Saturday, September 19, 2015

More Perspective on Housing Starts

Average Annual Housing Starts, 1959-Aug. 2000:

3 Worst 5-Year Periods for Housing Starts from 1959 to 2000:
Sept. 1965 - Aug. 1970:  1,365,000
Jan. 1979 - Dec. 1983:  1,375,000
Mar. 1989 - Feb. 1994:  1,210,000

Average Annual Housing Starts, Sept. 2000-Aug. 2015 - a 15 year period which includes the housing "bubble":

So, the worst 5 year period for housing starts in the modern era was followed by a 6 year period with average housing starts (1,517,000 from 1994 to 2000).  This was followed by a 15 year period that has had lower housing starts than any of the other 5 year periods in the modern era.

Yet, practically everyone I speak to says home prices and rents are going up because we have a too much money!

Apparently, there is nothing too outrageous for there to be a national consensus about it.  There are arguments about efficient markets, which have not been settled.  But, is there any question about what would happen if we had stable regulatory and institutional foundations that weren't bent on adjusting market conditions?  We'd be building houses!  And all the people complaining about demand and bubbles would be buying them.  Because we are only insane in the abstract.  We are usually sane in practice.

The problem isn't market efficiency.  The problem is that consensus leads to policy, and policy has a real effect on the market.  We can argue about whether, on a scale of 1 to 10, market efficiency is a 7 or a 9.  But, public policy can go off the rails.  It is capable to getting pinned down at 1.

Group beliefs are strange.  They say you can't reason someone out of something they weren't reasoned into.  But, we all think we were reasoned into our beliefs.  When consensus beliefs reach a certain level of absurdity, strange group behaviors begin to take shape.  There is some tipping point, where the absurdity of the belief itself creates a barrier against accepting rational, disciplining information, because it becomes difficult for the group to correct in a face-saving way.  We naturally rationalize the delusion.  We are now engaged in a subconscious national Ptolemaic drama.  The problem is that this process is unpredictable because it is not constrained by reason.

This particular issue is not even moderated by arbitrary bipartisanship, because the absurdity has captured virtually the entire spectrum of political viewpoints.  Antipathy to finance is a useful unifying device.  But, the worse things get, the harder the face-saving becomes.  I don't have a confident image of how the American populace will be able to allow housing to normalize.  The mechanisms that trigger new housing will probably lead to recovering prices before they lead to a supply response that will be strong enough to eventually lower rents, and therefore eventually prices.  Can we get there?


  1. Very thought provoking. Query, what happens if you correlate housing starts with a relevant demographic measure such as growth of labor force, or some more targeted demographic? If the prime-age house buying demographic (30-35?) is growing more slowly, housing starts should grow more slowly to keep S&D in equilibrium.

    1. That's a good point, nick. Here is a post where I adjusted homeownership rates for demographic effects:
      Here's a previous post, similar to this post, where I looked at population growth and housing starts:

      Keep in mind, though, that there is a difference between home ownership and housing consumption. I think a big contributor to confusion about housing is that we tend to think of homeownership as a demand factor. But, I think it is probably more coherent to think of home ownership as a supply factor and housing consumption (rent or imputed rent) as a demand factor. Households don't stop consuming housing when they cease being homeowners.

      Adjusting for demographics, etc, can shed some light on these issues, but I think the best way to get a full picture of supply and demand is to look at rent inflation. If rent inflation is high, there must be some constraint on supply, because there is no natural reason why housing supply shouldn't meet demand with relatively stable costs (in terms of rent), regardless of temporary rise and fall in housing demand growth.

      There has been tremendous constraint on supply since 2006 because credit markets collapsed. But before that, there were regulatory constraints that I have concluded are concentrated in the major coastal metro areas. These area are easy to identify. They have very high rents and low levels of housing starts.

  2. Great post. Consider also the regulatory constraints on the supply of retail space in the very places where there is dense populations.
    If I want to sell a tangible good to Kevin E. in the City of San Clemente, I can only do so in a zoned retail space. And no push cart vending allowed.
    No wonder a chronic complaint among retailers is what is going to happen to rents....