« Security Funds for WY- but Not Lady Liberty | Main | Unca Max Tells Thanksgiving Story »

November 27, 2003

Is Growth Real?

Now, the fact that growth is way up shouldn't be a surprising. Between massive government debt and personal debt being cashed out in tax cuts and home equity loans, it would be a shock if there wasn't some boost to the GDP growth numbers. Whether it's sustainable is the big question.

But with so little job growth accompanying the GDP numbers, it does make you ask what's going on.

And one answer is that the GDP statistics may be partially bogus.

Most folks recognize that you have to adjust GDP numbers for inflation, but they don't know that the government applies more than Consumer Price Index adjustments. One key game they play is trying to account for "quality" inflation (also called "hedonic pricing") as well as dollar inflation-- for example, the idea that the same dollar buys more computer power than it did a few years ago. But this has some perverse effects.

Check out this chart:

Read the text carefully-- despite the whole late 90s tech boom, business is not spending any more on tech equipment than in the mid-90s. The government just counts what's being bought as more valuable because of getting more power and quality for the same price. That's good news for quality of life, but it goes against the intuition that growth in GDP measures "more stuff" being made.

As the accompanying article notes:

That difference between the so-called real and nominal numbers clears up a mystery of the past few years. Even as the government was saying that real tech spending was rising, manufacturers were moaning about falling sales.

Savvy researchers wouldn't dream of using the adjusted spending numbers to analyze company results. "You don't pay yourself in real GDP, and you can't take real GDP to lunch," Lonski said.

Moreover, tech companies themselves don't put much stock in the gross domestic product numbers.

"To my knowledge, performance improvements in processing power have never figured in Hewlett-Packard's analysis of market strength," said spokesman Brian Humphries. "We focus on units shipped in a given period."

Still, misleading interpretations of the data are sometimes fed to the public by uninformed commentators. Wall Street analysts and economics writers often cite adjusted data on technology spending, thereby exaggerating the sector's gains.

Let's be clear-- it's no conspiracy that the government uses such quality-adjustments in GDP data. Quality improvements are worthwhile and worth measuring.

But it does mean folks should be careful about using GDP growth as a pure proxy for the health of the economy, especially when you are worried about jobs.

Some more reading on the topic:

  • Hedonic Price Indexing for upbeat GDP forecasts - Hedonism?-- notes that hedonic pricing may work well on pure commodity goods like computer memory, but will likely be inaccurate for more subjectively feature-laden goods like cell phones. Also, cross-national comparisons of GDP growth with countries that don't use hedonic adjustments (such as Germany) are likely to be distorted.
  • Hedonic Indexing: Now We Can All Have Productivity Miracles!- argues that hedonic pricing overstates US productivity gains in cross-national comparisons.
  • The Bush Boomlet? - makes the argument that fixed business investment is actually still falling in dollar terms, if you don't make the hedonic adjustments. See this disturbing graph:

    Posted by Nathan at November 27, 2003 07:58 AM