# Migration and the adjustment of housing and infrastructure

## Population can change faster than the capital stock

Migration adds potential workers and consumers, while also changing the demand for homes, transport, utilities, schools, and health services. These effects operate on different schedules. A person can enter a local labor market and need accommodation before a new dwelling or rail connection can be completed. Over time, additional workers can help expand construction and public services, but that response depends on skills, location, financing, and project readiness.

Canada, Australia, and the United Kingdom offer useful evidence on this timing problem. Their published population estimates, household projections, and housing-affordability assessments are different statistical products. They cannot be placed in one table as equivalent measures of a migration-induced housing shortage. A useful comparison first identifies what each publication measures, then connects population change with the adjustment of local capacity.

The relevant unit is often a city or commuting area. National population growth can coexist with spare housing in one region and severe pressure in another. The national labor supply may increase while a particular construction project still lacks a qualified electrician. Geographic and occupational matching determine whether a potential supply response becomes available where demand rises.

## Australia's migration figures count population movements

The Australian Bureau of Statistics' December 2025 release reports net overseas migration of approximately 306,000 people for fiscal 2024–25. It separately records approximately 568,000 migrant arrivals and 263,000 departures, with rounding affecting arithmetic. The release distinguishes these migration estimates from visa grants and warns that individuals can depart under a different visa category from the one held on arrival. [ABS, Overseas Migration, 2024–25](https://www.abs.gov.au/statistics/people/population/overseas-migration/2024-25).

The distinction is operationally important. A visa issued to someone already resident does not necessarily add another resident or require another dwelling. A border crossing is not necessarily a migration event. A net annual figure also conceals gross turnover, household composition, and the locations in which arrivals and departures occur.

For housing analysis, the migration contribution must be connected with the wider population account, including natural increase and internal migration. The age and household composition of the change matter. A family joining an existing household, a student entering communal accommodation, and a worker forming a new household can have different immediate housing implications even when each adds one person to a population count.

These distinctions do not eliminate additional demand. They identify how to estimate it without equating each migrant with a new household. They also prevent the opposite mistake of assuming that shared accommodation has no consequence for rents, crowding, or the housing choices available to existing residents.

## Canada's supply-gap estimate has an affordability objective

CMHC's June 19, 2025 assessment estimated the housing construction required over the following decade to restore specified 2019 affordability conditions. It distinguished ownership and rental markets and reported substantial differences among metropolitan areas. The assessment focused primarily on average-income households in market housing, rather than measuring every form of low-income need or homelessness. It is a dated model-based affordability exercise, not a count of dwellings attributable solely to migration. [CMHC, 2025 housing supply-gap release](https://www.cmhc-schl.gc.ca/media-newsroom/news-releases/2025/cmhc-releases-latest-housing-supply-gaps-report).

An affordability target can require more housing than a demographic household projection because the model allows housing costs and household choices to adjust. Conversely, an apparently small demographic gap can coexist with substantial unaffordability. A household may remain in crowded accommodation or delay independent formation when prices are high. Observed households therefore do not reveal all latent demand.

The Canadian publication also emphasizes the long construction process. That is the bridge between population and capacity: a demand change enters a market with an existing stock and a pipeline of projects at different stages. The pace of completed supply depends on what was approved, financed, and started earlier, as well as on new decisions.

## England's projection is conditional on demographic assumptions

The ONS's October 2025 household projections for England use a migration-category variant and explain why it better reflected then-available short-term migration information than the principal population projection. The release explicitly states that household projections do not forecast how many homes should be built. They show household formation conditional on demographic assumptions and do not incorporate all local economic and housing-policy influences. England's coverage should not be presented as a UK-wide household total. [ONS, Household projections for England, 2022-based](https://www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/populationprojections/bulletins/householdprojectionsforengland/2022based).

This provides a third object alongside Australia's observed migration estimate and Canada's affordability model. The publications can inform a comparative framework, but their headline values answer different questions. An analyst should preserve the difference between observed population change, conditional household formation, and additional supply needed to meet a chosen affordability standard.

Projection variants are useful for planning because population and household behavior are uncertain. They allow agencies to examine what would change under different assumptions. Treating one variant as a certain outcome would obscure the reason for publishing alternatives and could lead to overconfident infrastructure commitments or inadequate contingency planning.

## Exhibit: a delayed response can leave an accumulated gap

Consider a hypothetical city that receives an additional population of ten thousand at the start of the first year. Assume that the additional residents form four thousand separate households, with no immediate sharing response and no departures. The city initially has five hundred suitable vacant dwellings. A construction response delivers one thousand additional dwellings in year one, fifteen hundred in year two, and one thousand in year three. These completions are additional to a baseline that already accommodates other population changes.

| End of year | Cumulative additional households | Initial suitable vacancies plus additional completions | Remaining accommodation gap |
|---|---:|---:|---:|
| First year | 4,000 | 1,500 | 2,500 |
| Second year | 4,000 | 3,000 | 1,000 |
| Third year | 4,000 | 4,000 | 0 |

The gap closes by the end of year three, but the transitional shortage is material. Using year-end values, the sum of the first three annual gaps is thirty-five hundred household-years. That is a simple discrete measure of unmet separate accommodation at those observation dates, rather than a direct estimate of rent increases, homelessness, or welfare loss. Within-year completion timing would change a continuous measure.

Now assume that four thousand households would form eventually, but sharing reduces observed household formation to three thousand during the first year. The measured first-year gap against available dwellings falls to fifteen hundred. The physical dwelling stock is unchanged. A smaller observed household count does not therefore prove that the original accommodation pressure disappeared; part of it may have been absorbed through altered living arrangements.

The example also distinguishes a lower flow from a lower level. If the additional population remains, a subsequent reduction in new arrivals does not remove the demand already present. It may slow further increases while the construction pipeline catches up. Comparing the latest migration flow with the total dwelling shortage can confuse the rate of change with the accumulated stock adjustment.

## Construction workers also need housing

Migration can expand the labor available to build homes, but the relationship is conditional. Workers need relevant qualifications, recognition where required, access to sites, and complementary equipment and materials. A large aggregate labor inflow does not automatically remove a shortage in a specialized trade or a location with little accommodation.

Suppose, in the hypothetical city, five hundred additional residents work in construction and each supplies sixteen hundred hours annually. They provide eight hundred thousand hours. At an assumed direct labor requirement of two thousand hours per dwelling, that is enough for four hundred dwelling equivalents before accounting for supervisory, material, financing, and infrastructure constraints. The coefficient is illustrative and does not estimate construction productivity in any of the three countries.

Those workers are already included in the ten-thousand-person population addition. Their housing requirements should not be added a second time if the household assumption already includes them. Their contribution to construction should likewise be counted only to the extent that it raises total output rather than replaces other employment or is directed to nonresidential work.

The supply contribution can arrive later than the demand contribution. A newly arrived worker may need accommodation immediately but require training, licensing, or a job match before contributing to completed housing. Conversely, an experienced worker recruited directly for an active project may contribute quickly. A realistic assessment uses cohort and occupation information rather than a uniform delay for every migrant.

## Infrastructure has separate capacity and operating constraints

New dwellings do not by themselves ensure adequate utilities or transport. A development may require a water connection, electricity reinforcement, school places, or road capacity. Some infrastructure can be expanded incrementally; other projects have large minimum scales and long lead times. The relevant capital stock is therefore broader than residential buildings.

Service provision also depends on staffing and operating budgets. A new clinic without personnel does not deliver the same capacity as a functioning service. A school expansion can require teachers, transport, and recurring expenditure alongside classrooms. Population planning must connect construction schedules with the supply of those complementary resources.

Fiscal responsibilities can be distributed unevenly. One level of government may receive additional tax revenue while another finances local services. The resulting timing and coordination problem is not resolved by a national estimate of migration's long-run fiscal contribution. Funding arrangements must connect the location of population change with the institutions responsible for accommodation and infrastructure.

Existing spare capacity matters as well. A city with available homes and underused services can absorb growth differently from one already operating close to capacity. However, a vacancy is useful only if it is habitable, appropriately located, and available to the relevant households. Counting every empty property as immediately usable supply overstates flexibility.

Location can prevent a national construction response from resolving a local gap. Suppose one region adds one thousand suitable dwellings while another receives households needing those dwellings. A national balance can show sufficient additional supply even if the properties are beyond a feasible commuting distance. Prices and migration within the country may eventually adjust, but that adjustment can be costly and slow. The relevant housing stock is therefore connected to jobs, transport access, and service capacity. A model that assumes residents can move freely among all available dwellings should identify that assumption. Otherwise, apparent national sufficiency can conceal the local constraints that influence recruitment, household living arrangements, and infrastructure demand.

## Affordability and distribution require more than a national average

Housing-market adjustment can redistribute income between owners, renters, developers, and workers. Rising rents may affect recent arrivals and existing residents differently depending on tenure and income. Higher construction activity can expand employment while raising demand for scarce materials and land. A national increase in output does not identify who bears the transitional costs.

Interest rates, credit conditions, land-use rules, construction costs, and domestic household formation also affect housing outcomes. Migration should be analyzed alongside these influences rather than assigned the entire observed movement in prices. Establishing a causal contribution requires a design that separates population changes from concurrent local economic conditions.

The most informative comparison would follow the arrival and departure of people, the formation and suppression of households, suitable vacancies, project starts, completions, and service capacity within matched locations. It would distinguish an observed outcome from a scenario and state the assumptions linking people to households and households to dwellings.

The Canadian, Australian, and English publications offer complementary starting points for that work. They show why the timing of adjustment and the definition of the target matter as much as the national population total. A coherent policy response connects migration and labor-market planning with housing and infrastructure delivery over the period in which residents actually need them.

See [productive labor and demographic change](/research/demographic-change-and-the-supply-of-productive-labor/) for labor-supply accounting, and [fiscal rules and investment delivery](/research/fiscal-rules-and-investment-delivery/) for the stages between public authorization and operating services.

---

American Economic Senate