The title of this post comes from an old joke about economists. It’s a play on one of the standard critiques of the discipline: that economists are so wedded to the realm of pure theory that they take the output of their models to be more conclusive, and therefore more “real,” than any empirical evidence to the contrary.
It’s a good joke, which is why I use it here, but the serious criticism behind the joke is outdated. Over the past few decades, economists have increasingly turned to empirical methods in their research; this has led them to findings that run directly counter to what you might expect from the output of a simple theoretical model. A classic work of the “empirical turn” is David Card and Alan Krueger’s 1993 study of a minimum wage increase in New Jersey. When they compared fast food restaurants in New Jersey to those across the border in Pennsylvania, they found that the minimum wage increase had not led to a decrease in the former state’s fast food sector workforce — despite what a simple “Econ 101” supply and demand model might suggest about the relationship between labor costs and employment.
People on the left (including myself) have often been critical of those who doggedly weight crude economic models over actual evidence. This is one of the accusations that supply skeptics often level at YIMBYs: that instead of looking at the world around us, we rely on a vulgar caricature of the housing market. It’s an unfair accusation, and it grows more unfair with each year as empirical researchers uncover more evidence that building more housing leads to greater broad-based housing affordability.
Ironically, as the empirical case for YIMBYism grows ever stronger, it is the anti-”Econ 101” crowd that is now falling back on airy theorizing and simple models. It turns out that their problem with non-empirical economics wasn’t the lack of rigorous real-world investigation after all; it was just that they would rather have everyone use models that are rigged to produce their preferred outcomes.
Case in point, a recently published working paper co-authored by Michael Storper, one of the supply skeptic crowd’s favorite academics. Titled “Inequality, not regulation, drives America’s housing affordability crisis,” the paper seeks to demonstrate that increased market-rate homebuilding will not make high-cost American cities affordable on anything like a reasonable time scale. The authors’ primary tool for doing this is a simulation of a hypothetical supply shock to six high-cost cities.
It’s an odd methodological choice given that we have no shortage of actually existing supply spikes to study. For example, over the past couple of years, Austin, Texas has seen both a significant spike in homebuilding and a precipitous drop in average rents. There’s some debate among economists and housing policy experts about the extent to which the rent decline can be attributed to the supply shock; it’s probably a few years too early to measure the impact of Austin’s building boom with any precision, but certainly not too early to attempt some preliminary estimates. The research on upzonings in Auckland and Minneapolis is already piling up, and before too long we’ll also be able to gather some initial findings from New York City’s recent zoning and planning reforms.
Instead, Storper and co. provide us with a model with just a handful of inputs: their assumed price elasticity, filtering rate, and rate of supply increase. Which is not to say that they ignore empirical evidence entirely; they just make selective use of it in their literature review and ignore the methodological flaws in papers that support their preferred conclusions.
For example, Storper et al repeatedly cite a paper that purportedly finds new market-rate homebuilding can actually increase rents in the surrounding area for low-income tenants. But they fail to note that the paper didn’t adjust for inflation, and that doing so causes the observed effect to vanish. Meanwhile, they wave away the empirical evidence on chain-of-moves filtering (where higher-income residents move into newer and pricier rental housing, making their prior homes available to residents on the next income rung beneath them) by suggesting that people who move into recently vacated housing may then “experience higher housing cost burdens.” That strikes me as a strange claim; if none of the recently vacated housing is actually becoming more affordable, why wouldn’t more people in the chain of moves stay put? No one’s forcing them to move up the chain. At the very least, if you’re going to argue that participating in the chain of moves simply increases the rent burden for most households, it would be helpful to offer some explanatory mechanism.
But the biggest problem with this paper is one that is common to the genre: it is arguing with a straw man. Very few of the people they call “deregulationists” think that regulatory reform on its own is sufficient to make cities like San Francisco affordable to non-college-educated workers. As I recently argued in a brief for the Roosevelt Institute, regulatory reform is a complement to targeted public investment, not a replacement for it. So-called deregulationism can actually make public investments in affordable housing and rental subsidies more effective.
It’s sort of funny that supply skeptics continue to grind out these papers that treat all YIMBYs like market fundamentalists. That has never been true—Sen. Scott Wiener, one of the OG YIMBY politicians, has always been a champion for deed-restricted affordable housing—but it appears even more ridiculous now that the most prominent YIMBY mayor in the country is a democratic socialist who wants the public sector to build hundreds of thousands of new affordable housing units. In both its rejection of empiricism and in its reliance on hidebound stereotypes about YIMBYs, this paper feels more like a time capsule from last decade than a novel intervention into the housing debate.