Fudging the numbers, scripting the happy talk

Available today at this link is an interview with John Williams, a Republican kind of guy and an economist who has been tracking how government statistics have diverged from reality over time. His Shadow Government Statistics project displays a much different economy than the official figures, an economy more like the one you and I experience. The cherry picking of numbers I razzed EPI about last week is a sneeze to the finagling of officials' hurricane.

Inflation, for example, is systematically understated, and has been since the early 1990s. Williams estimates about 2.7% should be added to official inflation figures to compensate for dubious statistical tricks. The consumer price index (CPI) should be at 7%, rather than the official 4%. Particularly annoying to Williams is the geometric weighting of the CPI, a scheme devised by Alan Greenspan and George I's chief economist Michael Boskin, and later incorporated into official numbers under Bill Clinton.

Inflation was previously measured by checking the price of a fixed "basket" of goods from one period to the next, a collection of goods such as might be purchased by an average consumer. Under the Greenspan/Boskin scheme, the substitution effect was incorporated. The idea is when steak gets too expensive, people will substitute hamburger, so hamburger instead of steak should be in the basket. The price of the basket with hamburger is, of course, not so expensive as the basket with steak, so inflation does not rise so much. In other words, the CPI has become the price rise of a deteriorating standard of living.

Other clever means of suppressing inflation include "hedonics," from the same root as hedonism, which allows products which have improved in quality – say a washer with electronic controls as opposed to twist dials – to be valued at a lower price. The theory is they are not perfectly comparable.

Williams dismisses out of hand any claim to legitimacy such machinations have, but he says the exercise was not primarily to amuse voters around election time, though this was definitely a welcome adjunct. No, the real intent was to reduce the payments and obligation sunder Social Security, since these are tied to the CPI. And it has done that. Williams suggests that if all the tricks and their cumulative effects were reversed, payments to Social Security recipients would be 43% higher.

Equally insidious, however, is the tendency of this operation to overstate economic growth. The most popular economic measurement is the growth of real GDP, real meaning inflation-adjusted. If inflation is understated, real GDP is overstated. (Maybe you just count the "for rent" and "for sale" signs.)

By Williams calculations, the economy is on the verge of the second part of a double dip recession, and may already be in contraction. Inflation is 7%. And unemployment – as measured by the formula of the Depression years – is running at 12%.

In some parts, maybe he is over the top. Using accrual accounting for Social Security is not particularly apt. But Williams adjustments do make the numbers and the experience of the average working person come into the same picture.