The CPI is not an accurate measure of inflation, someone anyone who is not rich realizes every time they go shopping.
I totally agree that daily and monthly expenditures are very important as they do affect us in the wallet.
But, you're interpretation is incorrect. It is clear that they both track average inflation the same. They are equally accurate because they average the same.
The EPI is more volatile. Notice how it goes lower just as it goes higher. As such, it is less precise for the average rate of inflation.
These are accurate and precise definitions of accurate and precise. "Accurate" means that the measure averages about the real thing it is measuring. "Precise" means that the measure is less volatile about whatever its average is.
Then there is a simple question of what basis is there to claim the EPI is "better"? Better at what? What's the methodology?
I just went shopping and prices were lower from the week before. Maybe it's got to do with where I live and shop, don't know.
The BEA has hundreds of tables of information regarding the rate of inflation. The is the CPI-U, CPI-W, CPI-U less housing and energy, and hundreds of detailed tables.
Here is a full report of the CPI and components, readily available on the BEA website,
http://www.bls.gov/cpi/cpid1203.pdf. It is a hundred and fifteen pages. It covers every single detailed. Wisky at home was -1.2. Women's dresses was 8.8. Hair, dental, shaving, and miscellaneous personal care products was 1.5. Flour and prepared flour mixes was 4.1. Fresh sweetrolls, coffeecakes, doughnuts was 5.5. Uncooked ground beef was 10.5. (holly crap). Eggs was -13.7. (We should have been eating more eggs). Other poultry including turkey was 12. Dried beans, peas, and lentils was 19.5.
This alone makes the premise, that the CPI-U is not precise or accurate, simply absurd.
Now, check out
The Everyday Price Index (“EPI”
Is Flawed | TheArmoTrader
This so called "inflation index" obviously puts significant weighting on "gas" tremendously over the other components. Had all the components been weighted equally (11% weighting for each), the "EPI" would have averaged out to 4.25%, which is a little more than half of what the real EPI index came out to.
Especially when that index (the EPI) only measures "39 percent of total household spending". So what they are basically saying here is that the inflation rate for 39% of my expenses is 8.2%. They are totally disregarding the other 61%. Doesn't sound like a smart thing to do.
Measuring just 39% of my expenses to gauge inflation is not really a great way to measure the inflation rate. 61% is a statistically significant portion of my expenses.
Matt Yglesias over at Slate.com made this very same point. To sum up his post (I suggest reading it, its really short), he basically argues that:
* Non-Everyday goods make up a large portion of our yearly expenses because they are more expensive.
* We buy and deal with "everyday goods and services" more frequently, thus giving us the perception that inflation is higher than we think.
* Ironically, everyday goods (like Food, Gas) are more prone to inflation from supply-shocks & demand increases from the global economy than from changes from "debasement" monetary policy by the $FED. Those expensive items are a better tell for monetary inflation.
All in all, the EPI is less precise. It is simply another methodology for measuring part of inflation. And it is questionable if it is all that valuable.
What you can do is take the CPI-U, subtract the yearly average, multiply by ten, then add the yearly average in. That will get you the same thing as the EPI.
Why not just go with shadowstats.com? That will give you a really big number. You'll like that, it's really big.
Or, rather, seeing as what really want is a CPI-ME, you should just calculate your own using your grocery list.
----
p.s. I was discussing the changes in the standard of living with Ken Stewart at the BEA. Here is what he says about how quantities and prices affect the CPI.
Within what we call a "weighting period", we effectively assume a fixed quantity of cars, bananas, etc when we add up indexes across items.
But, periodically, we do take a survey of expenditures. If the quantity of bananas or whatever has changed, we ‘link’ that into the index; i.e., we do not let the change in quantity affect the change in price in bananas, etc.
Notice, though, that over time the relative “weight” of an item could grow. Let’s say, for example, that the relative weight of bananas grows over time, because people spend more on bananas.
While we effectively don't let the quantity of bananas effect the rate of inflation for bananas, it now has a bigger 'weight' or relative importance in overall inflation than it used to, and so the future price change for bananas may well have a bigger effect than it used to on the overall rate of inflation.
So I guess one could say that, while the increasing quantity of bananas purchased does not affect the banana index, the resulting higher relative weight of bananas over time could make bananas a more important influence on overall price change in the future.
Yeah, that made it as clear a mud.
