September 01, 2011
The CPI, and Some Key Components
A perennial topic of discussion is the deficiencies of the CPI in measuring the things that are important to “real people”. I actually believe that there is a point to some of these critiques. In particular, we know the CPI is “plutocratic” in that the weights associated with the CPI bundle are consistent with the expenditure shares of a household somewhere in the 4th income quintile.   However, I think that other critiques -- namely that food, health, and transportation, are missed -- are misguided.
Figure 1: Log CPI (blue), food and beverages component (red), medical care (green), and transportation (purple), seasonally adjusted, all rescaled to 2008M09 = 0. NBER defined recession dates shaded gray. “The Luskin moment” is the month that Don Luskin argued that the US was on the verge of “accelerating prosperity”. Source: BLS, NBER and author’s calculations.
Note that while there has been a tremendous run-up in transportation prices (driven by gasoline prices), it has only regained its position relative to the overall CPI recently. That might be surprising to some, but I think that surprise reflects the fact that people pay attention when transportation prices go up, they don’t so much when they go down (or stay constant). On the other hand, the medical care component continues to outpace the overall CPI.
Posted by Menzie Chinn at September 1, 2011 05:21 AMdigg this | reddit
It's important to run this 10 years back where headline CPI has significantly outpaced core CPI, in large part from energy. If we look at just the last 5 years, because of the recession, things don't look quite as bad.
Posted by: Tristan Bruno at September 1, 2011 07:14 AM
The link to the Luskin moment reference is broken.
Posted by: Rodrigo at September 1, 2011 07:47 AM
In 2004, non-discretionary expenses consistently exceded half of disposable income. (This data is before the gas and food price increases this year.)
1. Upward price spikes highlight neglected savings.
2. Price declines almost never are large enough to bring discretionary disposable income back up to pre-skipe levels.
3. Demand for savings goes up for upward price spikes, but less income is availible for savings. This decade, prices do not fall enough to meet the increased demand for saving or make up for the inadequate savings during the high price period (not only are many people not saving, they are using consumer debt to meet their expenses). Expenses need to be lower (relative to discretionary disposable income) than pre-spike and for a longer period of time than the spike to get total saving to the point that savings can fall to pre-spike level (this assumes that savings were not being neglected before the spike, much more saving are needed if that's the case).
Posted by: aaron at September 1, 2011 07:50 AM
The Luskin moment link is broken. Perhaps this is the one that was meant,
Posted by: John B. Chilton at September 1, 2011 08:28 AM
links to Luskin's 9/14/2008 piece which stated in part:
Overall, the pessimists are up against an insurmountable reality: In the last reported quarter, the U.S. economy grew at an annual rate of 3.3 percent, adjusted for inflation. That's virtually the same as the 3.4 percent average growth rate since -- yes -- the Great Depression."
BEA shows that the annualized growth rate for 2008QII was only 1.3%, which followed 6 months of virtually zero growth. Luskin was forecasting a booming economy but BEA shows his forecast was WAY OFF!
Posted by: pgl at September 1, 2011 08:38 AM
There is a hidden cost that is the exported inflation,there is a hidden danger, the perverse effects of the exported inflation coming back home through private debts and or sovereign debts inflated in amounts.
As an illustration of above comments,may I suggest reading two Econbrowser posts and the last Bloomberg column
The Correlation between Money Base Growth and Inflation
Explaining Recent Trends in Household Savings
London Loses as Banks Book Trades Overseas
"This time is different"
One may believe that exporting paper assets is not drawing on a country liquidity,is not borrowing on the country macro accounts,is not inflating the country s external debts.
There is a price for each item of the banks balance sheet used and none in the buffet my neighbor macro accounts?
Posted by: ppcm at September 1, 2011 08:54 AM
my wife and i are both scientists, living in Boston , so it is "natural" to discuss inflation as the ratio of 1st year tenure track professor salary to house or college.
I think, by these measures, professionals, except at the very upper end, are not doing as well as their parents on housing (boston phenom) or college (national)
also, inflation is really a psychological phenomenon, because what we buy is what we want which is what we see around us (silence of the lambs, first we covet...)
so if the upper 1% are building larger homes, inflation is rising, because we covet larger homes....someday, economics will grow up and recognize that statistics are driven by psychology
Posted by: ezra abrams at September 1, 2011 09:25 AM
The highest economic priority has to be to use inflation to get people out of negative home equity in their mortgaged homes. Period.
Since homes are leveraged, this should be accomplished with a mere 10% cumulative rise in CPI from these levels.
Whether that 10% can be accommodated in 2, 3, or 4 years is the question. Ideally, it should be in a shorter time.
Price increases in food and energy are a very small price to pay in exchange for this. If a person has -$100K in home equity, and can get out of that by carpooling for a year, that is an easy tradeoff.
And price inflation is mostly in processed junk foods, not whole fruits and veggies from the farmer's market. Inflation in processed junk food is a positive, as Americans eat too much of that anyway. Food price action is doing a good job of steering people out of junk and into fruits and veggies.
Medical costs are a problem. This is directly due to government distortions and lawyer parasitism.
Posted by: GK at September 1, 2011 10:31 AM
@Aaron, thanks for the chart. It is unsourced. Do you have the source?
Posted by: Frank in midtown at September 1, 2011 01:09 PM
I used the BEA PCE data.
Posted by: aaron at September 1, 2011 02:08 PM
The basket weight of healthcare in the CPI excludes most healthcare spending. Thus CPI understates inflation for the uninsured, or when compared to total compensation rather than wages.
Posted by: benamery21 at September 1, 2011 06:27 PM
I imagine that the education component of the CPI is a bit controversial too, seeing as there are huge differences in how much schooling costs. Would it be better to split that component off too? I gather it's been running at 5-6% for the past decade, making it as much of a problem as healthcare.
Posted by: endorendil at September 2, 2011 02:49 AM
I can only say what's important to me: the difference between wage inflation and asset price inflation minus goods inflation. That is my standard of living index, and positive means I am getting richer. wages go up with a lag, so usually headline inflation hurts. Yes sometimes it goes back down, but its assymetric.
When people complain about inflation my observation is that it's not one particular item, its because their budget has shrank (goods inflation > wage+asset inflation). food and gas in particular are a tiny part of the budget, but serve as a signal that other stuff is going up too.
Posted by: dwb at September 2, 2011 04:24 AM
We are in 'Worse Situation' Than in 2008 Roubini
Roubini says "60% chance of recession"
Now, that's an interesting assertion. Are we in a worse situation that 2008? I'm not so sure. There are no real assets bubbles to speak of. Housing looks about two years away from seeing some health--but no price bubble. Stocks aren't over-valued. Inflation remains low.
What's worse is that the various government entities have shot their powder, and it's time to pay the piper. What's coming next is Volcker, but on the fiscal side.
In any event, I would welcome some thoughts on whether Roubini's assertion is right, and to what extent.
Posted by: Steven Kopits at September 2, 2011 06:02 AM
Some interesting 10-year CPI changes (Jul01-Jul11):
Posted by: Tristan Bruno at September 2, 2011 10:11 AM
Also, for reference, in the 10-year CPI change (Jul01-Jul11):
Posted by: Tristan Bruno at September 2, 2011 10:22 AM
There is no way we can be 'worse' than 2008, as both financials and housing are at levels vastly lower than in 2007, so there is nowhere near the same distance to fall.
The S&P500 had ZERO earnings in Q32008, for God's sake. That is a once-in-a-lifetime occurrance.
Europe could soon be in a situation like the US's 2008, though. This would cause a small recession in the US, but not a huge one..
Posted by: GK at September 2, 2011 10:27 AM
Tristan Bruno's stat that Core CPI increased 21% in the past 10 years, while "Headline" CPI--to wit, actual measured inflation--increased 27%. That Core & Headline don't equal over a 10 year period is more evidence of how rotten the Core is.
Headline CPI under-measures inflation at least moderately in my view. Core just seems like an excuse to let the mandarins at the Fed do what makes them feel most important: Inflate.
Steven Kopits, there are bubbles, tho' they aren't fully inflated. Treasury bonds might be closest. Student loan debt next. Sovereign debt globally is on its way. But in rough agreement with you, assuming Europe 'kicks the can down the road' and Asia doesn't hit the wall right away, there doesn't appear to be a bursting bubble to trigger Roubini's prediction.
Posted by: Bryce at September 2, 2011 12:10 PM
Any discussion of how the CPI is broken, without any mention of OER, is deficient - remember how we got here?
Posted by: Jim D at September 2, 2011 02:43 PM
Steve and GK,
Is a submarine that is sinking in worse shape as it sinks or lying on the bottom?
Posted by: Ricardo at September 2, 2011 03:45 PM