Wow, those are some bad GDP numbers

  • Share
  • Read Later

I just wrote a quick piece for on this morning’s GDP release. The conclusion, for those who’d prefer to skip all the numbers and stuff:

Unless it ends suddenly tomorrow, all signs are that this will stack up as the worst economic downturn in the U.S. since the Great Depression. But if government spending is able to begin stabilizing the economy sometime in the second half of this year, the scale of the decline in indicators ranging from GDP to employment to consumer spending will be much closer to that of the harsh recessions of 1981-1982 and 1974-1975 than the near-complete economic collapse of the 1930s. Will the stimulus work? That’s the $770 billion question the new GDP numbers really don’t answer.

Also, after I wrote the piece, I got an e-mail from former TIME economics correspondent Bernie Baumohl, who now has his own forecasting firm. He addresses the question of why the downward revision in GDP was so big:

One area was in miscalculating the change in inventories.  In the initial report, the government believed the change in business inventories increased by a $6.2 billion annual rate. That turned out to be flat wrong. As new data arrived, we see that inventories instead plummeted by a $19.9 billion rate. A second miscalculation was consumer spending.  The advance report noted a 3.5% decline, while today’s release revised it to minus 4.3%.  Net exports were also problematic for forecasters. Last month, foreign trade was viewed as adding to GDP growth (by 0.9 percentage points). Not so. The revised report said that trade had so deteriorated, foreign trade ended up subtracting from growth (by 0.46 percentage point.)

Basically, the issue seems to be that the economy made a really dramatic shift in the fourth quarter, and because the BEA’s early GDP estimates rely on a lot of numbers extrapolated from previous quarters, they missed the severity of the shift. Here’s hoping that will happen in the opposite direction two or three quarters from now.

Finally, Barbara pointed me to a very interesting passage from the new (to the U.S., at least; it was published in the U.K. in 2007) book The Numbers Game, by Michael Blastland and Andrew Dilnot, on GDP estimation biases in better times and how they differ from country to country:

The problem is that the fastest-growing parts of the economy are often new, comprising businesses that simply weren’t there before. If we counted every cent of economic activity as it took place, we’d know all about them, but we don’t. Not until they file their returns to the Internal Revenue Service do we get a proper sense of where to find the latest boom. So in order to measure economic growth, what do we put in our sample of businesses? Businesses that we already know about, what else?

Oddly enough, this does not lead to an underestimate of growth, since the problem is known and understood. The figures are adjusted, with am estimate of growth in new areas. However, that estimate is usually too high, and initial reports of GDP growth in the United States are almost always revised down.

In the UK, by contrast, this new growth is simply ignored until the tax returns come in to tell us what it was. … This has led to an initial underreporting of GDP growth in the UK typically by about half a percentage point. When growth moves along at about 2.5 percent per year, that is a big error.

In consequence, for the last ten years the UK has believed itself underperforming compared with the United States, when in fact it might have been doing every bit as well. By the time more accurate figures come out, of course, public attention has moved on, thinking the United States a hare and the UK a sloth.