Statistical Games


It’s fair to say that among those of us buried deep in the paranoid depths of the alt-finance community, a certain suspicion exists when it comes to government economic statistics. I don’t think it’s too much of a stretch to claim that among the broader population, cynicism about such stats is also increasing. We have after all seen a senior policeman testify in parliament to the corruption of official crime stats, just as rigorous studies have shown grade inflation in school exam results. Why should economics be immune from this malign trend?

Still: suspicion is one thing; hard evidence another. Given the convoluted maths and assumptions built into economic stats, it’s hard for lay observers to point to smoking guns showing clear manipulation. But just occasionally we get insider commentary offering pretty good confirmation; in this instance, Mike Bryan, vice president and senior economist in the Atlanta Fed’s research department:


I remember a morning in 1991 at a meeting of the Federal Reserve Bank of Cleveland’s board of directors. I was welcomed to the lectern with, “Now it’s time to see what Mike is going to throw out of the CPI this month.”


His article on the subject is worth reading.


Then there is the convenient timing behind some methodological changes. Recently it was announced that some European governments would soon be including criminal activities such as drug dealing and prostitution in their GDP figures. For years – nay decades – this wasn’t deemed necessary, back when these governments weren’t running perpetual deficits for as far as the eye could see, and weren’t faced with the grisly combination of rapidly aging populations and unaffordable welfare schemes. But now, faced with these problems, they suddenly decide that they’ve been under-counting GDP all along. Nice coincidence.


The most obvious objection to including drugs and whoring in stats is that there’s even more guesswork involved than with measuring legal activities. Pimps and dealers aren’t always that upfront about reporting their transactions to Inland Revenue. They don’t pay tax on their work, so the government isn’t gaining from their business. So why does it make sense to include such ill-gotten gains when measuring say, the government’s deficit as a percentage of GDP?


If the government were planning to legalise these activities then a good case could be made for these adjustments. But as it is, it looks like another dubious attempt by the state to paint a rosier picture of reality than is actually the case.


Last year the US miraculously increased its GDP by $500bn – equivalent to around a 3% increase in the size of the economy – thanks mainly to redefining R&D spending on “intangibles” like music and films as “investments”. Peter Schiff offered a good critique of this:


“In essence, the new methodology is an exercise in double accounting. For instance, suppose a company employs an accountant who works in the sales department, who is then transferred to the R&D department at the same salary. He still counts beans but now his salary will be billed to the R&D budget rather than sales. In the old methodology, the accountant’s impact on GDP would come only from the personal consumption that his salary allows. Going forward, he will add to GDP in two ways: from his personal consumption and his salary’s addition to his company’s R&D budget. The same formula would apply to a trucker who switches from a freight company to a movie production company (for the same salary). If he moves refrigerators, he only adds to GDP through his personal spending, but if he hauls movie lights, his contribution to GDP is doubled. It makes no difference if the movie bombs.”


This isn’t quite as blatant as the Nigerian government’s recent changes to its GDP computation methods, which resulted in a 90% increase (as a result Nigeria is now statistically the biggest African economy), but still. Accounting bodies can spend years debating changes to GAAP and IFRS standards that affect private companies – and rightly so: consistent accounting standards are a basic, vital requirement for any functioning economy. The same consistency in GDP standards – both across time and across countries – seems lacking, perhaps because whereas GAAP and IFRS are administered to private companies by outsiders, GDP amounts to governments marking their own homework. It’s not surprising that when the going gets tough, they find ever-more ingenious methods of giving themselves “A” grades.


Sometimes the official numbers involve such large adjustments that no belief in subterfuge is necessary in order to call the process into question. Take the Q1 GDP print for the US economy. Initially this was reported as 0.1% in the advanced estimate released in May. Then it was marked down to -1.0% in the second revision, and finally -2.9% in the latest revision released on June 25. Perhaps more changes are to come (see the chart below for the same period in 2008) by which time of course the media will be reporting the latest initial GDP estimates, with little-to-no attention given to old data adjustments.


Q1 2008 GDP


We live in a strange world where the government can report misleading, substantially inaccurate growth data that appears to be based on a fair amount of wishful thinking and questionable methodology – and which is then taken at face value by media, so influencing countless business and investment decisions. As The Return of the Great Depression author Vox Day notes, back in 2009 average GDP revisions were larger than the delta that distinguished growth from recessions. Now the revisions are nearly three-times larger.


Everyone has heard of the butterfly effect. Individuals would surely have made different investment decisions had for example, Q1 2008 GDP been reported as negative from the start, rather than 13 months down the line. All of which could have resulted in substantial difference in capital market prices as opposed to what actually transpired.

TLB recommends you visit Gresham Law

for more great articles and pertinent information.

See original article here


It’s time to privatise national statistics.

Be the first to comment

Leave a Reply

Your email address will not be published.