SHELTER COSTS STILL DEFLATIONARY
Last Friday’s CPI report was hardly cause for concern with regards to the inflation outlook. The core year over year rate was astonishingly low at just 1.2% while the exclusion of food and energy actually reduced the year over year rate to just 1%. While the PPI reports are clearly showing higher rates of inflation it’s also clear from the consumer inflation data that these costs are not getting passed along. This is just one more clear sign of weakness at the household sector.
While food and energy tend to catch the media spotlight (they are the most volatile and noticeable after all) they’re not the largest components of consumer costs even if you combine the two. Shelter costs are by far the largest input in the consumer cost equation. A look at last week’s data shows the deflationary tendencies in this data. Econoday highlighted this over the weekend:
“In the core, shelter costs continued to stand out—as weak. Just when you thought they could not get softer, they did. Shelter costs were flat in August after rising only 0.1 percent in each of the prior fourth months. Again, this reflects the weak housing market and also a sluggish travel market for lodging while away from home. Shelter costs are so week that on a year-ago basis they were down 0.5 percent in August.”
We actually include housing costs in the equation you get a similar deflationary trend. Based on my data housing costs are just marginally deflationary at -0.25% year over year. I’ve taken the CPI data a step further and actually added the Case Shiller data. Since roughly 70% of Americans are homeowners I replaced 2/3rds of the housing component in the CPI data with the house price data. Thus, we still account for the rental costs, but don’t entirely ignore the real cost of a mortgage.
The BLS has a relatively controversial way of calculating the housing component of the CPI – they use what is called owners equivalent rent. This estimates what your house might rent for if you were so inclined. What’s misleading here is that the data doesn’t really reflect consumer costs and thus psychology. The last decade is a prime example. Rising home values have a wealth effect. Consumer spending is bolstered in such an environment. Likewise, consumer spending will deteriorate when an asset as important as your home deteriorates. Since the housing peak in 2006 we have seen an 8% rise in the OER figures from the BLS despite a 30% decline in housing prices. Clearly, there is a disconnect here. If you look at my alternative CPI chart it’s clear that inflation was running quite hot leading up to the housing bust and then fell off a cliff and has remained deflationary or low since:
The CPI data leaves a bit to be desired. All in all, I actually think the CPI report is quite thorough, incredibly granular and very transparent. But that doesn’t mean it is not without flaws. I certainly can’t say that swapping the Case Shiller data out with the OER figures is a more accurate indication of inflation, however, it would seem to me that there has to be a more accurate way to account for the changing costs of a monthly mortgage payment than the CPI report currently does. Considering how vital this is in the monthly spending decisions of households you would think that this component would be more accurately accounted for. Based on my simple findings I would argue that inflation (and deflation) has been dramatically under & over reported in the last decade.
Thanks for the chart!
What if the government had actually used these numbers for CPI, with a target of keeping inflation between 2% and 3% annually?
1. COLA’s would have kicked in, and that would certainly have caused concern.
2. Interest rates would likely have been pushed up by the fed, thus averting some of the housing bubble.
The whole point of measuring CPI is to help structure economic policy, right? So if this model could have averted the bubble, why not use it instead?
Might not have stopped lax lending standards however, which had a sizeable role in pushing up housing prices.
Also, when you get the time to think about it, can you explain why fuel costs are excluded from CPI (if indeed they are). I seem to recall hearing that fairly often. I realize it is volatile on a month to month basis, but that could be addressed by some kind of rolling average, couldn’t it?
I have looked at some presentations that state that inflation has been intentionally understated by the goverment to serve some agenda (re-election). Presentations usually have their own agenda, but at least some of the information is plausible.
Finally, measuring a monthly mortgage payment is pretty tricky. Changing interest rates, loan balances, and different loan types all play a part in arriving at that number. Probably simpler to stick with a “price”, rather than a payment.
I’m pretty sure that the basket of goods doesn’t look at a new vehicle payment, but the price, even if the buyer is more focused and impacted by the payment terms.
It would seem to me that the best way to include housing costs is the use the average historical cost of a monthly mortgage payment. I dont know if that data is compiled.