Two Ways of Viewing Capital and Real GDP Since 2000

In the closing paragraph of my last entry I offered two hypotheses about the post-2008 US economy. The first is that “real GDP has shifted to a lower path because of a shrinkage in the economy’s productive capital stock — a problem that better monetary policy (not feeding the boom) could have helped to avoid, but cannot now fix.” It is reasonable to suppose that the capital stock has shrunk, I argued, because the housing boom diverted investible resources from more productive capital formation into housing construction. The second is that potential output, as estimated by the Congressional Budget Office’s method, “is currently overestimated because capital wastage has not been fully recognized.”

Here again is the chart that frames the common account of our recent macroeconomic history, showing the paths of actual real GDP and of the CBO’s estimate of potential real GDP, this time in natural logs so that a constant growth rate corresponds to a straight line with constant slope:

White Image 1

This picture of the estimated “output gap” suggests no unsustainable boom in the US economy before 2007. There was no bubble. There was merely a return to full employment after the previous “dot-com” recession of 2001 pulled output below potential. The Great Recession of 2007-09 then appears not as a reaction to an unsustainable path, but as a bolt from the blue, an exogenous shock. The initial drop in real GDP has to be explained by going off chart, e.g., by reference to the bursting of the housing bubble. But the housing bubble is itself unexplained by macro data, not part of any general malinvestment-and-overconsumption boom.

Next, as Market Monetarists have emphasized, households responded to the start of the recession by hoarding money, reducing aggregate demand. As I showed last time, there was indeed a jump in hoarding (as measured by the ratio of M2 balances to GDP) during 2009. The Fed failed to increase the quantity of M2 in response, so aggregate demand did fall, which in a sticky-price world brought down real output. (In 2009 the CPI and PCE price indexes also fell, but in this view not enough to clear the markets.) This nominal shock helps to explain some part of the severity of the recession, but it can’t be the whole story. It can’t explain why the economy has remained below its potential output level for more than six years. It cannot explain why recovery to potential output has continued to fall short for so long. That remains a puzzle. The “output gap” has shrunk only because the potential output path has been revised downward, a revision explained by shrinking labor force participation.

The account of macroeconomic events that I prefer can be framed by taking the same path for actual real GDP, but instead contrasting it with a simple constant growth-rate path that extrapolates from the 2000-2003 trend, as follows:

White Image 5

This picture suggests that, between 2003 and 2008, real GDP rose unsustainably above its old trend. The recession brought a return to reality, and then some. Consistent with the view that the unsustainable boom was fueled by Federal Reserve credit expansion, here is the bulge in real M2 before and during the period:

White 4

Since 2009 the economy has followed a lower real GDP path, with no tendency to return to the old dashed path, let alone to the bubble path of potential output as estimated by the CBO. To explain that, I suggest, we need to recognize a drop in the stock of productive capital goods due to the misallocation of investment to housing construction during the housing boom.

Consistent with capital shrinkage, the Bureau of Economic Analysis shows gross private domestic investment making a negative contribution to real GDP for nine consecutive quarters, 2007Q3 to 2009Q3 inclusive. The CBO method of estimating potential output does not recognize any capital wastage during the period, however. The CBO’s data website reports a continuously rising value for its capital services index, an input to its estimate of potential real GDP, during 2000-2014. This is of course consistent with its continuously rising estimate for potential output.

White Image 4

I don’t know the literature on econometric estimation of the size of the capital stock well enough to criticize the CBO’s method in any detail, or to propose an alternative method that would give us a better way to estimate whether the path of capital accumulation has been shifted downward. I would be grateful for pointers to any sites that use a method distinct from the CBO’s to provide explicitly derived estimates of the path of productive capital.

[Cross-posted from]