Regarding official govt stats, I recently voiced a few opinions on the employment numbers, ( http://www.boomantribune.com/story/2012/10/5/92611/9177 )and had a former stat gatherer / # cruncher from the census dept. who posts here tell me I was an idiot and to basically shut up since I did not know what I was talking about.
In the interest of people knowing what they are talking about, I add this post as food for thought. I know this is a progressive community, but I also know Boo prides himself in being informed, so I present this as a contribution to broaden viewpoints and discussion.
A partially trigger for many comments was Jack Welch’s comments, with many all over saying he was nutzo. (Welch Replied to those criticisms… http://finance.yahoo.com/news/jack-welch-jobs-report-032200419.html. )
I also find it ironic also that a former Clinton aide was quick to dismiss Welch as a conspiracy nut for his comments, when the same aid impugned Bush’s authority for cooking the books for political benefit. But I digress…
Anyway… Boo subsequently put up a front page post ( http://www.boomantribune.com/story/2012/10/5/184540/363) saying the numbers are not important -fake or not- from a political standpoint since people vote reality (Agree!), although the same ex census dude jumped in defense of the numbers’ reliability with a sweeping overly generalized statement about those who think he and his former colleagues “cook the books”: they are “f’n idiots”. All on Friday.
Meanwhile, Secretary Hilda Solis criticized the conspiracy theories on Friday, as well. “This is a methodology that’s been used for decades. And it is insulting when you hear people just cavalierly say that somehow we’re manipulating numbers,” she said on CNN.
That’s correct, yet that’s not entirely accurate. The methods used are constantly shifting and adjusted to create that accuracy — an oxymoron of sorts. Yes, they’ve done that sort of thing for decades — with (mostly) the best of intents I’m sure — but that’s not to say that they are bomb-proof reliable, and that flexibility allows for the plausibility of occasional nudges — which deeper analysis suggests sometimes happens, unless purely coincidental.
Anyway… Here are a couple of links that detail what these numbers are from John Williams, whose job is to provide businesses with reliable econometric measures for their business planning, and he’s paid commensurately for his accuracy. I’ve followed John Williams for years, was a subscriber of his data analysis services, and even interviewed him when I used to do my own net-radio show.
Just below is a link to a written interview he did with Kate Welling of Weeden Asset Management (Part of Luethold) back in 1996 or so. (Again, discerning reality from the tea leaves of data is what that firm is paid to do. If they fail, they get fired — unlike a TBTF bank.).
http://www.shadowstats.com/pdf/779-626538446.pdf
While written in 2006, it is valuable because it discusses William’s history in the business, his experience of having his own analysis go bad, and his discovery that what had changed were the methods used by the BLS, etc., to build / calculate the final data he was digesting. When he dug into the numbers and began reconstituting them with the raw data by the old methods used by the BLS, etc., his models and forecasts became accurate again. With that, he has made a specialty of tracking the the way govt. provided stats are created, following the changes that have been made to those methods over the years, and crunching the raw data to replicate the old ways when possible. His evidence is not conspiratorial — it is disclosed and verifiable.
Another link to a more timely interview is from last Friday (10/5/12) is below — a stream / download interview he just gave on currently popular South Cal financial radio program available to a national audience over the net in which he discusses the nature of government provided statistics and a change in this reported data in the way seasonal adjustments are applied. While he finds the jobs # a bit surprising, he’s frustrated by the fact that the BLS does not immediately supply raw data to compare with what they’ve reported (Unlike those who report, for example, GDP which has lots of stat fixes), and therefore it is a bit of a black box in terms of verifiability.
http://www.financialsensenewshour.com/broadcast/fsn2012-1006-3.mp3
Specifically, pay attention to his discussion on the quality of what’s published (very little transparency) vs. what is eventually published and revised long after the number is no longer on the news cycle radar. What he can see in the seasonal adjustment data is that 1st half Job growth is being moved into the 2nd half, while the 1st half numbers were not revised downward to compensate, in essence giving both periods the benefit of the same job growth if looked at statically today.
Additionally, consider how the U6 job number stayed flat, which adds up per his statistical crunching vs. the headline U3 number. That also jibes with other “flat / slightly off growth” inferences from other reported economic data.
Bottom line — He comments on the methods used, and the pros and cons of the numbers provided, be it transparency, lack thereof, and the history of manipulation from both parties. He says outright statistical manipulation is very rare, that data adjustments are imperfect, and that systematic “improvements” to how adjustments are made to raw numbers(under either party) always skew to making the economic environment better appear politically more palatable.
NOTE: He gives Kudos to the Obama admin for having numbers that consistently make sense / seem more plausible / less political that his predecessors. However, this recent number does not fit that bill, and he explains why.)
Lastly, here is a link to a summary / primer on the nature of broad economic data / numbers from Williams’ public website. It cuts more into how he got where he is and where the data is simply the old saying about statistics: they are what they are, and are only as good as your understanding of that.
Where Williams falls IMO: neither conspiracy nut nor gullible dupe. Sober. And I think folks here understanding why will be more informed and less manipulated by our leaders, as if we need to trust them any less.
http://www.shadowstats.com/article/primers_intro
**
All that said, my point is to keep these numbers in perspective. No doubt folks like Mr. Rude (and his former D & R colleagues) who grind through to produce quality output are likely honest people doing an honest days work. But that misses my point. This data is what it is. It’s not infallible and makes many compromises to do what it does, yet it is held as definitive by many in the investment/economic analysis biz and political communities. I deal with the former, and I think it clouds their judgement. The former run campaigns with these numbers, which IMO misleads the masses in different fashion. To Boo’s and William’s point, people on the street see through it – they know when things suck regardless of GDP or employment stats.
But others defend that data as if its sacrosanct and infallible, a mistake. And the data is often used for political purposes, to say things are improving during an election, for example. (Bush I tried this by adjusting GDP’s formulation, apparently…)
All said, I’d be interested in others’ genuine (non rude) views on this, specifically those who give this subject a full hearing so they are better equipped to make sense of the subject. I don’t expect anyone to change on a dime with this. I think it makes sense, though, to keep handy in your hip pocket as a reference the next time an R or a D tries to use them for political advantage. I would like the input from census guy number cruncher as to what his actual experience was in the pits of all of this.
Thanks
