So here's my somewhat rambling criticism of ShadowStats inflation figures and methods. 1. John Williams of ShadowStats sells a newsletter, rather than inflation data. Given the contents of his newsletters (e.g. he has been forecasting a hyperinflationary collapse in the US since 2005), one should at the very least be aware of possible bias and treat the ShadowStats publications with a pinch of salt. 2. Williams has been very reluctant to participate in any sort of a dialogue with academic economists who critique his methods. His methodology is not public and not open to peer review (unlike the methods used by the BLS etc). This, again, is a warning sign. 3. Some of Williams's claims are falsified by simple arithmetic. For instance, he estimates that the "formula effect" caused by the adoption of the geometric mean in lieu of the Laspeyres index (this has been done to account for some substitution effects) is to lower the YoY CPI by arnd 3.0%. This is arithmetically impossible and the actual effect is arnd 0.3%. Again, a big warning sign. 4. Williams has stated on record that he doesn't actually go back and recalculate his alternative CPI using the old methodology, contrary to what you'd imagine. Instead, he applies the "spread" (which BLS calculates) to the official figures. The strange thing is that the BLS estimates the magnitude of the spread to be between 0.3 and 0.7% annually, rather than something arnd 4%, which is the spread according to ShadowStats. Williams has never explained clearly how exactly he arrives at the spread, which is a problem. 5. It's exceedingly curious that, while ShadowStats CPI has apparently been running at the annual rate of smth like 5% (just eyeballing it), the price of a 1 year subscription to the ShadowStats newsletter has stayed at $175 for the last 8 years (that's deflation, it seems). I find this very inconsistent and, as a result, rather problematic. 6. I attach the plots of two other alternative, independent measures of inflation. The first is the PriceStats inflation index (online prices, formerly known as the MIT BPP), while the second one is the AIER EPI. None of them (and no others I have seen) produce estimates of inflation close to that of ShadowStats. So this is why I, generally, am very skeptical of ShadowStats. I could add a few other points, but that's the idea.
"5. It's exceedingly curious that, while ShadowStats CPI has apparently been running at the annual rate of smth like 5% (just eyeballing it), the price of a 1 year subscription to the ShadowStats newsletter has stayed at $175 for the last 8 years (that's deflation, it seems)." Rofl.
The inability to pin you down and the "stand corrected" remark taken in proper context (not just yanked out of the 8 page thread like you just did) was because you started off the thread saying you would be "Careful suggesting that he (Santelli) knows what he is talking about" and then, after pages of debate with me and Mav trying to get you to tell us what he was wrong on, you back-peddled to saying you just didn't like to watch him and that you were just stating your own personal preferences. You can't just imply someone is wrong and then have your ultimate answer be "because I don't like him" - which is essentially what you did.
I recall it differently, TT. I gave you specific points where I disagreed with Santelli and, thus, thought he was wrong (see page 10 of the old thread). Subsequently, in my discussion w/Maverick, I provided a lot of specific facts to support my opinions. Just because you then disagreed with me doesn't mean that I was vague or elusive. Eventually, I also concluded that I am not a big fan of Santelli's style. I don't understand why this is back-pedalling. At any rate, this is my recollection of the other thread and I just don't see it the way you do, I'm sorry.
The AIER EPI certainly has some merit, though the weighting factor given raises some questions. The BPP simply doesn't have enough data yet.
The rest of his "somewhat rambling criticism" smelled of trying too hard to rationalize. To me anyway.
I'm not 100% sure, but I think I am generally paying more for stuff than say 2 years ago... I mean - wtf - are people too stupid to notice - or do they need someone with "authority" to tell them otherwise...?