1. QUESTIONS
Transportation and land use planning have traditionally focused on maximizing mobility, yet this aggregate framing does not distribute benefits equally; it favours those already well-served by the system while leaving the travel needs of vulnerable groups unmet (Pereira et al. 2017). A growing body of literature has shifted away from this framing and focused on foregone or suppressed travel: trips people wanted or planned to make but did not, typically due to inequitable infrastructure, systemic social barriers and financial constraints (Palm et al. 2024).
When transport barriers are the cause, foregone travel becomes a mechanism of transport-related social exclusion, a process by which inadequate mobility prevents people from participating in economic, social, and civic life (Kenyon et al. 2002; Lucas 2012). Transport-related social exclusion emerges from the interaction of individual factors such as income, race, and immigration status with broader structural conditions, including inadequate transport provision and failures in local service delivery (Lucas 2012). Despite growing interest, evidence on foregone travel remains limited in several ways: studies rarely separate general foregone travel from transport-related foregone travel, race and immigration status have received limited attention despite their importance, and there is a lack of research examining this problem in Canadian cities (Palm et al. 2024). This paper addresses all these gaps. We ask: to what extent do Greater Montreal residents make the trips they want, and who is most affected both in terms of foregone travel overall and transport-related barriers specifically? We hypothesize that both foregone travel and transport-related foregone travel are disproportionately concentrated among disadvantaged groups.
2. METHODS
Data were obtained from the Montreal Mobility Survey Wave 6, an online transportation survey conducted in February-March of 2026 in Greater Montreal. Participants were recruited through both in-person and online methods; the survey was completed online in English or French. Response rates could not be calculated, as the survey was distributed through multiple recruitment channels, making it impossible to determine the total number of individuals exposed to the survey invitation; of the 9,141 responses initiated, 5,236 (57.3%) were completed and retained after data cleaning. The analytical sample closely mirrors the 2021 Census of Population for Greater Montreal across key characteristics such as gender (women: 51.2% sample vs. 50.3% census), age (65+: 18.4% vs. 19.2%), and income (under $60k: 18.4% vs. 19.1%) (Statistics Canada 2023). Survey weights were calculated to match the sample to the Statistics Canada (2023) census-tract-level distributions of age, income, and gender, as well as the 2024 mode shares from the Montreal Origin-Destination Survey. A full description of the recruitment, data cleaning, and validation processes is provided by Wu et al. (2026).
The survey included a section on travel over the past seven days, with the following question asked for grocery stores, healthcare facilities (including pharmacies) and leisure destinations:
"In the last 7 days, compared to what I wanted, the number of trips I made to [destination] was… fewer than I wanted / about right / more than I wanted."
Respondents who reported making fewer trips than they wanted were additionally asked whether transportation was a challenge and, if so, what specific barriers they faced.
Our analysis proceeds in two stages. First, we characterize the prevalence and distribution of foregone travel and transport-related foregone travel through weighted descriptive analysis by income, age, gender, ethnicity, and immigration status; unweighted estimates are presented alongside weighted estimates for transparency. Second, we estimate two weighted quasibinomial logistic regression models to identify associations between sociodemographic characteristics and both outcomes. Model 1 estimates the odds of any foregone travel across the full sample (1 = reported making fewer trips than wanted for at least one purpose, 0 = otherwise), and Model 2 estimates the odds of transport barrier attribution among those reporting any foregone travel (1 = cited a transport challenge as a reason for missed trips, 0 = foregone travel reported but no transport barrier cited). Rather than conducting separate bivariate tests for each sociodemographic variable, regression models are used as the primary evidence, allowing simultaneous control of all the variables; coefficient-level p-values are interpreted cautiously, with emphasis placed on effect sizes and confidence intervals. Survey weights were applied throughout using the svyglm function from the survey package in R (Lumley 2004).
3. FINDINGS
Overall, 21% of respondents reported making fewer trips than they wanted for at least one purpose, with weighted and unweighted estimates showing similar values. Leisure had the highest share of foregone travel at 16.6%, approximately three times higher than grocery (5.6%) and four times higher than healthcare (4.2%) (Table 1). Foregone travel differed across demographic groups. Lower-income residents (under $60k) reported higher rates (26%) than the highest earners (over $180k, 17%). Age differences were modest, ranging from 20% among the youngest group to 22% among the oldest. Women reported slightly higher foregone travel than men (22.8% vs 19.4%). Immigrants and non-white residents also reported slightly higher rates than Canadian-born and White residents, respectively.
Among those who made fewer trips than they wanted, 35% cited transportation challenges as a reason for any purpose, around 7% of the entire sample. Transport was cited as a challenge by 42% of those with foregone grocery trips, 35% of those with foregone leisure trips, and 20% of those with foregone healthcare trips. Across income groups, lower-income residents cited transport challenges in 38% of foregone travel cases compared to 35% among the highest earners. Younger adults (18–29 and 30–49) reported higher rates of transport-related foregone travel than the 50–64 and 65+ age groups. Women showed higher transport-related foregone travel than men (37% vs 33%). Immigrants and non-white residents (41% and 46%, respectively) also reported higher rates than Canadian-born and White residents (33% and 30%, respectively).
Weighted logistic regression results are presented in Table 2. In Model 1 (any foregone travel), higher-income groups showed significantly lower odds of foregone travel compared to those earning under $60k, with odds ratios ranging from 0.66 ($60k–$120k) to 0.54 (over $180k). Women were more likely to report foregone travel than men (OR = 1.25, p < 0.05). The 50–64 age group and non-white residents showed marginal associations (p < 0.10), while migration status was not significant. In Model 2 (transport-related foregone travel), the income and gender effects observed in Model 1 were no longer significant. Ethnicity was significant in Model 2, as non-white residents had 79% higher odds of attributing missed trips to transport barriers compared to White residents (OR=1.79, p < 0.05). The 50–64 age group was significantly less likely to cite transport as a barrier (OR = 0.50, p < 0.05), while no other age groups differed significantly from the 18–29 reference category.
Approximately 20% of Montreal residents reported making fewer trips than they wanted, with lower-income groups and women disproportionately affected. Among transport-related foregone travel, however, income and gender were no longer significant, and ethnicity was the only significant factor, with non-white residents disproportionately more likely to attribute their missed trips to transport barriers. Reducing the barriers these groups face would enable currently suppressed trips, generating new activity participation among underserved groups. Future research could assess the extent to which spatial accessibility, online substitution, and residential self-selection independently shape foregone travel and examine purpose-specific barriers across grocery, healthcare, and leisure destinations. Multi-season panel data would allow assessment of seasonal variation, while longitudinal analysis could test whether transport investments reduce foregone travel among disadvantaged groups.
This study has several limitations. First, as an online survey, this study may underrepresent residents with limited internet access, low digital literacy, or limited proficiency in English or French, who are more likely to experience foregone travel. The prevalence of foregone travel among the most marginalized groups is likely higher than what our sample captures. Second, data were collected in winter 2026, which may inflate foregone travel rates relative to other seasons. Weather challenges related to transport (e.g., snowy sidewalks or bike lanes) were cited as a contributing factor by only 2.1% of the full sample and by 32% of those who reported transport-related foregone travel. Total trip counts in 2026 were slightly lower than in fall 2024, consistent with the winter timing of data collection, though broadly comparable to earlier waves (Wu et al. 2026), suggesting that seasonal effects exist but appear modest. As foregone travel was not measured in prior survey waves, a direct cross-seasonal comparison is not possible; this remains an important avenue for future research. That said, winter conditions may disproportionately affect some groups, such as recently arrived immigrants who have not yet adapted to cold-weather travel environments, potentially amplifying the immigration effect observed in our results.
ACKNOWLEDGMENTS
We would like to thank Prof. Ahmed El-Geneidy for giving us access to the Montreal Mobility Survey. This paper was supported by the Canadian Institute of Health Research (CIHR) (CIHR PJT - 195797) and the Social Science and Humanities Research Council of Canada (SSHRC) through the Vanier Canada Graduate Scholarships (CGV - 198918).
The analysis was developed with the assistance of Claude AI (Anthropic) to support code development and grammar refinement. The research questions, analytical decisions, interpretation of findings, and final manuscript remain the sole responsibility of the authors.