Facets of the Housing Shortage, A Series: Part 6

Approximately 97% of American land is rural. There is no lack of homes in most of that space. There is no housing shortage in 97% of the country.

Among those who don’t dispute the housing shortage problem, there is unanimity that there is a shortage in about 0.5% of the country’s land, which is within the few coastal metropolitan areas I refer to as the Closed Access cities (Boston, New York City, San Francisco, Los Angeles, San Diego, and Honolulu, if you’re being thorough).

In the rest of the country, there are fast-growing, slow-growing, and shrinking places, and there is some broad agreement that some fast-growing cities have struggled to permit enough housing to remain affordable.

Housing is sort of an extreme example of the 80/20 rule. It’s more like a 97/3 rule, or if you really dig into locational details, a 99/1 rule. At any given point in time, under any number of supply conditions, there will be plenty of housing in 99% of the locations where homes could be built. No matter how bad housing is (families paying elevated rents, migrating away from personally cherished locations for economic reasons, delaying the formation of households and families, homelessness, etc.) there will not be an economic case for a 2,000 unit mixed use residential development 20 miles outside of Altus, Oklahoma, or in countless other locations across the country.

Sometimes this leads to people talking past each other, because when there is a shortage of homes in the 1% of the country where homes would be built (with the range of economic and cultural stresses that the shortage causes) a skeptic can still say, “There are cheap houses in Altus.”

Why does everyone make a big deal about the Dust Bowl? There were jobs in California.

That’s the unserious end of the discourse. Some rain in Oklahoma would have been the preferable outcome, preventing a number of measurable social stresses. Of course, when preferable outcomes are unavailable, people move to less preferable outcomes. That doesn’t mean that the less preferable outcomes deserve political support, or that it would be ok to oppose rain in Oklahoma because going to California is an option.

Housing supply truthers are even worse than that, because saying “There are cheap homes in Altus.” is the economic equivalent of marching the Okies back to where they came from.

Even among those who take the shortage seriously, there can be a tendency to mistake and misapply this 99/1 rule in a way that leads to underestimating the scope of the problem. At this point, the shortage has stretched to nearly every large metropolitan area - fast growing and slow growing. At this point, a region really needs to be perpetually shrinking to have avoided evolving into a shortage position.

And, just like the deniers might say that the United States can’t have a housing shortage because there are cheap homes in Altus, a well-meaning YIMBY might mistakenly say that Buffalo can’t have a housing shortage because there are neighborhoods in Buffalo where homes are cheap. It’s just as wrong at the city level as it is at the national level.

I have written a few posts about whether new units “pencil”. Sometimes, upzoning skeptics will say that zoning reforms won’t help because other issues - financing, supply costs, local preferences - mean that builders can’t profitably complete units even if zoning allowed them. My point is that at any given time, there are countless projects that don’t pencil for every project that does. That ratio doesn’t change in different supply conditions. I can run the numbers right now for a project that wouldn’t pencil in suburban Phoenix, central Cleveland, or in gentrifying parts of Brooklyn. That project doesn’t matter. It’s a 99% project.

And, I would actually agree, in a macro-level sense, that at the current moment, upzoning won’t immediately heal the American housing market because financing did become the binding constraint for single-family homes in 2008. That cratered our homebuilding industry so much that it is still struggling to re-establish the capacity to construct a sustainable quantity of homes each year. Under current conditions, upzoning would change which homes are constructed more than it would change the total quantity.

That is why there is a national shortage (in the 1% of locations where there is demand for housing) even where zoning hadn’t been so binding that it created economic and social stresses like rent inflation and homelessness. So, in the short run, the general struggle to re-establish capacity to build is the main governor on the reversal of the national shortage. In the long run, upzoning will eventually be necessary to reduce economic and social stresses caused by inadequate housing in locations where zoning is binding enough to produce outmigration.

This all means that there is a shortage in Buffalo, even though it is a slow growing city. There is a shortage in Phoenix, even though zoning hadn’t been a binding constraint on total quantity before 2008. There is a shortage in New York City because zoning has been a binding constraint. But under current conditions, upzoning in New York City might mostly move construction activity to New York from other locations, reducing the scale of elevated housing expenses in New York City, but not necessarily speeding up the process of correction, nationally.

Upzoning will make housing more abundant and more affordable, in the long term, and it will make more lifestyle choices available to families under various socio-economic conditions. But, there are many tricky confounders that convince both skeptics and supporters that upzoning won’t help, and which lead to peculiar outcomes that might even make it look like it didn’t work, in hindsight.

The 99/1 rule is one of those tricks. Vincent Rollet at MIT recently noted that in a broadly upzoned New York City, most of the new construction would happen in neighborhoods where zoning had kept density far below demand and where demand would produce high property values for new units. Even within New York City, it’s a 99/1 rule, or at least 90/10.

A recent study by the Urban Institute helps to highlight this issue. They looked at some targeted upzoning reforms in Philadelphia and New York City. There were 9 local upzoning reforms in New York City between 2016 and 2021, which led, on average, to an increase of more than 2 housing units per upzoned parcel while other parcels across New York City averaged little more than 0.1 new units per parcel.

That is a significant effect. But, nearly half of the new units came from one neighborhood - Gowanus - that only represented less than 10% of the upzoned parcels. In other upzone neighborhoods, like Jerome Avenue and East Harlem, the effects were small or negligible. In practice, 99/1 rules at several scales.

And, this is where I think some essentialism creeps into general points of view about the breadth of the shortage and the value of upzoning. Even in New York City, there are many locations where upzoning would not lead to adequate construction. If you made sure to avoid the locations best suited to more infill growth, you could upzone 90% of New York City to minimal effect. And, yet, broad upzoning would easily lead to construction activity multiples of New York’s current rate of building.

If the Urban Institute’s neighborhoods are representative, if New York City erases its shortage of housing amounting to 20% growth, more than 95% of the parcels in the region wouldn’t host any significant construction at all.

You could say something similar about Buffalo. Buffalo doesn’t have a 20% shortage, but there are now homes in Buffalo with elevated rental value. Residential construction trends since 2005 have accumulated about 20,000 fewer new homes than previous construction trends would have, in a region of about 400,000 homes.

But, the 99/1 rule lends itself to essentialism. There are always many places where homes don’t pencil. Since Buffalo has been growing slowly, and even has shrunk at times, its shrinking neighborhoods have become symbols of the whole region. So, there is this notion that the way Buffalo has to grow is to reclaim its abandoned neighborhoods and fill them back in. And, until those abandoned neighborhoods begin filling in on their own, Buffalo can’t claim to have a housing shortage or an affordability issue. There are affordable neighborhoods in Buffalo that nobody is moving into just like there are affordable homes in Altus, Oklahoma that nobody is moving into.

YIMBY activists will sometimes make that claim, neutering their own message and advising policy choices that actively undermine their own moral and practical goals. America can have an important housing shortage even while Altus, Oklahoma exists within America. New York City can have an important housing shortage even while Jerome Avenue and East Harlem exist within New York City. Buffalo can have a shortage even while there are vacant lots in Emerson and MLK Park.

Should New York City require a minimum number of new homes on Jerome Avenue and in East Harlem before they upzone Gawanus? What would be the point of that? This attitude that cities like Buffalo shouldn’t or don’t need to encourage more lenient lending, new apartments, or single-family rentals because some formerly populated parts of the city haven’t recovered yet is basically creating the same outcome as the absurdist policy I noted above - upzoning the 90% of the city where new projects won’t pencil.

And, as far as I can figure, this is all a product of logical fallacies, such as creating a category that Buffalo fits in, where it is defined by its least successful neighborhoods while New York City is put in a different category where it is defined by its most successful neighborhoods. The difference between New York City and Buffalo is more one of scale than one of kind. All the same things are happening in both cities. In an unconstrained market, there might be more potential for new residential construction in New York City than in Buffalo, but under current conditions, they are both on the same side of the tipping point. The numbers might be different, but the signs are the same, and the processes and reforms needed to get there are the same.

There are places in Buffalo where nothing would be built under any zoning regime. But broad upzoning, or lending reform, or construction capacity, or expanded rental market options, that applied to all of Buffalo, would increase construction and lower rents. That same basic point could be made about almost every major American city, or about the country in general.

America has an important housing shortage. I assert that normalization will require at least 15 million homes. Those 15 million homes will be constructed in neighborhoods making up less than 1% of American land. We all agree that there may be cheap homes on the other 99%. The existence of the other 99% doesn’t mitigate against the shortage. Some of that 99% is land that previously was part of the in-demand 1%. The existence of those peculiar parcels also doesn’t mitigate against the shortage. Unfortunately, the 99/1 rule sometimes steers both supply skeptics and supply advocates wrong.

Large parts of New York City are unlikely to see residential construction at any significant scale, regardless of changes in zoning, lending, or input costs. Small parts would see major investment, which would create a rebirth, financially and in quality of life, for millions of existing and new New York families. That point is generalizable across markets and geographical scales. Please apply it.