The reliability of government figures

Argentina takes the easy and straightforward approach to reducing inflation:

Since 2007, when Guillermo Moreno, the secretary of internal trade, was sent into the statistics institute, INDEC, to tell its staff that their figures had better not show inflation shooting up, prices and the official record have parted ways. Private-sector economists and statistical offices of provincial governments show inflation two to three times higher than INDEC’s number (which only covers greater Buenos Aires). Unions, including those from the public sector, use these independent estimates when negotiating pay rises. Surveys by Torcuato di Tella University show inflation expectations running at 25-30%.
PriceStats, a specialist provider of inflation rates which produces figures for 19 countries that are published by State Street, a financial services firm, puts the annual rate at 24.4% and cumulative inflation since the beginning of 2007 at 137%. INDEC says that the current rate is only 9.7%, and that prices have gone up a mere 44% over that period (see chart).

INDEC seems to arrive at its figures by a pick-and-mix process of tweaking, sophistry and sheer invention. Graciela Bevacqua, the professional statistician responsible for the consumer-price index (CPI) until Mr Moreno forced her out, says that he tried to get her to omit decimal points, not round them. That sounds minor—until you calculate that a 1% monthly inflation rate works out at an annual 12.7%, whereas 1.9% monthly compounds to 25.3%.

Threatening letters sent by the government to independent economists also shed light on INDEC’s methods. One was told that since the cost of domestic service was “a wage, not a price”, he should not have included it in his CPI calculations. “They have put a lot of effort and lawyers into such arguments,” he says.

What is remarkable is that one can quite reasonably rationalize this exercise in statistical fiction under the economic doctrine of “rational expectations”, which is the insane neoclassical idea that it’s not an increase in the money supply or debt outstanding or even a change in the level of prices that matters, but rather the expectation of future price changes on the part of consumers. So, if the government simply adjusts those expectations downward by lying about the current rate of inflation, then the subsequent behavior of consumers will cause prices to fall in line with those expectations regardless of how much the government ends up actually printing or borrowing.

Wikipedia reports this interesting fact about the concept: “Chicago economists applied rational expectations to other areas in economics like finance, which produced the influential efficient market hypothesis.” The EMH, it should be needless to say, has turned out to be a complete and utter flop, as demonstrated by Robert Prechter and numerous others using a wide variety of means.

What I find rather amusing about this article from The Economist is the clear implication that it is only an Argentine tactic and other government agencies, such as the Bureau of Labor Statistics and the Bureau of Economic Analysis in the United States, have not been playing exactly the same game for decades. Of course, as those who have read RGD will know, U.S. statisticians have been continually devising new and “improved” versions of CPI, each of which purports to show “true” inflation while methodically lopping off various sectors of the economy that are showing price changes such as those “volatile food and energy” sectors.

Consider this. The entire rise-and-fall in housing prices have had no effect whatsoever on U.S. CPI despite the fact that according to the latest statistics, $69.5 billion new home sales and $756.7 billion in existing home sales take place every year, representing around 5.5% of the $15 trillion in transactions that make up reported U.S. GDP.