Every attempt to come up with a single number (a price index) that reflects the change in the purchasing power (PP) of money is bound to fail. The main reason is that disparate items cannot be added together and/or averaged to arrive at a sensible result. However, some price indices are less realistic than others. In particular, some well-meaning private-sector efforts to come up with a consumer price index (CPI) that does a better job than the official CPI have generated some of the least-plausible numbers.
One of the most popular alternatives to the official US CPI is the CPI calculated by Shadowstats.com. As I noted in a previous post, it always seemed to me that the Shadowstats number was derived by adding an approximately constant fudge-factor to the official (bogus) CPI to essentially arrive at another bogus number that, regardless of the message being sent by real-world experience, was always much larger than the official number. As I also noted at that time, economist Ed Dolan did some detective work to determine the cause of the strangely-large and fairly-constant difference between the Shadowstats number and the official number. It turned out that Shadowstats had made a basic calculation error that caused its version of the CPI to consistently be at least 4.5%/year too high even assuming the correctness of its own methodology.
Another alternative CPI is called the Chapwood Index. The components of this index were selected based on a survey of what Ed Butowsky's friends and associates spend their money on (Ed Butowsky is the index's creator). The prices of the 500 most commonly purchased items were then added together to generate the index. Not surprisingly, considering the methodology, the result is not a realistic measure of the change in the dollar's PP or the cost of living. As evidence I point out that if the roughly 10%/year average increase in the general price level estimated by the Chapwood Index during 2011-2014 is correct, then the US economy's real GDP must have been about 25% smaller at the end of 2014 than it was at the end of 2010*. In other words, if the Chapwood Index is an accurate reflection of PP loss then the US economy now produces about 25% less goods/services than it did four years ago. This is not remotely close to the truth.
When assessing the validity of economic statistics it's important to use commonsense. A statistic isn't valid just because it happens to be consistent with a narrative that you wholeheartedly believe.
*I arrive at this figure by approximately adjusting nominal GDP by the Chapwood Index, that is, by using the Chapwood Index as the GDP deflator.
http://tsi-blog.com/blog/blog-default/
