Why It Costs So Much to Live There

Housing is most expensive where productivity is highest. This is not a coincidence and it is not, in itself, a problem: people bid for access to places where their work is worth more, and prices reflect that. The problem is what happens next, and it depends entirely on whether anybody is allowed to build.

Demand has to clear somewhere. In a city where supply responds, population growth shows up as buildings: more units, modest price movement, arrivals absorbed. In a city where supply does not respond, the same growth shows up as price, because the only way to allocate a fixed stock among more people is for them to outbid each other. The comparison between American metropolitan areas over the past fifty years makes this unusually legible. Cities that built absorbed decades of growth with real prices barely moving. Cities that did not saw comparable growth turn into some of the highest housing costs in the world. The difference is overwhelmingly regulatory rather than geographic.

Two conditions make supply elastic: land that can be built on, and rules that permit building on it. Both are necessary. Available land under a prohibitive planning regime is not usable supply, and permissive rules in a city with nothing left to develop do not help either. Most of the variation between rich-world cities comes from the rules.

The cost of getting this wrong extends far beyond the housing market, and this is the part that is easy to miss. Agglomeration (the productivity gain from density, through thicker labour markets, faster circulation of ideas and viable specialised suppliers) is large and well measured. It is why the expensive cities are expensive. When housing costs prevent workers from moving to where they would be most productive, the output they would have produced does not exist anywhere. Estimates of the cumulative loss to the United States alone from constraints in a handful of high-productivity metropolitan areas run into the trillions. The modelling assumptions are contested; the direction and rough magnitude have survived a great deal of scrutiny.

A common objection is that new construction is expensive and therefore helps only the wealthy. Studies that follow chains of moves find otherwise. Someone entering a new building vacates a unit, which is taken by someone vacating another, and within a few links the chain reaches income levels well below the price of the original construction. The effect is smaller than building cheap housing directly, and it is clearly not zero, which is what the strong version of the objection requires.

The politics of all this is coherent, and treating it as ignorance leads to strategies that reliably fail. The costs of new building (disruption, changed views, and above all restrained appreciation of the asset in which existing owners hold most of their wealth) fall on people who live there now and vote in local elections. The benefits go to people who do not live there yet and cannot vote at all. Concentrated costs, dispersed benefits, and a decision rule that weights only the first. The reforms that have worked have generally been those that moved the decision to a level of government where the beneficiaries are also constituents.