GDP Report is Good News; Deflator Critics are Wrong

Some people (mentioned here, for example) are denigrating the GDP report, which appears to show a 1.9% rate of real growth, on the grounds that the deflator – used to convert from nominal GDP to real GDP – is artificially low because it doesn’t include imports (which have become more expensive, due primarily to changes in the price of one dark, viscous liquid that is included in their number). As far as I can tell, these people are confusing production with terms of trade. It’s true that the “average American” is worse off when the terms of trade deteriorate, but that isn’t what GDP is supposed to measure. And frankly, it’s not what I, as a macroeconomist, care about.

If it’s a really big deal to you that people who have jobs can’t (on average) buy quite as much with their paychecks as they could 3 months ago, then go ahead and criticize the GDP report (but be careful how you phrase your criticism). Personally, I care a lot more about the people that don’t have jobs, or potentially won’t have jobs in the future. The biggest reason that I care about the GDP report is that more production means more demand for labor. (I’m speaking in relative terms, of course. In absolute terms, a 1.9% growth rate is probably not enough even to keep employment stable, let alone absorb normal labor force growth – but this has nothing to do with the deflator.) Production increased by 1.9% in the second quarter. As far as the implication for labor demand, it doesn’t make a damn bit of difference what you could or couldn’t buy with that extra production.

The second big reason I care about the GDP report is what it suggests about trends in productivity. What happens in one particular quarter isn’t, in itself, of great consequence, but it does tend to make us revise our estimates of what will happen in the future. If productivity is growing quickly in the present, then we can be more confident about its future growth. I haven’t tried to do the calculations, but since employment fell significantly in the second quarter, productivity must have risen by considerably more than 1.9%, and that sounds like good news to me.

Again, the terms of trade are irrelevant, unless you think you can extrapolate changes in the terms of trade the same way we extrapolate productivity. Does the fact that oil prices rose quickly in the second quarter make you think they will rise more quickly in the future? Maybe according to some theories, but my theory is based on the observation that oil demand is much more elastic in the long run than in the short run. If prices rise in the short run, while demand is elastic in the long run, that means future demand is likely to decline and push prices down rather than up. Obviously that’s a limited view and an oversimplification, but you’re going to have to do some pretty fancy theoretical gymnastics to get the result that rates of oil price growth are persistent over long periods of time.

I think the 1.9% growth number is the right number to look at, and it’s better than I expected. Apparently, it is not quite as good as the consensus expected (2.3%, according to Briefing.com, as reported by Yahoo.)

But the other number you might want to look at is inventory investment, which (as James Hamilton points out in the article I linked in the first paragraph) was severely negative in the second quarter. Inventory investment is not something that can be extrapolated. If anything, it is more likely to correct after an unusual move because the move was often the result of a mismatch between planned production and actual sales. (For example, in the second quarter, the drop in inventories may have been the result of having better than expected sales, which would lead us to expect positive inventory investment in the third quarter.) Once you account for declining inventories, growth of final sales was probably enough to imply a positive growth trend for employment, once all the inventory and cyclical labor usage issues wash out. That sounds like good news to me.