I see Paul Krugman has picked up on my friend Jared Bernstein's post asking where the Fed's 2.0 percent inflation target came from. He argues that it is a pretty much arbitrary compromise between the idea that the target should be zero (the dollar keeps its value constant forever) and the idea that we need some inflation to keep the economy operating smoothly and avoid the zero lower bound for interest rates. This is far too generous.

I recall reading the literature justifying the 2.0 percent inflation target last year when I was researching my book with Jared, Getting Back to Full Employment. I was shocked to see how weak the argument was. Not only is there not much justification for 2.0 percent, there is not much justification for any target. After all, most central banks have not had any explicit inflation target for most of their existence. These countries didn't get hyperinflation and their economies did fine. There isn't much evidence that an inflation target will prevent bad things from happening.

While as a rule stable prices are better than unstable prices, economists have been largely untroubled by unstable prices in other areas. Look at the real value of the dollar over the last three decades. Rises and falls of 20 percent over a two or three year period are not uncommon. This means that the price of large chunks of our economy fluctuate in large and unpredictable ways, apparently without negative effect, or at least not enough of a negative effect to concern the bulk of the economics profession. So if these big changes in relative prices of imports and exports  don't have much consequence, what's the big deal if inflation is 2.5 percent rather than the 2.0 percent we all are supposed to treasure? (If the dollar falls by 20 percent, as a first approximation, exporters are getting 20 percent more for their products, measured in dollars. They get the same amount of foreign currency, but it buys 20 percent more dollars.)

The other part of the 2.0 percent story that I really loved is that some economists argued for it by saying that it could be reconciled with the dream of zero inflation because of errors in measurement. Krugman briefly refers to this view, but doesn't explain the meaning of this claim. The argument here is that our price indices overstate the true rate of inflation and that the overstatement is on the order of 2.0 percentage points annually.

If you just read that and your hair didn't stand on end, read it again. Many of us have been writing on the stagnation of wages and income over the last three decades. (For example, Thomas Edsall had a good piece in the NYT earlier this week, although he gets the dates wrong.) If we have been overstating the true rate of inflation by 2.0 percentage points annually then wages and income have actually been rising very rapidly. Instead of near zero growth since 1980, real wages have risen by more than 96 percent. (Take 1.02 to the 34th power.)

There is no way around this implication. We know the rate of nominal wage growth. For example in the last year it has been around 2.2 percent. With a reported inflation rate of around 1.8 percent, this translates into 0.4 percent real wage growth (2.2 percent minus 1.8 percent equals 0.4 percent). But if inflation is overstated by 2.0 percentage points then real wages actually rose by 2.4 percent.

Not only does the measurement error mean that real wages have been rising very rapidly in the past, the implication is that they will continue to rise very rapidly in the future. Think of the Peter Peterson generational equity types who want us all to be upset about the Social Security and Medicare taxes our kids will be paying in 30 years. If the 2.0 percentage measurement error crew is right, then the Social Security trustees projections imply that the average real wage will be around 200 percent higher in three decades. And we are supposed to be concerned that our kids are paying a couple percentage points more in Social Security taxes when their wages are three times ours?

The implications of a measurement error of close to one percentage point are enormous. If the error is on the order of two percentage points then we have no clue what is going on in the economy. The fact that very serious economists, who enjoy great standing in the profession, could suggest a two percentage point measurement error with a straight face shows the lack of seriousness of the discipline. People should understand what is at stake here. It is big (and simple).