Originally published on Kellogg Insight by Sachin Waikar.
Based on the research of Charles Nathanson, Raven Molloy, and Andrew Paciorek
One reason for these soaring prices, of course, is the old principle of supply and demand. If there are fewer homes being built—because of regulatory constraints, such as zoning rules, or geographic constraints, such as mountains or rivers—then prices will logically go up.
But Charles G. Nathanson, an associate professor of finance at the Kellogg School, and coauthors thought a more nuanced study was needed to understand the true effect of supply constraints on housing affordability. Importantly, they believe one must define affordability as something other than simply the sticker price on a new home. After all, housing prices reflect speculation about future sale prices or rental rates. What is needed instead, the researchers posit, is a better definition of housing affordability. With that in hand, they could examine how much supply constraints really do impact housing affordability.
“There’s been a lot of work documenting patterns of home-price increases, but not as much looking at other changes in housing markets and how those affect affordability,” Nathanson says.
Nathanson, along with Raven Molloy and Andrew Paciorek at the Federal Reserve Board, argue that “imputed rent” is a better measure of affordability. This is based on current rents for renters and housing-consumption measures, such as housing square footage and the size of the backyard, for owners. Measures like these, the researchers explain, capture the present-day affordability of housing, as opposed to future speculation.
With their definition of affordability in hand, the researchers explored the impact of regulatory and geographic factors that constrain the number of new homes being built. They find that while supply constraints have indeed affected homes’ purchase prices, they haven’t affected overall housing affordability nearly as much.
“Supply constraints haven’t made housing that much less affordable relative to what you might think,” Nathanson says.
A Tale of Two (Hypothetical) Cities
Nathanson and collaborators first had to devise a methodology for the study.
“An ideal experiment would be if you had two different cities that were otherwise identical, and one would have decided to make it harder to build housing, and one didn’t,” Nathanson says. “Then you see how things play out differently in the first city versus the second as far as affordability.”
Since our universe doesn’t include identical cities, the researchers created a mathematical model to simulate that two-city scenario. Then they compared the model to data from multiple U.S. metropolitan areas, including information about geographic and regulatory constraints, such as the length of time required to obtain a building permit. The goal was to explore how these constraints end up affecting affordability.
“The approach helps us measure how difficult it is to build housing. Then we can use the model to compare places where it’s generally more difficult to those where it’s easier, to understand the impact,” Nathanson says.
They also analyzed the difference in affordability between older and newer housing in a given city at specific points in time. “Regulations on housing grew a lot between 1980 and the 2000s, and that constrains supply,” Nathanson says. “So we compare houses built in the 2000s in a given city with those built in the 1960s and 1970s there, to understand the effect of the regulations.”
A Not-So-Large Effect
The study yielded some surprising results.
Specifically, housing-supply constraints had little impact on measures of affordability other than home prices. For example, a one-standard-deviation increase in supply constraint increases rents by less than five percentage points, about half the increase for home prices. Similarly, supply constraints had no measurable effect on house size, for example.
The results suggest that reducing regulation in major metropolitan areas might not improve affordability much.
“If Boston adopted the regulatory posture of Indianapolis, where it’s easier to build housing, our model shows that would only lower rents by 12 percent, which is a lot less than rent has gone up in Boston over the last 30 years,” Nathanson says.
“That raises the question,” he continues, “of what we could do to make housing cheaper.” One possibility is making drastic regulatory changes, such as removing most home-building regulation altogether. This would be a far more dramatic move than the Boston-to-Indianapolis hypothetical, and the consequent increase in supply could enhance affordability, as could dramatic innovation in construction efficiency and speed.
But these shifts seem unlikely in the immediate future, and they wouldn’t loosen the physical constraints that prevent the building of more housing, such as the water around the San Francisco area.
A Question of Density?
The results, overall, paint a challenging picture for housing affordability in the U.S.: with regulatory changes unlikely to sufficiently move the needle, there’s no readily available tool that could improve affordability in the near term. And people don’t seem likely to simply live in smaller spaces that would increase density in cities, thus increasing the availability of housing.
“That doesn’t seem to happen in reality,” Nathanson says. “Instead of density increasing, a city’s population goes down by a lot. You just have a lot more people migrating out of the area.”