanzadeh, K., Rinne, T., Kajosaari, A., & Kyttä, M. (2021). Residential relocation and travel behavior change: Investigating the effects of changes in the built environment, activity space dispersion, car and bike ownership, and travel attitudes. Transportation Research Part A: Policy and Practice, 147, 28-48. DOI: 10.1016/j.tra.2021.02.016. https://doi.org/10.1016/j.tra.2021.02.016
  15. Faber, R., Merkies, R., Damen, W., Oirbans, L., Massa, D., Kroesen, M., & Molin, E. (2021). The role of travel-related reasons for location choice in residential self-selection. Travel Behaviour and Society, 25, 120-132. DOI: 10.1016/j.tbs.2021.07.003. https://doi.org/10.1016/j.tbs.2021.07.003
  16. Chen, F., & Costa, A. B. (2024). Exploring the causal effects of the built environment on travel behavior: A unique randomized experiment in Shanghai. Transportation, 51(1), 215-245 (online 2022). DOI: 10.1007/s11116-022-10325-5. https://doi.org/10.1007/s11116-022-10325-5
  17. Deng, Y., & Zhao, P. (2022). Quantifying residential self-selection effects on commuting mode choice: A natural experiment. Transportation Research Part D: Transport and Environment, 104, 103197. DOI: 10.1016/j.trd.2022.103197. https://doi.org/10.1016/j.trd.2022.103197
  18. Acharjee, A., & Sarkar, P. P. (2022). Longitudinal study of travel attitude and travel activities on residential self-selection: A case study of small-sized Indian city. Cities, 129, 103934. DOI: 10.1016/j.cities.2022.103934. https://doi.org/10.1016/j.cities.2022.103934
  19. van Wee, B., & Cao, J. (2022). Residential self-selection in the relationship between the built environment and travel behavior: A literature review and research agenda. In: Urban Transport and Land Use Planning: A Synthesis of Global Knowledge (Advances in Transport Policy and Planning), 75-94. ISBN 9780128240809. DOI: 10.1016/bs.atpp.2020.08.004. https://doi.org/10.1016/bs.atpp.2020.08.004
  20. Tao, Y. (2024). Linking residential mobility with daily mobility: A three-wave cross-lagged panel analysis of travel mode choices and preferences pre-post residential relocation in the Netherlands. Urban Studies, 61(2), 273-293. First published online 2023. DOI: 10.1177/00420980231181049. https://doi.org/10.1177/00420980231181049
  21. Schimohr, K., Heinen, E., & Scheiner, J. (2023). Travel-based residential dissonance as a motivation for relocation: An analysis of movers in Germany. Travel Behaviour and Society, 33, 100639. DOI: 10.1016/j.tbs.2023.100639. https://doi.org/10.1016/j.tbs.2023.100639
  22. Liu, T., & Ding, C. (2024). Revisiting built environment and travel behavior: A natural experiment accounting for residential self-selection. Journal of Transport Geography, 115, 103794. DOI: 10.1016/j.jtrangeo.2024.103794. https://doi.org/10.1016/j.jtrangeo.2024.103794
  23. Shi, L., Jiang, Y., Chen, F., Zhu, K., Nielsen, C. P., Wang, Y., Tian, F., Wu, J., & Chen, X. (2025). Built environment, car ownership and PM2.5: Stronger causal estimates from a quasi-experiment. Journal of Transport Geography, 127, 104301. DOI: 10.1016/j.jtrangeo.2025.104301. https://doi.org/10.1016/j.jtrangeo.2025.104301
  24. Chetty, R., Hendren, N., & Katz, L. F. (2016). The effects of exposure to better neighborhoods on children: New evidence from the Moving to Opportunity experiment. American Economic Review, 106(4), 855-902. DOI: 10.1257/aer.20150572. https://doi.org/10.1257/aer.20150572
  25. Stolze, D., Siebert, F. W., & Haustein, S. (2026). Forced or desired residential relocation? Moving voluntariness and its effects on well-being and travel behaviour. Travel Behaviour and Society, 45, 101346. DOI: 10.1016/j.tbs.2026.101346. https://doi.org/10.1016/j.tbs.2026.101346