Why the Fed Doesn’t Fear Inflation, But You Do
The core PCE deflator for March recorded a near-record 1.1%. Should we worry that deflation is taking hold?
Well, first of all you should recognize that the PCE, unlike the CPI, is frequently revised and by significant amounts. As the chart below shows, this is only a near-record because there was a massive revision that raised the 2010 low from 0.7% or so to 1.1%. We should be wary, in my opinion, to draw any strong conclusions from (and certainly wary of implementing policy based on) a data series that can have the rate of change revised by 60%.
But still, core PCE is near its lows while core CPI is not. Should we be concerned about deflation? Should the Fed?
There are a number of reasons for the difference. A persistent difference of about 0.25%-0.5% is consistent with differences in the type of formula used and other “normal” differences. The Fed favors the PCE because it has a broader representation of the economy – in that it doesn’t focus “just” on consumers – and because it adjusts more quickly as the composition of spending changes. However, if you are looking at how the costs to you the consumer change, the CPI is the index that you should be looking at.
The main reason that core PCE is currently so much lower than core CPI is that PCE has a much lower weight on housing. And, thanks to the Fed’s loose money policy, it is housing that is driving the CPI higher. The difference in housing weights currently adds 0.31% to CPI compared to PCE. The PCE makes up for this low weight in housing by having a much higher weight on medical care (about three times the CPI weight). Why the huge difference in the medical care weight?
The CPI and PCE metrics are meant to measure different things – the PCE is broader, but the CPI measures specifically expenditures by consumers. Consequently, the difference in medical care weights occurs because the CPI measures spending by consumers, while the PCE includes spending by Medicare, Medicaid, other government entities, the employer portion of health insurance, and other non-consumer payers. Which do you think is more relevant for consumers? And which do you think is a better representation of what a typical consumer spends: 42% on housing and 6% on medical care, or 26% on housing and 22% on medical care? In simple terms, do you spend more for your house, or do you spend about equal amounts on both? I suspect that for most Americans, especially those who are employed and those who are currently receiving Medicare, spending on housing is vastly higher than direct spending on medical care.
Isn’t it convenient for the Fed that right now, they can focus on a metric that is pointedly underweighting the category of expenditure that is most directly being affected by quantitative easing? This is one reason that I do not expect QE to stop any time this year.
Leave a Reply
>>Isn’t it convenient for the Fed that right now, they can focus on a metric that is pointedly underweighting the category of expenditure that is most directly being affected by quantitative easing?<<
In fairness to the current Fed (which I despise), weren't they also using the PCE when housing prices were crashing, thereby leaving the then rate of inflation overstated?
Yes, but actually in 2008 the two measures were right on top of each other. Both were at 2.5% in July 2008. In 2004, PCE was higher (which is weird) and then it was about even for the next several years, which is as ifit was high since it tends to be lower than CPI. But they were going the same direction at the time, similar contours – right now, PCE appears to be dropping sharply while CPI is treading water.
Personally I think that will change, as we project CPI to rise from here and I doubt the non-housing part of PCE will continue to be such a drag. A lot of it looks idiosyncratic (really? Medical care inflation is soft?)
________________________________
Mike, Regarding housing and inflation, is it not possible that an unusually large share of home buyers right now are investors who plan to rent them out, thus simultaneously increasing housing prices while suppressing a corresponding rise in rents? In other words, if a vacant, foreclosed home is first bought (marginally moving prices up), and then rented out (marginally adding to the supply of rentals and suppressing rents), couldn’t we see (perhaps for some time) a dynamic where house prices rise but rents are relatively stable (thereby dampening housing’s contribution to CPI)?
Home prices aren’t particularly low compared to rents – so if investors are paying higher prices and planning to rent them at lower rents, then i suppose we should thank them … but they’re idiots! If they’re paying higher prices, it’s because they think they can let them at higher rents, right?
Mike, I don’t think anyone sets out to buy houses high and rent them low; this would be just the inevitable consequence of too many investors pursuing the same strategy at the same time. A lot of investors right now seem to have the attitude of “investment x looks expensive, but so does everything else, so what the heck, I’ll buy”. Certainly this dynamic is at play in, to cite just one example, the “high-yield” bond market. With so much cheap money sloshing around, looking for a home (pun intended!), weird outcomes are becoming more and more normal.
I’m leery of this particular argument since it is exactly the opposite argument that was made when the housing market collapsed, and lots of Wall Street economists said “that will drive rents HIGHER because people will prefer to rent now instead of buy.”
It bugged me at the time because the Wall Street economists should really have a better grasp of Econ 101. The absolute prices of substitutes almost always move the same direction. What changes in this case is not the absolute price, but the RELATIVE price. This is a key point! So if home prices rise 20% and rents rise 3%, then the economics works…and inflation goes higher even though renting gets to be “cheaper” relative to owning. In practice, this is what has happened – rents move a fraction of what home prices do. BIG movements in home prices have always (always!) been reflected in movements in the same direction, but much smaller, in rents…
Thanks for the explanation of CPI versus PCE. Surprising to me is that medical care inflation isn’t rising just as fast as housing.
One can understand why a broad-based index would incorporate Medicare and Medicaid.
At the same time, though, Medicaid and Medicare pay lowball, administered prices. Every year official budgets assume their payouts will be frozen. But then Congress unfreezes them before doctors down scalpels, hop in the Mercedes and head for the golf course in protest.
Apparently the health industry offsets its Medicare and Medicaid loss leaders by taking it out on the private sector via price discrimination, with thousand-percent mark-ups for uninsured walk-ins. How all this affects the various inflation indices I haven’t a clue.
All I know is that when I first set foot in NYC a third of a century ago, it cost $1.50 to cross the George Washington bridge. Now it’s $13.00. Dollars to dimes, as it were.
Commodities and commodity producers are getting hammered by the economic data. I do think this makes some sense. But it sure does make the rest of the equity market look like the Coyote right before he looks down.
It’s really not so much the level as the market but the divergence between the commodities and the equity market that is disturbing…as well as the completely opposite stories being used to rationalize each of those positions. I think it was Dire Straits who said “Two men say they’re Jesus…ONE of them must be wrong…”