Is Housing Holding Back the Recovery?
Evan Soltas thinks so. With both inflation and unemployment figures finally pointing in the direction of a rate hike, only meek housing numbers portend any extended period of low interest rates. This is important because, as Evan notes, housing has consistently been a key driver of monetary policy and the crash in construction leaves a big dent to this day.
I’m a little less convinced this is cyclical factor deserving monetary nourishment as much as a structural change in the housing market itself. This isn’t something I’m too sure about, and this is largely a note for my benefit. While Evan says “It [the weak recovery] is no less apparent if you include multifamily housing”, I’m not so sure.
Now, single-family houses are definitely the most important part of the market, which is why a weak recovery in 1-unit structures will persist in a graph of the market as a whole. But the consistent improvement in 5-unit housing starts, that is apartment buildings, is curious.
It tells us a little bit about where the recovery is and is not. People in New York City live in apartments, people in Tennessee do not. It tells us a little bit about the changing structure of home ownership. People rent apartments but own houses. And surely enough, rental vacancy has fallen over a third from its double-digit peak in 2009. The divergence in housing starts is just the market responding the a change in consumer preferences.
A reasonable doubt may be that consumers are starting to rent precisely because the recovery is so week, and that I am diagnosing a symptom, not the disease. However, that is unlikely to dismiss this theory. Homeownership rates have been in persistent decline since well before the crisis, and the recession didn’t really accelerate the trend.
This is a trend that will certainly become more important in years to come. While Evan is worried that the housing market remains depressed, the rental market is headed towards steamy recovery.
That kind of “v” recovery in rental prices is exactly the kind economists love to see. Unfortunately it’s not for the economy as a whole, confirming a longer term shift in underlying economic dynamics, towards a nation of renters. While single-family housing starts may be underwhelming, apartment creation will be necessary for affordable housing over the next decade.
And while the consensus that the taper didn’t affect the overall recovery, but did slam housing, seems to be true, lumber futures – a good gauge of where markets think housing will go – have been going strong.
It’s clearly true that tapering, and tightening in general, will slow the recovery of housing relative to recovery of the economy as a whole. But that’s not necessarily a bad thing. Rents are a superior gauge of economic activity as they are more reactive to underlying labor market indicators – like job creation, consumer confidence, and employment – than home prices which ultimately reflect an asset and hence are sensitive to the magnitude of the initial drop in price.
Sure, single-unit construction dominates the market. Yet homebuilder ETFs have shown a stronger recovery than housing prices in general, or one-unit construction in specific, would suggest. The market, I think, is pricing in the eventual need to vastly improve America’s multifamily apparatus, as there are no signs a decline in home ownership will halter.
And, as I’ve written before, the argument that student loans reflective of inadequate demand is holding back housing is not strong. Fact is, an overwhelming number of students want to live in places where they can’t afford to buy a home.
But more importantly, those jumping at the correlation between student loans and home ownership over a few years of data are not paying enough attention to large, structural shifts in economic geography over the past fifty years. For nearly a century, suburban growth eclipsed urban. Millenials, the fraught group in question, are changing that. Two-thirds of young graduates now want to move to cities for a better job, compared to a fraction not too long ago. Not to mention the more than 80% that are willing to move to any city if needed.
Dad is no longer a company man, nor mom a housewife. Rather, graduates are likely to vie for shorter commitments focused on training and, in a number of cases, with a higher probability of relocation in the future. Not to mention the logistical, locational difficulty of maintaining a dual-income family (specifically outside of urban centers).
And that’s the demand side. Jobs that cater to college graduates are slowly disappearing from middle America toward coastal centers that capitalize on economies of scale and network effects. Vox notes that the age at which graduates first purchase a house is becoming later. True, but not necessarily relative to household formation itself – something happening later across the country driven by graduates. (Not to mention, as a commenter on Twitter mentions, the increasing necessity of a post-Baccalaureate degree).
So if the argument is that further monetary accommodation will improve the housing market, I’d buy. However, that’s different from the argument that high investment in single-unit housing is the weakest link in the recovery. And the Fed shouldn’t be making choices about structural trends.
I’m in favor of keeping rates lower for longer only as an insurance that the left tail of doing too little is worse than the expectation of doing too much. Not all labor market indicators are decidedly healthy – employment of 25-54 year olds, for example – and current levels of inflation aren’t overwhelming enough that we can be sure they’re here to stay. We can afford to be a little behind the curve.
Pingback: Assorted links
I suggest you compare household formation to housing starts.
Back in the 1960s and 70s the baby boomers growing up created sufficient demand to justify 2 million housing starts a year.
Now,the rate of household formation is half what it was back then and there is no way it would justify even 1.5 million housing starts each year.
Given weak household formation you should only expect 1.1 to 1.3 million starts.
This info is worth everyone’s attention. How can I find out more?