For the last few years there has been a continuing debate about government statistics. Some people argue they are seriously flawed, some think they are fine. The site most cited in this argument is Shadow Stats. Shadow stats made some claims regarding inflation which I (regrettably) bought into. The BLS later debunked these claims. For a summation of this situation, go to this link.
I have not found a single article or paper from a person in the field of statistics or economics that makes a credible claim that the government statistics are so seriously flawed as to be devoid of any substance. In fact, after doing a search of the social science research network's paper database (www.ssrn.com) I could find no paper near that point. In addition, I'm a big consumer of the financial press and I have not seen any raging debate about government numbers.
In short, if you are going to argue that the government numbers are flawed, please present credibly evidence -- not a website, not an "I just know it" argument. And Shadow Stats has already been debunked.
Subscribe to:
Post Comments (Atom)
7 comments:
I invite Bonddad to defend the peculiar birth-death model used by the BLS to reduce unemployment numbers by postulating huge numbers of jobs being created by new businesses started during a severe economic downturn.
I also invite Bonddad to explain how banks rated by the FDIC as "well capitalized" can fail shortly after receiving this designation.
Finally, I would ask Bonddad to explain how mark-to-market accounting was banished by government action just at the time when it would have rendered many major banks insolvent.
Government statistics honestly follow the rules when they do not obstruct policy. When policy dictates, the rules are waived.
The BLS has defended their birth death model. Here is a link to the BLS webpage that explains it.
http://www.bls.gov/search/?cx=011405714443654768953%3Abtgxl8qv780&cof=FORID%3A10%3BNB%3A1&ie=ISO-8859-1&prefix=&q=birth+death+&filter=0&sa=Search#967
Alot of initernet ink was spilled about how the stress tests were terrible. How many of the banks that were tested have failed?
The mark to market model is dealt with by the Financial accounting standards board. Here is a link to their website. http://www.fasb.org/home Feel free to write them a letter.
I asked for anyone who dislikes the numbers to please provide evidence- - as in an academic paper written by an economist or an accountant that has been published and peer review. You have not done that.
The following was written on 5/20/08. Note that the GDP I predicted at that time, 1.8% lower, was later revised by the BEA from a revised +0.9% to -0.7%, a difference of 1.6% lower:
"The Consumer Price Index for All Urban Consumers (CPI-U) increased 0.9 percent in March, before seasonal adjustment"
http://www.bls.gov/news.release/cpi.nr0.htm
That is an annual rate of 11% inflation not the 3% reported after the "seasonal adjustment".
"Real gross domestic product -- the output of goods and services produced by labor and property located in the United States -- increased at an annual rate of 0.6 percent in the first quarter of 2008"
http://www.bea.gov/newsreleases/national/gdp/gdpnewsrelease.htm
The 0.6% increase was based on the aforementioned bogus inflation data. Had the untampered data been used the resulting GDP number would have been -1.2%. Likewise, the prior quarter would also have registered and negative number.
U-3 (official unemployment rate), April 2008 - 5.0%
U-6 Total unemployed, April, 2008 - 9.2%
http://www.bls.gov/news.release/empsit.t12.htm
The underlying unemployment rate is almost twice the headline rate. Thus with inflation and unemployment running at about 10% each the misery index is at 20%.
Oh, did I mention the Birth/Death jobs adjustment for April was 267,000 including 45,000 new construction jobs? This despite a 12 % drop in home building, layoffs increasing by 67% and claims for unemployment rising. The 20,000 job loss would have been a 287,000 job loss without this adjustment.
From CNN:
"Surprisingly, the Labor Department reported that seasonally adjusted energy prices did not rise from March's levels. Though unadjusted gasoline prices were up 5.6% in April from a month earlier, seasonal adjustments made to that estimate resulted in gasoline prices being down 2% on that basis.
That's despite the fact that gasoline prices posted a string of 16 straight daily record highs in April, according to the motorist group AAA. Other measures of national average gasoline prices, including the U.S. Department of Energy and AAA, showed gasoline prices and rose an average of 9% to 10% over the course of the month. Crude oil prices rose 11% in April."
Thank as always for your stimulating articles.
The data is not "untampered." Seasonal adjustment is a standard statistical tool to adjust data is such a way as to see what is really there instead of seasonal changes.
For example, it's standard for retail to hire starting at the end of October and fire after the Christmas season is over. Without adjusting for this data we would be saying in October "retail is going on a tear" and in February "Retail is really hurting." The same is true with the seasonal adjustments you mention above.
As for the birth/death model note that home starts have been increasing for the last 3-5 months. In addition, in the article I asked people who question the model to please provide an academic source for their statement. As I said, I've looked for a raging debate about the birth/death model and have only found it among anonymous sources on the internet. When I see it among academic/professional peers then I'll take notice.
The U3 U6 statement is far more valid and has more teeth. Using a broader unemployment metric that includes unemployed for economic reasons etc... makes far more sense. Or at least, the two should be reported together.
Bonddad, the B/D reference was from 2008, not this year. At that time the data that I presented was taken from the BLS. It showed the discrepancy between the reported data, declining homebuilding, layoffs, rising unemployment claims and the statement that 45,000 construction jobs had been created that month. Not an academic assessment but sometimes you've just got to use your head for something other than a hat rack.
Thanks as always and I don't agree with those dart throwers at HP and I did make that known on one article.
OP --
I understand your point. Still, the point is there is a tremendous amount of "the government data is flawed" stuff going on in the "tubes". But none of it is backed-up by or written by anyone with the qualifications to actually say it's bad.
For me, the point is it feels like there has been a slow almost unnoticeable drift into "black helicopter" territory in the "tubes." Anyone can say anything without backing it up with data (and no, I don't think you do this at all). But there seems to be so much of it that it's just time to say, "what are your qualifications to actually state that?" What I find really interesting is so far, no one has been able to point to some raging academic debate.
Very informative.you are doing a wonderful job. thank you.
Post a Comment