Tuesday, August 31, 2010

More Fallacies About The GDP Accounting Identity

A reader asked me to comment on this article by John Tamny about why "weak GDP Does not a weak economy make".

First of all, it should be noted that GDP is far from perfect. There are problems, relating to for example inevitably imperfect data collection (which is why there are so many revisions) and the fact that GDP may not reflect for example capital consumption or factor income movements.

But Tamny's particular critique is not related to that and makes little sense (Though as we shall we see he makes some good points about the causal effects of improts and trade deficits):

"In another account covering Friday's GDP revision, one newspaper noted that "Friday's GDP report showed a surge in imports, which grew at the fastest rate in 26 years, during the second quarter. Growth in imports far exceeded U.S. exports and wiped out more than three percentage points of U.S. growth in the quarter." If so, let's thank heavens for a number that was revised downward.

A higher calculation, if higher due to reduced imports, would logically signal a weaker economy and the reason why is basic: all consumption - and imports are consumption - is the result of production first. In the real world we trade products for products, and since there's no evidence of compassion on the part of global producers, the surge in imports to record levels points to a substantial increase in productivity stateside in order to pay for those imports."


But if you do not exclude imports, the mere act of demanding (purchasing) goods will constitute creation of wealth. Assume that Americans had all decided to quit working and forfeit all capital income. Then clearly they wouldn't have been producing something. But assuming they could find some foreigners willing to supply them goods they might still enjoy a high level of consumption. Would it still make sense to say that the value of their [non-existent] production would be high simply because they order a lot of imported goods? That clearly makes no sense.

Furthermore Tamny contradicts himself in the following paragraph by complaining about the Keynesian "elevation of demand", because not excluding imports would in fact mean the elevation of demand.

Still, while Tamny expresses himself in a misleading way in this context, he is still right when he points out that trade deficits don't causally reduce GDP. But his error is that he in his rightful attempt to refute that fallacy, he tries to deny the accounting identity truth that given a certain level of domestic demand, imports will reduce GDP.

The accounting identity is valid, but it may be misinterpreted. If information exists about a certain level of domestic demand, and we find out that foreign producers covered a higher proportion of that demand, then it is valid to revise down the estimate of the value of domestic production. But that shouldn't be misinterpreted as imports having caused the decline in the value of production.
It is just as consistent with the accounting identity Y=DD+NX (Where Y is GDP, DD is domestic demand and NX is exports minus imports) to assume that a decline in NX caused an increase in DD as it is to assume that the decline in NX a decrease in Y.

And indeed, given the fact that the mirror image of increased imports is an increase in capital inflows and given the fact that it is associated with increased benefits from deepening division of labor, it is likely that increased imports caused a higher value of real GDP.

So Tamny was right in rejecting the argument that higher imports causes a reduction in GDP. But he was wrong in questiong the GDP accounting identity. Instead he should have pointed out that the GDP accounting identity does not in fact imply that higher imports causes a lower GDP, as it might just as well be that it causes higher domestic demand.