Sunday, April 24, 2022

Some Thoughts on Kevin Drum's Housing Supply Post

 A few weeks ago, Kevin Drum wrote a blog post that expressed skepticism about "America’s housing crisis."  He got quite a bit of pushback against that, so he has followed up with a new post, doubling down on his skepticism.

It is a very interesting post, mostly because Drum does such an impressive job at convincingly feeling the parts of the elephant that make his point. The post is a study in rhetoric and in the limits of empirical debate to inform a conversation.  I would take some issue with most of the specific points of fact, but in a way that I suspect a skeptic would consider pedantic.  And, those points of fact are wrapped in a broader assertion that is expressed as a generality - a feeling - that hampers pedantry, and so the whole argument is somewhat protected from criticism.

I suspect that for a reader, this comes across much like my own work does.  My books are full of empirical reinforcement of a core idea, and so for the reader, they are either deeply argued concepts or Gish gallops, and the problem for those kinds of arguments is that they present a skeptic with a lot of work. The skeptic will require a deep engagement with the long line of points in order to have their mind changed, or to confirm their skepticism. The other choice is to write it off as a Gish gallop without doing that work. The problem is that that work is not free! Writing off the full argument can be a rational response! So, I find that with my work, many of the readers are in two categories: (1) those who weren't skeptical or for whom something about the work overcame their skepticism, and they either do deeply engage in it with a sense of self-directed discovery or they accept it as basically true, either in whole, or with some minor remaining potential caveats, and (2) skeptics who don't engage with it deeply. I'm not sure that there is a single skeptic that has deeply engaged with my work, and I can't blame them. It is necessarily a complex web that I have tackled and it would take a lot of work.

It's a bit of a paradox. If my work is worthwhile, it is essential. But, if it isn't worthwhile, then anyone who engages with it skeptically will find at the end that they wasted a lot of effort.

Anyway, back to Drum, the generality I take issue with is comments such as this:

But beyond that there are always individual places that are popular and expensive, and there are individual neighborhoods within those places that are even more expensive. These areas change from decade to decade as different cities get hot, and there's really no way around this. This doesn't indicate a housing shortage any more than high prices for Gucci bags indicates a shortage of purses.

It seems like folksy wisdom, and surely there is a basic truth to it. New generations tend to have it better than older generations, and frequently cry foul at conditions that their grandparents would have embraced or accepted. The current generation is no different than any other generation in this regard. But, this point is simply wrong. This is a new era in American housing. The most economically dynamic cities have been shedding domestic migrants by the hundreds of thousands for years now. That is anomalous. One can see this both within the history of those cities and in other cities. NYC and LA are dominant because when they originally were ascendant, they were magnets for the working class. Of course they grew so large because, given their economic draw, they were affordable!  Looking elsewhere, the norm until recent decades was for metro areas to grow when they are economically ascendent and to stop growing when they struggle.  During the early and mid-20th century, the Detroit metro population grew by nearly 3% annually. Back then there weren't individual cities with incomes far above the norm, like there are today. Over the early 20th century, people moved to opportunity and regional incomes were converging as a result. Today they can't, so some metro area incomes are skyrocketing away from the norm.  This is new!

Also, it isn't the nice neighborhoods that are getting expensive. It's the cheapest neighborhoods in the most housing constrained metros.  Here is a graph of home prices in the LA metro since 2000.  The pressure here is to trade down on the socio-economic status of a neighborhood in order to get into the LA metro. Young Americans aren't complaining that they can't move straight into Beverly Hills. They are actually bidding up home prices in Compton to try to get into a massive metro area that refuses to grow. Population growth of American cities with the most buoyant local incomes has been recently countercyclical. They barely ever grow, and when the economy starts to do better, they actually start shrinking.  This is really weird!  In fact, a decent portion of their ascendent local incomes is really just a side effect of the poorest residents systematically being forced to other cities. This is something that I would expect Kevin Drum, of all people, to be really bothered by. This isn't just the way it's always been.

On his more empirical claims, I will just lay out one of my pedantic points.  He highlights two facts that appear to downplay the supply crisis narrative.  First, he compares cpi housing inflation with cpi inflation excluding shelter.  He finds that housing inflation is up 7% more than non-shelter inflation since 2000. He says that of the data he highlights, this is only one of two metrics that point to a housing shortage, but they are "very tiny ones". I have a pedantic sub-point here. He uses "cpi housing" here, which includes furniture, utilities, etc.  There is a better metric - "cpi shelter", which only includes the actual shelter. But, even that includes things like lodging. You can even dig down deeper and look at "cpi rent of primary residence".  Here's a chart.

The measure he uses shows 7% excess inflation since 2000. "Shelter" excess inflation is 10%. "Rent of primary residence" excess inflation is 21%.  Is that still "very tiny"?

Then, he highlights "rent as percent of household income" from 2000 to 2020, which increased from 22% to 25%.  He counts this as a lack of evidence that rents are skyrocketing.

But, think about this for a minute. If rent inflation is high (and in fact, much higher than his "housing" metric suggested), and yet total spending on rent is relatively level, then that means that American households are greatly decreasing their relative real housing consumption.  Here are a couple of graphs showing how the long term trend in gross rent/GDP started rising in the late 1970s.  Before 1980, there was generally below-average rent inflation and growth in real housing consumption, which averaged out to a level 8.5%ish of GDP being spent on shelter. After 1980, rent kept taking more of US incomes, but after 1980, this was a combination of high rent inflation and below-average real housing consumption. The real consumption hasn't receded fast enough to make up for the inflation, so total spending on housing has increased.

There is one period where the 5 year rolling average in real housing expenditure growth was faster than real GDP growth - 2009-2012.  That isn't because we were building lots of housing then.  Ironically, that's because in 2007-2008, the Fed was convinced that we had too much housing, and as I described in "Building from the Ground Up", they basically lowered American incomes until Americans were willing or able to purchase less housing.  Since Americans were actually lacking in adequate housing and already spending more of our incomes on it than we would like to, this was catastrophic. That error - the same error Drum is still making about the supply problem - was basically the cause of the financial crisis and recession.

Since the crisis only worsened the housing supply issue, Americans have spent the last decade growing real housing consumption at the lowest relative rate since WW II because we have no other choice. Yet, as long as we continue to be deprived of adequate housing supply, this will only partially offset rising rents.

There are several implications of this. First, this certainly doesn't sit well with Drum's general assertion that younger generations expect to hit the ground running, increasing their housing consumption more and earlier than their parents and grandparents had. They are increasing their real housing consumption at a lower relative rate than any previous recent generations did.

Some people will try to suggest that the reduction in real housing consumption is voluntary - smaller families, the norm that as real incomes rise real consumption of housing tends to grow at less than a 1:1 ratio with income, etc. But, if that was the case, rent inflation should be lower. The fact that rent inflation is above average, even though real housing consumption has been low for decades, points to a substantial supply problem.

So, this is a case of Drum highlighting two "parts of the elephant" that look like they push against the housing supply narrative, when, in fact, taken together, they point to a supply problem.

Where I would like to conclude, however, is something that Drum gets right that many people don't.  He closes with:

What about all those investors snapping up houses so the rest of us can't buy them? Isn't that killing off the housing supply?

No. Housing supply is the same regardless of who owns the homes. Besides, corporate and real estate investors buy a small fraction of all the houses sold in America. There are a few specific areas where they're very active, but that's it.

The real lesson from this trend is not that housing supply is tight, but that in certain areas starter homes are selling too cheaply and apartment rents are too high. That's why it's profitable to buy low-end homes and turn them into rentals. This suggests that in certain places there's an imbalance of what's being built and what people want, but that's likely to balance out before long.

This is correct (except for thinking that it isn't evidence of tight supply)! And, I find this interesting for two reasons.

First, it would be easy for a skeptic to produce an empirical retort to his closing statement very similar to his retort here against the housing supply narrative.  They could show that low tier prices have been increasing along with rising institutional ownership. They could argue that when corporate owners buy up more of the housing stock, they ruthlessly raise rents, and that they are raising rents even though new construction has been growing for a decade.  They could point out that homeownership rates are lowest in the places where home prices have been bid up the highest. They could point out that prices are, in fact, rising the fastest where rents are rising the fastest.  Drum is right on this, and they would be wrong, yet there are many empirical points just lying in wait to support a number of sets of premises.

I suppose that means you should be skeptical of all of us. Or, unfortunately, it means that before you can really settle on a point of view here, you're going to have to put in a lot of difficult work parsing the various empirical assertions.

Second, this is a very good example of how mortgage suppression since 2007 has completely muddied the rhetorical waters on this issue, and made it very difficult to do that parsing.

Drum is right that the prices are too low and rents are too high in many places. The reason that is the case is because we killed off low-tier mortgage lending in 2008, which is part of what has made housing supply even worse.

Here's a chart of home prices from 2002-2015, by ZIP code income. The myth about 2002-2006 was that loose lending drove up the prices of low-tier homes to unsustainable levels, and busting the hobgoblin at the center of that myth was part of what we tried to do in 2008. We did it disastrously well.  Over the entire period from 2002-2015, low-tier home prices rose much less than high-tier prices. And, in fact, for the most part, they rose together from 2002-2006, and then we shit all over low tier mortgage lending and created a low-tier housing bust in the shadow of the primary housing bust that people are still making excuses for today. The crappy outcome of sucking 30% out of low-tier home prices relative to high-tier prices is bad enough. We should talk about that. But, that's not the end of it.

At the end of the day, though it's complicated, prices of homes are going to tend toward the cost of alternative new homes. So, you can't just push home prices around to wherever you want. If you push home prices unsustainably low, like we did to low-tier prices with suppressed mortgage lending, they will eventually want to rise again. And, the way they will rise is through rising rents.

As even Drum must agree, based on the charts he highlights, the housing bust did little to bring down rents.  So, now, the next chart, is a scatterplot where the 2002-2015 price appreciation is on the x-axis, and the subsequent rent inflation from 2015-2021 is on the y-axis.

Quite systematically, where a lack of mortgage financing drove prices down, rents increased excessively, until prices rose back up.  The graph (not shown) of price changes from 2002-2021 (rather than 2015) is flat. Prices at the low end and high end have increased similarly. But, to get there, low end rents had to skyrocket, because we made it extremely difficult for low-tier owner-occupiers to create demand as buyers (as opposed to renters), and so supply in the most credit constrained cities has been in the crapper for a decade.  With price/rent ratios still pushed down because of a lack of owner-occupier buying power, the only way for prices to rise back to their natural level was for rising rents to make up for low prices.

(Notice how the recent rise in low-tier rents and prices would seem to bolster the argument of people who falsely believe that the institutional buyers are the cause of all of that!)

Given all of this, if we aren't willing to stop suppressing mortgage lending, then institutional buyers funding build-for-rent neighborhoods is the only solution left for "balancing out" those markets, as Drum suggests they will.  There is a budding movement to stop that from happening, which is just one more step down the road of creating supply constraints and then trying to solve the damage of those supply constraints with more supply constraints. At least on this issue, it seems, Drum will be on the side of allowing new supply.

Or maybe, he'll take the same approach he takes about current supply conditions, when he writes:

Why are you willing to force people to move away from the places they were raised just so you don't have to look at an apartment building near your home?

I'm not. As far as I'm concerned, you may build as many apartment buildings as you want near me. I'm not trying to prevent higher density, I'm just gathering data about the amount of housing in the United States.

So, maybe he'll just watch as all his neighbors fight to block to build-to-rent neighborhoods so that the price and rent levels can't come back into balance, continuing to feel like his conscience is clean, and that the problem isn't severe enough to worry so much about his neighbors' consciences.  In the meantime, we are at quite a crossroad, because, if we find ourselves in a position where working class families can't get mortgages to build their own homes, cities keep fighting multi-unit infill developments, and now we start blocking the single-family build-for-rent developments that are one of the dwindling number of options for reversing our housing shortage, things are gonna get real dark.