The initial NLSCY cohort comprised 22,831 children and was tracked through biennial waves from 1994 to 2009 to measure what determines their educational trajectories. A first move at age 14 or older is associated with a predicted probability of on-time completion roughly 10 percentage points lower than that of non-mobile students, after controls. The study observes an association between mobility and educational completion and proposes plausible mechanisms, without establishing what school policies can or cannot fix. Funding schools without funding roofs limits the effectiveness of both.
The Essentials
- A first move at age 14 or older is associated with a predicted probability of on-time completion of 80%, compared to 90% among non-mobile students (Calhoun & Haan 2026, Population Space Place).
- The NLSCY cohort used in the study covers 1994 to 2009 and allows comparison of forms of residential and school mobility. Both types of mobility are associated with lower on-time completion, and their combination is most unfavorable.
- Three mechanisms explain this effect: loss of social ties, disruption of learning habits, cognitive burden of adapting to a new environment.
- Traditional educational policies—curriculum reform, pedagogical support, standardized assessment—operate within the framework of a stable student; housing-school coordination could complement these interventions.
- The 2030-2040 horizon will pose the question even more acutely in major cities where rental pressure forces families into unplanned moves.
At Fourteen, Changing Neighborhoods Changes Trajectory
The research by Calhoun and Haan, published in 2026 in the journal Population, Space and Place, seeks to establish which moves experienced during schooling weigh on educational outcomes. The analyses distinguish residential and school mobility within an initial NLSCY cohort of 22,831 children, with follow-up through biennial cycles until 2009.
The most striking result concerns adolescence. A first move at age 14 or older is associated with a predicted probability of on-time completion of 80%, compared to 90% among non-mobile students. This effect withstands neutralization of standard socio-economic variables: household income, parental education level, family structure. In other words, two adolescents with identical incomes, in structurally similar families, do not have the same probability of completing secondary school depending on whether they moved during their high school years.
This 15-point figure occurs in a specific context. In advanced economies, a secondary diploma conditions access to higher education and, ultimately, professional stability. Research on wealth transmission shows that material inheritance reproduces inequalities generation after generation; a dropout in secondary closes doors long before these mechanisms operate. Residential instability can thus lock in a trajectory as early as adolescence.
The Invisible Breaks of Changing Neighborhoods
To understand why moving fragilizes education, we must look at what the student loses when changing address, not just school.
The first mechanism is the rupture of social ties. At fourteen, relationships with peers often constitute the emotional driver of schooling. An adolescent who moves loses friends, study groups, and social reference points. They must rebuild a network in a new context, at the precise moment when high school demands the most cognitive energy. This social recomposition mobilizes time and attention at the expense of learning.
The second mechanism is the disruption of learning habits. Each student builds routines: routes to school, neighborhood libraries, informal study spaces, supportive teachers. A move wipes them out all at once. The pedagogical continuity that school policies seek to guarantee within educational establishments fragments on the outside, without education systems having tools to measure it.
The third mechanism is the cognitive burden of adaptation. Settling in a new neighborhood mobilizes mental resources: new route, new social norms, new daily geography. For an adult, this takes a few weeks. For an adolescent preparing for exams, it happens at the worst possible time.
These three mechanisms have something in common: they operate outside school walls. A better student-teacher ratio, a revised curriculum, a strengthened tutoring system—none of these interventions restore lost ties or lighten the burden of adaptation.
Residential Mobility Against School Mobility: A Distinction That Changes Everything
One of the most decisive contributions of Calhoun and Haan’s study is separating what public policies have long conflated. Changing homes and changing schools are two distinct events, with distinct effects, and confusing the two leads to wrong diagnoses.
School mobility, changing schools during education, had already been documented in the literature as a risk factor. But residential mobility acts differently, and not necessarily through a school change. An unfavorable association persists for neighborhood moves without a school change.
The Canadian NLSCY cohort, which tracks the same individuals from 1994 to 2009, provides the data continuity necessary to distinguish episodes of residential and school mobility, to date when they occur, and to measure consequences at different ages. This type of longitudinal data is rare, and that is precisely what makes the conclusions robust.
The study also compares the effect of moves at different ages. Early moves, before adolescence, have measurable but less marked effects on secondary completion. Adolescence constitutes a specific window of vulnerability, something researchers in cognitive development had suspected, and this cohort confirms through data.
Educational Policies Reach Their Structural Limits
Since the 1990s, school reforms in Canada as in most OECD countries have invested in pedagogical levers: class size, teacher training, support for struggling students, standardized assessment. These efforts have produced documented results in certain contexts. But they share the same implicit assumption: the student is stable. They live where they live, and school policy can reach them where they are.
Calhoun and Haan’s data break this assumption. The study observes an association between late mobility and lower on-time completion after controls, without attributing 15 percentage points to it or evaluating educational policies.
Canada is not alone in this. American research on residential mobility among low-income families has documented similar effects since the 2000s. Studies on beneficiaries of Section 8, U.S. federal housing assistance, show that stabilizing a family in a neighborhood improves children’s school outcomes, sometimes more than direct school-based interventions. Housing as educational infrastructure: the idea is no longer theoretical.
AI in the classroom improves learning outcomes, and digital tools open real possibilities for pedagogical continuity at a distance. But a technology for personalized tracking does not restore the social ties of an uprooted adolescent. These two dimensions are not interchangeable.
The Effect of Rental Markets on Educational Trajectories
We must name what produces residential instability. Families do not move by choice in the great majority of cases, especially when it comes to their children’s adolescence, a time when parents intuitively know that stability matters.
Households with modest incomes experience residential moves linked to rental market and employment constraints. Rental markets in Toronto, Vancouver, and Montreal have faced affordability pressures. These moves touch first renters, single-parent families, new immigrants, and part-time or precarious contract workers.
The link between rental precarity and educational fragility is thus, in part, the mirror of the link between labor market and housing. Professional and rental precarity can contribute to residential instability and, by association, to less favorable school outcomes. The causal chain is long, but each link is documented.
The data indicate an association between mobility and on-time completion, without evaluating the effectiveness of school policies. An integrated policy could combine educational measures and residential stabilization measures.
Coordinating Housing and School: What 2030 Will Impose
Residential instability is associated with less favorable educational outcomes, without proof that school alone cannot mitigate them. The upstream intervention instruments thus fall to housing policy as much as to school policy.
Several countries have begun experimenting with integrated responses. In Finland, the long-term social housing model, which covers roughly 15% of the rental market, stabilizes families in neighborhoods over decades, which mechanically reduces the frequency of forced moves. The effect on school trajectories has not been measured through cohorts as long as the Canadian one, but residential stability has been integrated as a social policy objective since the 1970s.
In the United States, the Moving to Opportunity program, which relocated poor families to wealthier neighborhoods, showed that children moved before age ten benefited durably from better educational and professional trajectories, confirmation from another angle that where you live matters as much as the school itself.
Coordination between housing and education policies is envisioned by some public agencies as a priority for the 2030-2040 decade. Rental pressure in major cities should maintain, and even amplify, the frequency of forced moves among low-income families. Without targeted intervention, the effect documented by Calhoun and Haan risks worsening.
Two concrete directions are emerging. The first is the extension of rental protections for families with adolescents in school—longer leases, limits on rent increases during high school years, preferential access to social housing for families facing lease termination with children nearing the end of secondary school. These instruments exist in embryonic form in certain Canadian provinces; none has yet been evaluated at scale.
The second direction is the coordination of data systems between housing agencies and education ministries. Today, a ministry of education does not know whether a struggling student changed address three times in a year. Cross-referencing this data would allow identification of students at risk of dropping out for residential reasons, and trigger targeted support—enhanced pedagogical assistance, funding help for extracurricular activities to rebuild social ties, contact with social services.
The signal to watch in coming years is twofold: the frequency of forced moves in school cohorts in major cities, and the emergence of pilot programs explicitly coordinating housing and schooling. If such programs develop in Canada, Australia, or Nordic countries, their evaluation will allow testing whether an intervention on the roof can, in fact, shift the diploma curve.
A study published in 2026 associates certain forms of residential mobility, measured in NLSCY data from 1994 to 2009, with on-time secondary completion, which suggests the value of coordination between housing and education policies. The challenge is not to choose between investing in schools and investing in housing; both remain necessary. It is to stop treating these two policies as if they do not share the same beneficiaries.
Sources
- Calhoun & Haan (2026), “Which Move Matters More: Comparative Impact of Residential and School Mobility on High School Completion in Canada”, Population, Space and Place, Wiley Online Library
- National Longitudinal Survey of Children and Youth (NLSCY), Statistics Canada, 1994-2020, cohort data cited in Calhoun & Haan 2026
- Policy Horizons Canada, prospective reports on housing-education issues for the 2030-2040 horizon (Government of Canada)
- Moving to Opportunity Program, U.S. Department of Housing and Urban Development, longitudinal studies on the effects of relocation to wealthier neighborhoods on children’s educational and professional trajectories
