What a mess. As Kevin Phillips reports in a Harper’s article, “Numbers Racket: Why the Economy is Worse Than We Know” (full text here) government agencies have made scads of tiny, incremental changes to economic statistics over the years that, in the aggregate, have completely changed our basic measures of economic health and wellbeing. What’s especially troubling: the changes have been consistently biased to make the economic outlook seem rosier than it otherwise would. Says Phillips:
If Washington’s harping on weapons of mass destruction was essential to buoy public support for the invasion of Iraq, the use of deceptive statistics has played its own vital role in convincing many Americans that the U.S. economy is stronger, fairer, more productive, more dominant, and richer with opportunity than it actually is.
Look, for example, at inflation. The nifty chart below is from ShadowStats.com, a website devoted to tracking this sort of statistical tomfoolery. The blue line is the inflation rate as it was calculated before President Clinton came into office; the orange line is the current inflation rate; and the yellow line at the bottom is an experimental measure that—if made official—would ratchet down reported inflation rates yet another notch.
Unemployment figures are similar—the unemployment rate would be higher if we excluded the military from the workforce, as we used to, or if we added back in people who haven’t had a job in over a year, but would still like to find one. Using the old methods, today’s ‘misery index’—the sum of the unemployment and inflation rates—would look a lot more miserable than the official figures currently show.
But here’s the rub: statistical changes in the way that inflation, unemployment, and so forth are calculated aren’t simply tomfoolery. Many of the changes are actually pretty reasonable—or, arguably, even necessary.
Find this article interesting? Support more research like this with a gift!
The problem is simple: times change, and to stay relevant, economic statistics have to change with them.
Take, for example, a rapidly mutating sector like consumer electronics. If the cost of the average cell phone goes up from year to year, is that because underlying prices went up, or because consumers were willing to pay a bit more for newer, better, cooler phones? Possibly both, but it can be hard to disentangle the two. So statisticians have developed a technique called “hedonic adjustment” which factors in consumers’ willingness to pay for higher quality products when calculating inflation rates. (Search this page for the word “hedonic” to explanations of how hedonic adjustments are calculated for various classes of consumer goods.)
Obviously, hedonic adjustment can be taken too far; there’s no good reason to consider the price of cell phones when setting cost-of-living adjustments for the very elderly, or for folks who are too poor to afford a cell phone.
But over the long haul, quite a few things resemble cell phones: price increases have accompanied increases in quality, and it’s hard to tease the two apart. Today’s low-end housing ain’t great, for example, but it’s a major step up from the crowded slums of a century ago, or even the 1960, when one in five homes lacked indoor plumbing. People may pay more for housing, but they get more too.
What all this means to me—the problem may not be that statistics about inflation and unemployment are biased. It’s that we’re asking too much of the statistics themselves. We want numbers that provide a a reliable gauge of short-term blips in prices and job prospects, but also of long-term changes in how we live. We want measures that ensure consistency over the long haul, but that also adjust to our changing preferences and standards of living. And we want figures that are valid for both the poor and the middle class, who have very different patterns of spending, and different perceptions of their financial needs.
Statistics like that may be a chimera: no single economic measure that can do it all. In fact, the most fundamental measures of how we’re really doing—measures of the strength of community ties, or people’s happiness—aren’t really in the realm of economics at all. It should be obvious, but it’s not: the search of a reliable “misery index,” or its converse, probably shouldn’t start and end with measures of what we can buy.
Regardless, the ever-rosier biases of the massive government statistical apparatus is something to keep in mind. So the next time you hear that inflation isn’t all that bad, even though the price of food and energy is going through the roof, you’ll know what’s going on.