Monday, April 11, 2016

Private Investment and Recessions

I saw this reference to an upcoming Auclert and Rognlie paper (HT: EV) on inequality and the business cycle.  I can't see the full paper, but the abstract says:
A temporary rise in inequality, if not accommodated by monetary policy, has an immediate effect on output that can be quantified using the empirical covariance between income and marginal propensities to consume.
This seems like a strange idea to me.  Aren't falling profits and falling private investment leading indicators of coming recessions?  If that is so, then it seems odd that a lower propensity to consume would be a causal factor in a production contraction.  Maybe the key here is the monetary policy bit.  Maybe they are using interest rates as the proxy for monetary policy, so that a model that produces a downward shift in the natural rate while stable monetary policy is defined as a fixed interest rate, would lead to contraction.  I don't know if that's what's going on, but otherwise, it seems strange.

This led me to call up this graph:

Thinking about our current context, private domestic investment, minus residential investment, actually doesn't look that bad.  It's not blowing the roof off, but after the deep drop in the recession, we have recovered to pretty typical investment levels.

If we look at the top line, which includes residential investing (or the bottom green line which is the measure of residential investment), it looks a lot worse.  The recovery level is below all other post-war recovery levels, and is still roughly at levels we would normally consider recessionary.

This is the story of this decade.  Millions of construction workers have been sidelined.  How much of the persistent cyclical downshift in labor force participation was due to our misplaced concerns about homebuilding?  Unemployment has dropped in construction during the recovery, but labor force has not recovered.  (The employment and unemployment lines are stacked.)  So, 1% of the labor force was in construction and now has either left the labor force or moved to other industries.  Meanwhile, rents are skyrocketing and residential investment remains about 1%-2% below any past levels.
Meanwhile, real GDP growth seems to be about 1%-2% below typical recovery levels.  These seem like pretty obvious dots to connect.  We shut down mortgage lending, and for a decade, we have basically put 1% to 2% of the economy on mothballs.  There is one way to fix this economy, and only one way.  But, it would require the equivalent of admitting that we shouldn't have gone to war.  These sorts of decisions are not easy to reverse.
Given two choices: (1) it was probably a bad idea to invade Iraq, even if Hussein was a really bad guy, or (2) maybe, in general, banks were just kind of doing their jobs, and we should have taken steps to stabilized the mortgage industry as early as 2006*.
Which one would a plurality of Americans more easily cop to?  The almost universal initial reaction to the idea that I would even suggest #2 is indignation.
* I would say even as far back as 2003-2004.  One of the factors that I see in the data is that the expansion of the subprime industry itself seems to have been mostly a reaction to sharp cuts in mortgage growth at the GSE's and the FHA - especially in 2003 and 2004 after Congressional witch hunts into GSE accounting practices coincide with backpedaling at both Fannie and Freddie.  The rise of subprime, itself, was a sign, not of excess, but of stress - and of both supply and demand deprivation.
This seems hard to believe when residential investment was so strong in 2004 and 2005.  But, residential investment was inflated because (1) Closed Access housing policies in the large metro areas pressed housing expansion into parts of the country where single units were more prevalent and units of a given value required more fixed investment and (2) the high prices of the Closed Access cities inflated brokers' commissions, which are included in the aggregate measure.  If we adjust for these things, residential investment was at normal non-recessionary levels.
Measures are stacked
(We can see in this graph how low multi-unit investment had moved.  A unit-for-unit exchange of expensive downtown urban condos in the Closed Access cities for single unit homes in Arizona or Texas would, on net, reduce total residential investment because more of the value would come from location - less from materials.  This substitution from metropolitan condos to McMansions in the heartland probably added about 1/4% to 1/2% of GDP to residential investment during the boom, with no net benefit to the real value of the housing stock.)


  1. Great post.

    Who would ever think housing development and financing would become demonized in America?

    1. The Fed did it, Ben. Growth is slow because the Fed wants it that way and I have been writing about it and no one listens. Well, I now have proof, Christopher Phelan, quoted in the article I just wrote: