Are government economic statistics accurate? John Williams at Shadow Statistics says they are not, because the government has changed its measurement criteria over the years. Mr. William's shows us what the government's statistics would have been had the criteria not changed. Does this mean the "shadow stats" are correct, and the government's current measurements are wrong?

I'm skeptical. . . of the government's stats, of the shadow stats, and of the entire concept of measuring something so complex as a national economy. Indeed, I'm even skeptical of the idea that it makes sense to speak of a national economy. We are so interwoven with the rest of world that it's no longer accurate to speak of an American product, or a Japanese product, or a Chinese product. And yet, the attempt is made to measure things in these terms.

Ted Koppel, during his current Discovery Channel series about China, gives us the example of a couch that has some parts made in the U.S., and some in China. The couch is assembled in the U.S. and exported to China. Is the couch a U.S. product, or a Chinese product? I believe this question makes no sense. It's a null question. It's like asking whether 2+2=5 or 6. Neither answer is correct, because the question itself is invalid.

We have similar problems when we speak of price inflation. If the price of oil goes up is it because of supply and demand, an expanding money supply, currency exchange rates, or some combination of these factors? The answer is almost certainly that it's a combination of these factors, but there's absolutely no way to measure the relative weight of each factor. Given this, can inflation measurements based on prices have any real validity, using either the government's current criteria, the shadow stats criteria, or any other criteria? I think not.

We constantly think of prices in terms of supply and demand, but the size of the money supply is also a factor. It may be possible to have some rough idea of how much the money supply has expanded, but this still tells us nothing about the contribution of monetary inflation to price inflation, except in combination with really strong knowledge of supply and demand. The supply of a product can probably be measured to a certain extent, but demand can only be measured in terms of a broad range of prices which cannot be known to any great degree. Prices change constantly, from moment to moment and place.

Then we have the concept of adjusting current prices to account for inflation in order to compare them with past prices. Is this even possible? I think not. My house was built in 1968. As of today it has single paned aluminum windows that were installed when it was built. As of next week it will have double paned, coated, argon filled windows framed in virgin vinyl (so that the windows will expand and contract at a uniform rate). Both the old windows and the new are called windows, but they are as similar to each other as a horse and buggy is to a Formula 1 race car. How could we possibly compare the price of the old windows to the new? They are not the same product. Indeed, the old windows don't even exist as a product anymore.

Does my skepticism lead to impotence? How are we to direct our personal finances in a world of unmeasurable complexity? I suggest that we must rely on first principles. As citizens we must oppose the Federal Reserve's power to expand the money supply for the same reason that we oppose counterfeiting. As savers and investors we must diversify our portfolios, because no one can predict the future performance of any sector of our unmeasurable economy.