“There are three kinds of lies: lies, damned lies, and statistics.” This well known pun, which US author Mark Twain attributed to the British Prime Minister Benjamin Disraeli, has again been illustrated by the recent debate regarding US inflation and how to measure it.
In May the headline inflation rate in the US posted a whopping 3.6%, its highest reading since October 2008. More worrisome still is the fact that this rate has almost tripled in the last six months, despite which the Federal Reserve kept its cool, continuing to argue that headline inflation doesn’t matter. Focus should remain on the core rate, which excludes volatile components like oil and food prices.
A new polemic now enters the debate. The Dow Jones news agency reported that changes to inflation measurement are part of the ongoing negotiations between Republicans and Democrats regarding the US debt: “Lawmakers are considering changing how the Consumer Price Index is calculated, a move that could save perhaps USD 220 billion and represent significant progress in the ongoing federal debt ceiling and deficit reduction talks.”
Interestingly enough, this might be a proposal the two entrenched camps could agree on. Both want to do something about the deficit. But while the Republicans insist on reducing government expenditures and not increasing taxes, the Democrats want to increase taxes and keep government expenditures where they are.
Since “inflation” is generally thought “bad” by the broader public, governments traditionally try to show inflation to be as low as possible. Each revision of the official definition of inflation measured with the consumer price index (CPI) in the US over the last half century – not only there but also elsewhere – has resulted in lower official inflation rates – and as a mirror image, higher real growth rates in the economy.
The concept of a “core rate” was introduced in the early 1970s under US President Richard Nixon, when energy and food prices started to surge. Then in the early 1990s several debates led to an in-depth revision of the US CPI.
The main adjustment was to replace the fixed weights of goods and services in the representative consumer basket used to measure inflation with variable weights. This took into account changes in tastes and habits, e.g., CD players replacing tape recorders. Technological progress was another reason. A notable example is the iPad 2, which is twice as fast and half as thin as the iPad 1, but costs the same price – a clear technical “deflation” of this particular product. Now, the US CPI undergoes a revision of the weights every two years along these lines.
Also discussed was whether price-induced substitutions shouldn’t be part of the weighting scheme. If the price of potatoes increases, people tend to eat more rice, hence the weight of rice in the consumer basket should increase while that for potatoes should decrease. So far this factor hasn’t been made explicit in the calculation of the US CPI. However, it has been implicitly taken into account by using geometrical instead of arithmetical means in the CPI computation. This technicality will understate large price increases and overstate price drops.
The revision now under discussion makes these substitution effects more explicit. If agreed upon, official inflation will be once again lowered. This helps both to reduce government expenditures while increasing revenues.
Expenditures are reduced because many incomes derived from social security programs are indexed to the official inflation rate. Revenues are increased because inflation is nothing else than a wealth redistribution from creditors to debtors, or put differently, a taxation on savings. Understating real inflation by fumbling with its measurement allows the creation of more of it while mitigating some of the debt problems facing the US government.
In this context, the old German saying that “you shouldn’t trust any statistic you haven’t falsified yourself,” may be the best advice one can offer.
No comments:
Post a Comment