More on the CPI Posted by: Dale Franks
on Wednesday, March 22, 2006
I just want to interject a brief note here about Jon's observations on the CPI problem. The problem with the CPI that the Boskin Commission identified affects not only the future, but an unreliable CPI also should make us think twice about much of the debate over economic policy in other areas, too. The CPI is, in many ways, the 800-pound gorilla of economic statistics. It's how, for example the labor department calculates real wages.
After the Boskin Commission's findings were released, the New York Times asked economist Leonard Nakamura to provide an independent look at the CPI. Nakamura found that the CPI had overstated inflation by an increasingly large amount. Starting from an overage of 1.25% compounded annually in the 1970s, it increased to 2.75% by 1996.
By using a corrected CPI figure, Nakamura estimated that, rather than falling, real average hourly wages had risen by 25% from 1975-1996, GDP had grown by twice the official rate, and family real income rose by 19%.
Yet, in the debate one hears about incomes, and the policy proposals that flow from it, there is hardly any acknowledgement given to the fact that a faulty CPI invalidates many of the arguments that wages are under a tight squeeze.
If the CPI is off, determining real wages, real incomes, and real GDP growth is a practical impossibility. Nearly the whole structure of price, wage, and growth statistics are built on the assumption that the CPI is telling the truth about the level of inflation.
Moreover, while the Fed uses two other measures of inflation in addition to the CPI, the CPI and Producer Price Index (PPI) form the core statistics that the Fed use to make decisions on short-term interest rates. If the CPI overestimates inflation, that implies that the Fed has been making policy moves to raise rates, and indirectly cut back on economic growth, that were not justified by the actual rate of inflation. The FED, in short, jumped the gun to strangle growth, based on fears of inflation that didn't actually exist.
That, in turn, raises the obvious question of whether GDP growth over the last 25 years or so might have been even higher had the Fed not intervened in times like the "soft landing" they engineered in the mid 90s.
We may not have needed any kind of landing at all, which means that GDP growth was throttled for no good reason at all.
If the CPI is off, determining real wages, real incomes, and real GDP growth is a practical impossibility. Nearly the whole structure of price, wage, and growth statistics are built on the assumption that the CPI is telling the truth about the level of inflation.
I don’t think that’s really true. For example, many people consider "X as percent of GDP" (where X can be the federal budget, deficit, defense or health-care costs, personal earnings, ...) to be an important economic statistic, and it’s unaffected by CPI. Income-distribution statistics are likewise unaffected. True, a faulty CPI would have serious implications for growth statistics of all kinds, but there’d still be quite a bit left.
After the Boskin Commission’s findings were released, the New York Times asked economist Leonard Nakamura to provide an independent look at the CPI. Nakamura found that the CPI had overstated inflation by an increasingly large amount. Starting from an overage of 1.25% compounded annually in the 1970s, it increased to 2.75% by 1996.
Do you have a citation for this? This is sharply at odds with what the Boskin commission found and what later research has indicated. Iirc, the Boskin Commission concluded that inflation was overstated by 1.1 percentage points in 95-96 and by 1.3 percentage points in the years between 78-95. (tangentially, they noted that these discrepancies cannot be used going back indefinitely)
If the CPI overestimates inflation, that implies that the Fed has been making policy moves to raise rates, and indirectly cut back on economic growth, that were not justified by the actual rate of inflation.
The Fed is actually well aware of the CPI problem. Greenspan has addressed it in the past. I suspect they correct for this by examining more detailed, discreet measurements.
The validity of the CPI has been debated for years. A detailed critique would require too much of my time and be extremely boring. Suffice it to say that the man on the street experiences inflation/deflation from a completely different perspective. How much has my food bill, gas bill, rent, gone up last month? Why does my company posit 2.5% as the average of pay raises for this year?
Imagine trying to cook a 12 trillion pound 300 million cubic inch turkey by testing the temperature once a month.
It’s probably too much to ask, but we should do two things.
First, assign the smartest propellor heads going, to define some new measures. Not one or two, but less than five. Ask the consumers of the data what they need.
Second, put a stake in the idea of "cost of living increases". Make Congress propose and vote on annual increases for every federal program. No more of this "tied to the CPI", tied to raises given to the Post Office stuff. I want Congress to vote on whether or not my grandmother gets a raise in her SS payment and I want them to decide what that’s going to be. Every year.
The first is possible and likely will come to pass.
The second, given the lack of courage in Congress, never will.
The truth of the matter is that inflation can be perceived (inflation is the alteration of the supply of and/or the demand for money such that the perceived purchasing power declines) but not measured. Whatever the CPI purports to measure, it isn’t inflation.
