Lying With Statistics, Inflation Version

Barry Ritholtz has a good post today on how the inflation stats are being presented in a particularly misleading fashion. We already have the proud history of CPI being redefined in the 1990s to lower the published inflation rate. As an economist buddy explained:

There was a big fuss over the CPI some years ago. It’s a charged issue because the CPI determines the social security escalator. Without making your eyes glaze over, the issues fall into 2 types: basket composition and quality adjustments. Should the CPI measure the prices of an historically fixed basket of stuff, even if no one buys it, or should it change to reflect what people actually buy? Since people shy away from high priced goods and look for cheaper alternatives, over time the flexible basket will rise a little more slowly than the fixed. Quality adjustment is the more vexed of the two issues. It’s clear on things like computers, where crunch power can be defined and priced by the unit, but it’s less clear in things like automobiles, where so called improvements come in the form of pointless gadgets. Bottom line: the BLS made a number of adjustments, which probably reduced the measured inflation rate by about a half percent.

Ritholtz has a different focus, which is the producer price index: the announcements generally focus on core inflation, when “core” excludes food and gas because they are volatile. They happen to be significant cost categories for both households and businesses so excluding them renders the relationship between the government data and real world experience rather dubious.

From Ritholtz’s “Inflation Confined to Rest-of-the-World, Avoiding U.S.:”

The United States has this fascinating relationship with inflation — there is this school of thought that if we disrespect it, belittle it, then perhaps it will go away.

That is not, unfortunately, the case in the real world. I guess the thought process is that if we collectively pretend there is no inflation, the Fed won’t have to raise rates.

Let’s look at last week’s PPI as an example. We saw a hefty increase in component prices in March. There was a big 4.1% rise in energy, with gasoline up +8.7%. Food “only” saw a 1.7% rise in prices. If we ignore these two life necessities (Inflation ex-inflation), the core was unchanged. A 1.2% decline in light motor truck prices mitigated the increases everywhere else.

Here’s the good news: if you forego consuming food and energy, and stick to buying light trucks, you have no inflation worries.

Elsewhere in the world, where real data matters and rhetorial games are not a substitute for real economic analysis, they do have inflation worries. The UK’s version of inflation (CPI is out Tuesday in the US), showed sharp price rises:

The Office for National Statistics said output prices rose 0.6% on the month in March to stand 2.7% higher than a year earlier, up sharply from 2.2% the month before, a figure analysts had expected to be repeated this time.

The figure spooked financial markets and the pound rose to its highest in three months against the dollar of above $1.991. Markets broadly expect Bank of England’s monetary policy committee to raise interest rates again next month, taking them to 5.5%.

Analysts said the figures suggested firms were managing to put up their prices in the wake of a year of steep rises in energy costs. Core output price inflation, which excludes volatile food and energy prices, also picked up more than expected to an annual 2.9%, the fastest rate of growth since June 2006.

Note that Core Inflation prices are almost an afterthought, not the main focus of the inflation report.

We noted earlier in the year that The Office for National Statistics allows consumers to calcualte their own personal inflation rate, based on a variety of factors. I’d love to see the BLS try that trick . . .

Print Friendly, PDF & Email

One comment

  1. Richard

    As several posts in the comments section of the original blog entry pointed out, the US is hardly alone in fudging inflation figures. The CPI in the UK is a joke. It excludes house prices all together, but does include rent in calculating inflation. In a country where 70% of households are owner occupiers, the UK CPI relies on 30% of the population to measure housing costs. When you look more closely into those numbers, though, you realize how ludicrous the statistics are. 2/3 of renters in the UK live in state provided or subsidized housing, paying a fraction of market rents, so right off the bat, the number has very little to do with what’s going on in the larger economy. Further, about 4% of the population is reported to live rent free (how does that work?) so about 12% of the sample is reported to have no housing cost at all. Given the set up of the CPI, is it any wonder that the Bank of England has created a monster housing price bubble?

Comments are closed.