The housing shortfall is a supply constraint, not a financing constraint

There is an almost universal agreement that housing has become unaffordable in most rich countries, and an almost universal disagreement about which side of the market caused it. The debate is usually conducted in terms of interest rates, because interest rates are the variable that moves fastest and gets reported most. We think that is the wrong place to look, and the reason is a matter of arithmetic rather than ideology.

The price series says something, but not what people think

The OECD price-to-income index divides a nominal house price index by nominal disposable income per head. It reached 114.8 for the OECD average in 2024 against a 2015 base of 100, with Portugal, Canada and the Netherlands all above 130. Prices have grown meaningfully faster than incomes for a decade.

That is a real finding. But a price ratio cannot tell you whether the cause sits on the demand side or the supply side, because both produce the same signature. Cheaper credit bidding for a fixed stock raises the ratio. A constrained stock meeting stable demand raises the ratio. You cannot distinguish them without looking at quantities.

The quantity test

Here is the test we would apply. Between roughly 2010 and 2022, policy rates across most advanced economies were at or near their lowest levels in recorded history. Mortgage credit was widely available. Construction finance was cheap. If the affordability problem were principally a cost of capital problem, that decade is when it should have eased.

It did not. It worsened, in every major market, on every measure. A demand-side explanation has to account for why the most favourable financing environment in modern history coincided with the sharpest deterioration in affordability in modern history. We have not seen an account of that which holds together.

What the supply story has to explain instead

The supply story has its own burden of proof, and it is a more tractable one. It has to show that the delivery of new units was constrained by something that did not respond to cheap money. There are at least four candidates, and they are all measurable.

• Construction cost inflation. Materials and site costs rose faster than general prices. Lower financing cost does not reduce input cost.
• Skilled trades capacity. The rate at which approved units can physically be built is set by labour that takes years to train. A demand impulse into a fixed build rate produces price, not volume.
• Land use policy. Where planning restricts buildable land in the locations people want to live, credit expansion bids up a fixed stock.
• Public delivery withdrawal. OECD public capital investment in housing fell from 0.17 percent of GDP in 2001 to 0.06 percent in 2018. The channel that historically served the affordable end shrank by roughly two thirds, and the private channel that replaced it is least profitable precisely there.

Only the first of those is even indirectly affected by monetary policy, and only through developer financing costs, which are a modest share of delivered cost.

Why the distinction is worth money

If the problem is finance, it self-corrects when rates fall, and the correct planning assumption is that affordability improves through the easing cycle. If the problem is supply, rate cuts make it worse in the short run, because they raise willingness to pay against an unchanged stock. Those two forecasts diverge sharply within about eighteen months, and they imply opposite positioning for anyone with exposure to housing volumes, construction inputs or the rental market.

Our working assumption is the second. We would revise it if a large market delivered a sustained improvement in the price-to-income ratio driven by volume growth rather than by a price fall. That has not happened anywhere we can find.

Sources

• OECD, housing prices indicator, definitions and index construction
• OECD Affordable Housing Database, indicator HM1.2, house prices
• House price to income ratios by country, from OECD data