You Should Make Good Arguments for YIMBYism, Not Bad Ones

Casual analysis is no substitute for causal analysis

Many people cite falling prices in places which have built lots of housing as evidence that YIMBYism works, like Austin, Texas. However, this is not an adequate argument. Critics of YIMBYism can point to instances where housing was built, and prices still rose. The price of housing alone is not a sufficient statistic for the welfare improvements from expanding capacity, and arguing that it is needlessly weakens the case for expanding housing.

Here’s why. The quantity and price of housing is simultaneously determined by the people’s demand for housing, and developers’ ability to construct. We represent this with a graph of supply and demand. The slope of the demand curve is negative because people are willing to buy more as the price falls. The slope of the supply curve is positive because we are assuming that the cost of producing one more house is always increasing, which is a simplifying assumption. (If the marginal cost of producing one more unit were constant, then it would be flat.)

The case for YIMBYism is that by removing regulatory burdens, we reduce the costs faced by developers, causing their supply curve to shift along the demand curve. Quantity increases, and the price falls.

The trouble is that increasing quantity is also consistent with increasing price. Suppose that instead of the supply curve shifting, there is a surge in demand for the area. Consumers are now willing to spend more on housing, causing it to shift to the right along the supply curve. Both quantity supplied and price increase.

To find the effect of liberalizing housing laws, you need plausibly exogenous changes in the cost of producing housing, or shifts in the supply curve holding the demand curve fixed. Even here, though, we must be careful. Suppose that there is a shift out of the demand curve, increasing the price of housing. As a response to this, the local authorities liberalize housing, also shifting the supply curve. While the impact on quantity built is definite – it will increase – the impact on price is indeterminate. It could go up or down.

Someone considering the effects of deregulation with a regression might put deregulation on the X axis, with the degree of it appropriately weighted (there has been excellent recent work by Bartik, Gupta, and Milo (2025) using LLMs to categorize different zoning regulations; and also work by Jaehee Song (2025) on estimating minimum lot requirements), and put the price on the Y axis. If deregulation is affected by shifts in the demand curve, then you would spuriously believe that deregulation raised prices, and not the other way around!

When appealing to a broad correlation between building and rents, YIMBYs leave themselves open to debunkings which, while they miss the broader point, are technically right. It doesn’t have to be like this. There does, in fact, exist good evidence for reducing regulatory restrictions, which I will cover. But much more importantly, we can demonstrate from reasoning alone that allowing people to build more housing is always good. Whatever happens to price, it is better than the counterfactual in which less was built.

Incidentally, a similar thing happens when considering highway expansions. The speed of traffic, which is akin to the price, is not the only thing which we care about. It is certainly possible for adding “one more lane” to leave the price of driving the same – but this could only happen by allowing more people to make a trip altogether!

Thus far we have been implicitly assuming that there are no externalities, positive or negative. If there are negative externalities, then of course restrictions can raise welfare. This is not, however, empirically plausible. New York City is far more valuable than a shack in Death Valley. The credible empirical evidence, such as Ahlfeldt, Redding, Sturm, and Wolf (2015) shows that being located next to other stuff raises the productivity of firms and workers. What positive externalities do mean is that there is now no longer a monotonic relationship between deregulation and prices. A maximally regulated place, where the only structure allowed is a single shack, would of course be valueless; as the place partially deregulates, the value of the land and the cost of housing will rise before falling again as we deregulate still further.

I would also like to point out the danger of not adjusting for quality. The proper measure of price is adjusted for the things you are buying – it would hardly do to make housing “cheaper” simply by making it shabbier. Since new housing is, well, new, it will tend to be higher quality and thus higher price for that reason.

As noted, there are a few ways to find the effect of deregulation. The first and obvious one is to take an event, argue it’s exogenous, and then find the effect on prices and quantities. For example, Kate Pennington (2020) uses building fires in San Francisco. By demolishing the building, it allows for a larger housing unit to be built. She can then find the local effects of more housing, which leads nearby rents to fall. It’s the same story with Andreas Mense (2025), who uses weather-induced delays in when housing is completed, as does Xiaodi Li (2022) or with Asquith, Mast and Reed (2023).

However, these exogenous shock studies, while intuitive to explain to non-economists are not the ideal answer. We would miss the endogenous effect of people relocating from elsewhere in the city or country. If there are positive externalities The proper way to answer this is to build a general equilibrium structural models, and use plausibly exogenous events to identify parameters. Vincent Rollet’s job market paper is the best example of this that I know of, though there are others. The principal contributions of the paper are in moving away from the perfect competition simplification and having developers solve a dynamic discrete game with costs of tearing down a building and building up. He is able to show decisively that, whatever specifications you choose, removing zoning will reduce rents and raise welfare.

Anagol, Ferreira, and Rexer (2023) exploit a reform in Sao Paolo to the zoning laws, which was a general shift in the maximum allowed floor-area-ratio. Some areas already had FAR above the new cap, so you can use the difference in differences to estimate to parameters of a structural model. They find that the price of renting places fell.

The evidence is out there, and it is decisive. Allowing more housing will raise welfare, and on the margin, it will reduce rents. However, just because an argument is correct does not mean that one can make sloppy or fallacious arguments for it. Making poor arguments allows people to feel smugly superior in their misconceptions, and hold onto them for longer. Price is not a sufficient statistic for welfare, nor will casual analysis of prices and quantities tell us the effect of deregulation.