Let me summarize quickly. On the surface, the personal consumption report released by the government on Friday shows that real PCE is still at an all-time peak. In fact, a deeper look shows that real PCE ex medical care peaked in November. Does that mean the recession began in November? Could be….
As those of you reading this blog and my stories now, government statisticians traditionally include many items in personal consumption which actually don’t correspond to household spending, as you and I understand it. For example, all of the nation’s medical spending is counted as personal consumption, but 85% of that total goes through Medicare and health care insurers, and never touches consumer hands (see this story here). In an important sense, that money is government and corporate spending, rather than true consumer spending.
So I removed medical care from PCE, using the detailed statistics published by the Bureau of Economic Analysis on its website. These numbers are subject to revision, of course, but this calculation gives an indication of what is happening.
Below is a chart of real PCE, and real PCE ex medical care, starting with September 2007 =1.
You can see that once we take out medical care, the rest of PCE peaks in November and has been heading down since then.
I can extend this calculation. There are two major “imputed” categories of PCE—that is, big categories of spending which don’t actually correspond to any dollar outlays.
These are "rent" for owner-occupied housing and services furnished without charge by financial institutions. The first category, which totalled $1.1 trillion (at annual rates) in February, represents the estimated value to homeowners of living in their homes. The second category, which totalled $238 billion (at annual rates) in February, represents the estimated value to consumers of things like free checking. These two categories together--roughly about 13% of personal consumption--do not correspond to any direct monetary consumer outlays, so they will not drop immediately in recessions. Medical care, at $2.1 trillion, is another 21% of PCE, so at least one-third of consumer spending is not subject to household spending decisions, at least in the short run.
So I am now going to remove all three categories from real PCE, in the chart below
I took this back to March 2007. You can see that the two lines match each other, until November. After that, reported consumer spending is being held up by these three "non-consumer" categories--medical care, imputed rent of owner-occupied housing, and free financial services. The part of consumer spending actually under the control of households has dropped since then.
Two caveats on this calculation. First, I used the BEA's monthly detail, which is subject to major revisions. Second, I can't swear that these calculations are correct (it's saturday afternoon, and I should really be outside).