Why the CPI might underestimate inflation

[sophie]

I (like many others here) have been wondering how the CPI can claim 0% inflation or close to it, while costs seem to be going up all around us for everything except gas.

I was reading a Consumer Reports article on, of all things, toilet paper.  They reported that while prices have been stable, the amount of paper per roll has decreased by as much as 20-30%, depending on the brand.  Costco was the lone exception, but they reduced the number of rolls in a pack from 36 to 30.

Since then, I've noticed a lot of people buying groceries on Amazon complaining that product sizes have decreased.  Less chips in a package, smaller sizes of individual items, etc.

Is this a universal trend?  If so, it foils the CPI calculations that check prices of items but use unit measures that can hide this kind of shrinkage.  And if it's a widespread practice it could certainly affect the overall numbers.

[mathjak107]

all weight and quantity changes are reflected in the cpi changes

[mathjak107]

what you have to remember is the country is actually made up of  1500 mini economies .

the cpi is not a personal cost of living index . it is only a price change index that only logs price changes in a basket of goods and services .  some you may use a lot ,some you may not use at all . some i may buy once a month , some you may buy once a week .

so what inflation means to our personal rate is going to be very different .

we live in a rent stabilized apartment and have no rent increases voted for next year . but others in private homes are getting rent increases .

in fact my sister refinanced and saved so much all  her expenses are below where they were years ago  even with the increases .


consumers  report answered that very question about shrinkage  :

"The good news is that, as clever as packaging pickpockets are, they can’t fool the U.S. Bureau of Labor Statistics, which is the federal agency that calculates and keeps track of the Consumer Price Index and the CPI-W for wage earners and clerical workers (a subset index used for Social Security cost of living adjustments).

BLS pricing experts are aware of and adjust for shrinkage.

So yesterday’s 99-cent two-ounce candy bar that’s still 99 cents today but has been downsized to 1.8 ounces doesn’t get away with not having raised its price in the eyes of the Feds. Economists at the BLS count that as a 10 percent price increase.

[craigr]

The CPi might underestimate inflation for a very simple reason: It's a government agency run by a political appointee that may be trying to make his/her boss look good.

I have felt for years that the CPI probably underestimates inflation by some amount. What that amount is would vary I'd think by actual inflation. But it's not unreasonable for instance for typical inflation of say 4% to have a CPI representation that was more like 3-3.5%. Inflation that was high of around 12% maybe CPI would show 10%?

So you know it's close enough that people can argue about it all day long and not be able to prove/disprove it, but still would be enough to save the government a lot of money on cost of living index payments, etc.

The MIT Billion Prices Project would seem to validate what I'm saying above:

http://bpp.mit.edu/usa/

Basically, I wouldn't trust the CPI except in gross terms. Trust the markets on what they think inflation actually is. If companies are reducing product weight and not raising prices, then they think there is inflation. If that amount they are doing is higher than the reported CPI (so there is a 4% reduction in weight vs. say 3%), then the companies think the real inflation is 4%, not 3% as CPI may state.

The above is just an off the cuff example. But hopefully you get what I'm saying.

In 2011 I was in Argentina and official inflation was like 20%, but unofficially it was way over 30% and the market prices were showing this. Always believe the markets over the government.

The above is also why I don't like CPI linked investment products like TIPS. You don't want the people causing the inflation to tell you what inflation is. They have a lot of reasons to lie about it and historically, they always do lie about it.

[mathjak107]

a big factor is the quality of goods you buy . higher end goods carry higher mark ups  than cheaper goods so there is a quality factor involved in your personal rate of inflation .

i have to say just the fact i am now on cobra because i am to young for medicare has blown this years expenses out of the water  but that isn't an inflation issue .

but it all boils down to this country being 1500 different mini economies and we all see things differently .

[dutchtraffic]

CPI



Unemployment



Find all real data here: http://www.shadowstats.com/

[dutchtraffic]

Even admitted in this document:

Of the nation's 80.9 million families, 80.1 percent had at
least one employed member in 2014.

20% of all families have zero familymembers with a job.

source: http://www.bls.gov/news.release/pdf/famee.pdf

A family, as defined by the BLS, is a group of two or more people who live together and who are related by birth, adoption or marriage. In 2013, there were 80,445,000 families in the United States and in 16,127,000—or 20 percent–no one had a job.

[dutchtraffic]

Using the official CPI data in any of your calculations is absolutely pointless, as the data is always false, i assumed everybody would understand that?

Which also means a lot of portfolios end up looking a lot less interesting

[dutchtraffic]

Shadowstats, made up, eh..? No.

https://en.wikipedia.org/wiki/Shadowstats.com

"Indeed, over the past few years, some of Mr Williams's (shadowstats) views on economic indicators - the consumer price index, in particular - have been echoed by better-known and leading investment community figures, such as the bond investor Bill Gross, the strategist Stephen Roach and James Grant."

[craigr]

I think Shadow Stats overstates the case dramatically.

I generally think that CPI is probably understated by a small amount. Again, close enough that economists can argue back and forth on it and never reach agreement.

But to say that inflation is 10%+ when the CPI says it's round 3-4% is just too much. I've been in countries with double digit inflation and there is simply no way to hide what is going on. Prices change quickly and black markets come out of the woodwork. People start using foreign currency for transactions. Etc.

So mostly the Shadow Stats stuff just doesn't ring true to me because when inflation is truly double digits it's just plainly obvious there is a problem to everyone. The 1970s U.S. were the best example, but I was too young to remember much of the impact. However mortgage rates were in the mid-teens, etc. so again it's hard to hide from the reality of inflation that is +10%. It's just in your face each day.

[dutchtraffic]

It could be overstated for sure, the truth will be somewhere in the middle then.

Either way, backtesting portfolio results with govt. supplied economic data is just spitting out false data.

[AnotherSwede]

Which also means a lot of portfolios end up looking a lot less interesting

With a reported inflation of 1%, true inflation of 3%, nominal returns of maybe 4-5% doesn't look too great.

Stock and gold-portfolio is looking more and more attractive with NIRP and 3-ish % inflation. But first we need a crash.

[drumminj]

With a reported inflation of 1%, true inflation of 3%, nominal returns of maybe 4-5% doesn't look too great.

Really? That looks great to me -- Not only am I preserving my purchasing power, but my wealth is passively growing, with relatively low volatility and diversified risk.

Sign me up!

[AnotherSwede]

"Maybe" is the keyword
Taxes is the hidden keyword.

[MachineGhost]

Annual rates of inflation via the MIT Billion Prices Project:

Here's the problem with MIT & CPI.  No one buys hundreds of thousands of items every month all averaged together to get an inflation rate.  Personal inflation is like a stock portfolio.  The less stocks you hold, the higher your average return will be.  Hence why using the 1980 methodology is the most accurate.  You can't say inflation was so obvious in the 1970's but then change the calculating methodology thereafter to understate inflation going forward.  That just diverges the situation into a nominal inflation vs real inflation.  Our current real world experience does not mesh with official rates of [nominal] inflation.  The gap is just too large, but its not at Venezuela or Argentina levels...  yet.

It would be very easy to prove what real inflation is by crowdsourcing on the Internet by having people track the prices they paid for the same goods and services they buy every month.  Sort of like Gas Buddy.