Thread by @KAErdmann

Thread!
Shout it from the rooftops, people!

“Unlike 2005, when we had too many houses…” You see this line everywhere today.

It’s wrong.

It’s worse than you think. There weren’t enough homes, even in 2005.

Here’s my new paper out today @mercatus:

mercatus.org/publications/h…

“Housing is cyclical.” Another line you see everywhere, and this one is true. Before the 90s, housing *supply* was cyclical. Since then, because we’ve tied production up in regulatory knots, supply isn’t very cyclical, so housing *prices* are cyclical instead.

2/

“The Great Moderation” applies at least as strongly to housing supply as it does to GDP growth. All the chatter about Americans keeping up with the Joneses and going all in on McMansions is defied by data.

3/

mercatus.org/publications/h…

So, how is it that the Great Moderation in GDP was supposedly undermined by a massive bubble cycle in housing?
Here are housing permits, population growth, and vacancies. Note that population trends after 2000 are at least as important as housing trends.

4/

These trends were different in different places. Most localities had no building boom. There was a rise in housing vacancies before 2004, largely in rental units in cities where housing production was flat but population growth declined.

5/

mercatus.org/publications/h…

Contagion cities (AZ, FL, NV, etc.), had a building boom from 2003-2004, but even there, vacancies were correlated with sharp downtrends in population. In this case, population growth cratered AFTER the boom.

6/

mercatus.org/publications/h…

Contagion cities were the poster children of “oversupply”. Increased construction in 2004-2005 was followed by neighborhoods full of empty homes. But in cities growing by nearly 3% annually, extra inventory can’t stay empty for long, certainly not for years.
So what happened?

7/

The jump to nearly 8% vacancies in those cities from 2008-2010 was a result of a huge demand shock – a sudden lack of tenants, relative to the norm.
By the time Contagion vacancies peaked, cumulative building was BELOW trend.

8/

mercatus.org/publications/h…

The only city type that echoes the national trend of a housing boom that coincided with declining population growth were Closed Access cities (NYC, LA, Boston, SF).
Do you think in 2005 there was an oversupply problem in these cities and only these?

9/

mercatus.org/publications/h…

Compare the “Closed Access” chart to the “Less Growth” chart, included above. Closed Access cities have population growth similar to “Less Growth” cities (Detroit, Cleveland, etc.), but they have housing permitting rates that are lower comparatively.

10/

mercatus.org/publications/h…

“Closed Access” cities never had inflated vacancies. They persistently hemorrhage people - the most negative rates of domestic migration in the country. High rents are a leading reason for those migration flows. They are the definition of anything BUT housing oversupply.

11/

So, their low population growth is a result of a housing shortage. And, looking at cumulative population growth and housing permits, one can see that the housing boom was adding supply in Closed Access cities, and yet, population growth declined…

12/

mercatus.org/publications/h…

What’s going on? If they needed homes, and the housing boom was creating new supply specifically in the cities that desperately need it, then why was there a depopulation and mass migration from Closed Access cities to Contagion cities during this boom?

13/

Even though the change in housing demand was too small to show up amid the assertion of Census Bureau measures of household size (residents per housing unit), the moderate increase in demand must have been related to declining household size.

14/

mercatus.org/publications/h…

And I do mean MODERATE. Remember that first graph comparing population growth and housing starts?

Back when we ACTUALLY built a lot more homes during boom times, household size was not stable at all.

15/

mercatus.org/publications/h…

Back when housing supply was cyclical, Americans took all the homes available, and we just lived in them. It wasn’t hard. Where building can happen, it doesn’t lead to oversupply. The whole idea is at odds with macro-level housing markets in the real world.

16/

Here is a graph of household size in the U.S. and several other countries over time. Pre-1990s, Americans, and Germans, French, and Japanese households still today do not have stable preferences of household size.

17/

mercatus.org/publications/h…

What do the places with declining household sizes have in common? They could build many more homes than the U.S. did in the 2000s. Generally, their home prices remain moderate, and nobody worries about oversupply in those times and places.

18/

mercatus.org/publications/h…

So, why did we end up with a vacancy and inventory crisis? Because of the perception of oversupply. The Federal Reserve “took away the punch bowl just as the party got going” and then burned down the ballroom for good measure.

19/

mercatus.org/publications/h…

In August 2007, the Fed’s annual Jackson Hole meeting was all about housing. There is a fascinating presentation there by Ed Leamer where he shows that collapsing housing starts are a harbinger of a coming recession.
nber.org/papers/w13428

20/

Yet, he counsels against doing anything about it.

“The historical record strongly suggests that in 2004 and 2005 we poured the foundation for a recession in 2007 or 2008 led by the collapse in housing we are currently experiencing.”

21/

More from Leamer:
“With no lost sales to transfer forward in time, the low interest rates in 2002–2004 transferred sales backward in time, stealing sales that otherwise would have occurred in 2006–2009…”

22/

More from Leamer:

“..In 2007 the housing sector of the economy is now paying the Piper with very little possibility that a rate cut would make much of a difference. Once the wave has peaked and is crashing, there is not much that can be done to quiet the waters.”

23/

Before that meeting, the Fed forecasted stabilizing housing starts, adjusting the bottom downward after each disappointing quarter. After, instead of aiming for recovery in housing starts, Fed staff accelerated downward their forecasts of collapsing housing starts.

24/

We didn’t have a housing crash because of oversupply. The crash happened because of the PERCEPTION of housing oversupply. And, just to remove all doubt, in Ben Bernanke’s memoir, he claimed that the recovery was still slow in 2011 because of oversupply.

25/

Here is a graph of housing starts as a % of existing housing stock over the past 6 cycles, showing the timing of Leamer’s Jackson Hole presentation in 2007, and 2011, when Bernanke was still referencing oversupply as an economic headwind.

26/

mercatus.org/publications/h…

Look back again at the graph of population growth, vacancies, and housing permits in Contagion cities. By the time of the Jackson Hole meeting, Contagion cities were already in a contraction deep enough to halt decades-old population trends.

27/

Contagion city housing starts had already deeply contracted. In a regression of lagged changes in rates of permitting, population growth, and home prices, by 2007, vacancies were overwhelmingly correlated with negative shocks to population growth.

28/

mercatus.org/publications/h…

The 2007-2010 recession was PRECEDED by a deep and destabilizing LACK of housing production. And fourteen years later, it is long past time for us to break out of the upside-down oversupply mythology that breeds crisis and deprivation. We need to build more houses.

End.