The housing shortage and what to do about it

The housing shortage and what to do about it

Matthew Yglesias — Read time: 9 minutes

The housing shortage and what to do about it

YIMBY — now more than ever

Yesterday I argued that the Fed’s planned interest rate increases are too timid based on their own inflation projections. But Larry Summers is arguing that the Fed’s projections themselves are too optimistic and the inflation rate is likely to be higher than they say.

Housing costs are a key part of that argument.

I’ve been arguing for years that the United States has a serious supply-side problem related to land use and zoning regulations.

These days, though, the quantity of new houses being built is lagging behind the number of permits, largely because builders are having trouble getting sufficient labor and construction materials for acceptable costs. This is to say that on some level, the housing market is feeling the general squeeze as a ton of demand rushes through an economy that’s been hit by a couple of big supply-side shocks.

One interpretation of that supply crunch is that YIMBY/NIMBY issues are less relevant today because of all these non-permitting barriers to house-buildings, or relatedly because remote work is probably going to reduce the link between high white-collar salaries and specific metro areas. I think both of these ideas are mistaken; if anything, we’re in a “now more than ever” situation. When material shortages are limiting the number of units that can be built, it’s an especially big problem that we can’t build houses where they are most needed.

Expect rent inflation to accelerate

The Bureau of Labor Statistics’ rent of primary residence data series indicates that the U.S. saw a 3.3 percent increase in rent in 2021, while the Zillow Observed Rent Index says rents rose 16 percent during that same period.

As Tim Lee explains, the series diverge is because the BLS and Zillow are measuring different things. Zillow is looking at the average rent charged for vacant units that are on the market right now — if you show up in a new city and need a place to live, what would it cost you to rent one? The BLS is looking at the average rent that people are actually paying. There are two reasons for the divergence. One is that most people are on leases that last one year or longer, so they’re not actually renting on the spot market. The other is that if you’re a landlord, it’s costly to have a dwelling you own sit vacant for a month or two while you get it cleaned up and listed and wait for a new tenant to book it. Even in a hot market, it’s easy to lose five to 10 percent of a year’s rental income due to a vacancy. So while landlords aren’t averse to raising the rent when they think they can get away with it, the self-interested play is to charge tenants somewhat less than the market-clearing price in order to avoid a vacancy.

Down markets feature the same dynamic in reverse. One time my roommate and I were on a month-to-month lease and noticed a place around the corner that was clearly better than our apartment and was listed for the same amount of money. We tried to use that to get our landlord to cut our rent, but she said no and we felt we had to follow through on our bluff and actually move around the corner. We got a better place for the same amount of money, but moving is a huge pain in the ass and it arguably wasn’t worth the hassle. So it’s a basically symmetrical situation: if the tenants pay rent on time and the landlord fixes stuff on a reasonable timetable, both parties have incentives not to drive as hard a bargain as they would absent the existing relationship.

This is all just to say that average rents are stickier and less volatile than the vacancy rents that Zillow measures.

Both concepts are useful, and the BLS measure is the more relevant one if you’re trying to measure the welfare of the renting population. But the Zillow measure arguably tells you more about the macroeconomic situation. When we say the price of oil is $90 a barrel, we mean the marginal price of a barrel of oil — the price you would pay, right now, if you needed to buy some oil. There’s a lot of oil zipping around the world at all times that costs different amounts depending on when the contracts were made, etc., but “the” price of oil is the spot price.

In a paper published in February, Marijn Bolhuis, Judd Cramer, and Larry Summers find that the BLS series tends to converge on the Zillow series with a 13-month lag. That leads Summers to think the Fed is lowballing its inflation forecast. I’m not sure whether or not that’s right — his argument seems plausible to me, but bond market expectations are similar to the Fed’s forecasts and I think the traders know how to look up Zillow data. But either way, the housing situation is a significant constraint on growth.

Housing supply is increasing — sort of

Prices and quantities both matter, so one good question to ask is whether the supply of dwellings is rising amidst the price pressure. And the answer here is that it kind of depends on how you look at it.

If you want an optimistic take, this 10-year chart shows a clear pattern of growth off those Great Recession low points.

But if you zoom out, you’ll see that while we’re no longer experiencing catastrophically low levels of house-building, we’re still not back to what would have been a pretty “normal” level in the 1980s or 1990s.

This is particularly bad because the output of manufactured homes (“trailers”), though rising, has never recovered from its late-1990s mini-peak, which itself was way lower than the early-1970s peak (see this post on how the manufactured housing industry was sabotaged).

For a slightly rosier view, you can adjust the housing indicators for the fact that population growth is slowing down and look at new units divided by new people. This measure exhibits weird seasonal spikes because the monthly population accounts get adjusted, so don’t take it too literally. But just by eyeball, you see that new housing per capita really has recovered.

So my overall take on the housing supply situation is “not great, not terrible.”

Two more points I would make on this are that in some sense, we care about prices because we care about human well-being, and the Zillow-type information probably understates well-being in this case:

As this Morgan Stanley analysis says, part of the reason housing demand is so robust is that we’re seeing an increase in household formation — fewer adults living with their parents or roommates. This means an increase in individuals’ housing costs but a rise in their actual standard of living.

Most Americans own their home, so the increase in housing costs is actually an increase in their wealth rather than a decline in their living standards. And the homeownership rate seems to be rising.

Long story short, this housing inflation is different from the situation where gasoline gets more expensive and now everyone has less money and they’re mad. Most people don’t pay rent at all and are actually benefitting from the surge in housing demand. Some of the renters are taking advantage of rising income to get their own places while they were previously sharing a home. And we are adding new homes in the United States to meet the demand, at least to some extent.

But overall demand for everything is really high and there are various supply chain stresses, and it’s making it harder to achieve housing abundance.

Construction costs are rising fast

The government generates lots of different inflation indexes, including a Producer Price Index that has a handy sub-index called the Producer Price Index for Net Inputs to Residential Construction.

In other words, this is the price of stuff that you have to buy in order to build a house. It has gone up a lot since the pandemic started. So much so that if you deflate Zillow’s home price index by the PPI-NIRC, it looks like prices haven’t risen at all.

It would be a bit too cute perhaps to say that this shows that all the increase in house prices is purely due to a rise in construction costs.

But at a minimum, I do think it shows that there’s not some special new pathology in the housing market. We do have a longstanding pathology where the price of a new home in Los Angeles is much higher than the cost of building a house in Los Angeles. That’s because land in Los Angeles is very expensive and the zoning rules do not allow homebuilders to take advantage of the technology humans have developed to fit many homes onto a fixed parcel of land. In terms of recent changes in the housing market, though, what we’re seeing is that construction costs have soared.

This seems to be basically a question of commodity prices. There was a lot of attention paid at one point to a huge surge in the price of lumber, and prices have come down a bit from the peak. But there’s all kinds of other stuff in a house like copper and cement and it’s just all gotten more expensive. And that makes land use issues more important, not less important.

YIMBY more than ever

One view is that if the total number of houses that get built is limited by the supply of lumber and copper and cement, then it really doesn’t matter at this point if Austin or Nassau County or Wellesley, Massachusetts are raising too many regulatory barriers to infill development.

That seems wrong to me. First because overly prescriptive regulation prevents builders from economizing on building material. Most people prefer to live in detached houses so they don’t need to worry about intrusive noise from a house that they share a wall with. On the other hand, attached townhouses are cheaper because they consume less land and also because that shared wall means you don’t need as much lumber. That lumber-efficiency issue hasn’t normally been a very important consideration in the U.S. housing market, but if you drive around Ireland they have tons of shared-wall duplexes even in places where land is cheap (Ireland is very sparsely populated) because that’s a cheaper construction method.

But the other reason is that the land costs still count from a productivity perspective. Four units stacked vertically on a $1 million parcel of land generates more dollars’ worth of housing services than four units spread out across $700,000 worth of parcels. If they both consume an equal amount of lumber and copper and cement, the former is still generating more economic output per unit of input because it is taking better advantage of the available natural resources.

The one saving grace of the whole situation, of course, is that a larger share of work is being done remotely these days, which has made localized housing shortages less binding than they used to be. One reason to be a little less alarmed by rent inflation than the statistics indicate is that the most expensive cities have seen the smallest increases because people are now more able to exit from pricey markets if they want to. That absolutely does make YIMBY/NIMBY fights in certain specific places less significant than they used to be. At the same time, more flexibility in where people can live makes inflexible land use more significant in other places. We seem to be seeing booming demand for living in scenic parts of the Mountain West and coastal Maine. If those places don’t zone to accommodate new construction, we’ll just recreate the dysfunctional housing paradigm we had before, except instead of “decent commute to a handful of downtowns,” it will be based on access to recreational amenities.