Labor Shortages Are Housing Shortages

An employer who cannot fill a position at the offered wage has a labor shortage by definition. The interesting question is why the wage that used to work no longer does, and in a growing number of metropolitan areas the answer is rent.

The arithmetic is not subtle. When housing costs consume half of a median wage, a job that pays the regional average stops being viable for anyone who is not already housed. Workers do not negotiate, they decline, and the position stays open.

This shows up in the sectors that cannot relocate. Hospitals, schools, transit systems, restaurants, and municipal services all need staff physically present in expensive places, and none can offshore the work or raise wages indefinitely against a fixed budget.

The policy response usually addresses the labor side, through training programs and recruitment pipelines. Those help at the margin and do not touch the constraint. Housing supply is the variable, it is controlled at the local level, and it is decided by the residents least affected by the shortage.

The framing decides the response, which is why it is worth being pedantic about it. An employer describes a labour shortage because that is what they experience, and the description points every remedy at the wrong side of the equation.

A wage is only meaningful against a cost of living, and in the regions where these shortages are worst the cost is dominated by one line item that has risen faster than wages for a long time.

The housing data and affordability measures both exist to make that relationship visible, and they are published on a schedule that predates the current argument by decades.

What they show is a distribution rather than an average, which is the part that gets lost. A regional median can look tolerable while the units available at the relevant price point have effectively disappeared.

The sectors that cannot escape are the ones that reveal it. A hospital, a school, a transit system and a restaurant all require staff physically present in the expensive place, and none can relocate the work to where the housing is.

That immobility is what converts a housing problem into a service problem. The ward that cannot be staffed and the route that cannot be run are the visible end of a construction shortfall from ten years earlier.

The measurement layer has its own difficulty worth acknowledging. How the figures are built determines what counts as compensation and what counts as cost, and the categories were designed when housing consumed a much smaller share.

The supply side is largely a matter of local law. The statutes that govern permitting set what may be built, where, after what process, and the process is where most of the constraint actually lives.

Each procedural step exists for a reason and most of them are defensible individually, which is the recurring pattern in everything this publication covers. The aggregate is a permitting timeline nobody designed.

The training response is the one that gets funded, and it is not useless. It moves people into occupations and it does nothing whatever about whether those people can afford to live within commuting distance of the work.

That is why the programmes appear to underperform. They are being evaluated against a shortage they were never addressed to, and the shortfall they were addressed to may genuinely have narrowed.

The uncomfortable conclusion is that a labour market intervention cannot fix this, and the intervention that could is slow, local, and controlled by the people with the least incentive to permit anything.

None of this means every shortage is housing in disguise, and the argument should not be stretched that far. Some are genuine skills gaps, some are wage suppression, and some are employers unwilling to pay what a market requires.

The claim is narrower and testable. Where a shortage is concentrated in immobile occupations, in the regions with the highest housing costs, the labour explanation is doing less work than it appears to.