U.S. Second-Quarter GDP Growth Revised Up Sharply

Note that although the 1.7% increase in consumer spending was better than the first quarter (0.9%) and was surely helped by economic stimulus payments, it was foreign trade rather than stimulus that was the key driver of growth. The stimulus payments were not responsible for the 13.2% increase in exports, or for the 7.6% decline in imports. The stimulus payments would have raised imports, not lowered them.

A GDP growth rate of 3.3% does not mesh well with most perceptions of the economy's performance. In part that reflects the fact that GDP is doing so much better than domestic spending. In effect, all of the growth in GDP is being diverted into exports and import substitution. The U.S. economy as a whole is now growing its spending more slowly than its production—or not growing spending at all—after years during which spending outpaced production.

It is also possible that GDP is not really doing as well as the published figures suggest. There are two major puzzles in the GDP figures, which do put question-marks over whether the economy could really have been doing as well as the 3.3% headline growth rate indicates.

1. The revised 3.3% growth rate puts second-quarter growth above most estimates of the economy's trend—a surprisingly good outcome for a quarter in which industrial production, employment, and hours worked were falling, and the unemployment rate was rising. It is puzzling that according to the GDP accounts, the growth of real GDP in the goods sector accelerated to 6.2% from 0.9% in the first quarter, while industrial production went in the opposite direction, contracting 3.2% in the second quarter after growing 0.4% in the first.
2. The Commerce Department is unable to account for all of the GDP growth on the income side of its account. GDP measured from the expenditure side and from the income side of the accounts should be the same, but there is always a "statistical discrepancy." In this case, there is a large and growing statistical discrepancy between GDP growth and income growth, since measured income growth is running well below measured GDP growth. Over the past four quarters, the statistical discrepancy has risen by $255 billion (from -$143.4 billion to positive $111.8 billion). That means that $255 billion of the $575 billion growth in nominal GDP over that period cannot be explained by measured income growth. As the data go through multiple revisions over the coming months and years, either the Commerce Department will "find" more income growth, or GDP growth will be revised down.

The mismatch between the GDP evidence on the one hand, and industrial production and income evidence on the other, implies a strong possibility that GDP growth has not been running as high as the published numbers suggest.

But one possible area of mis-measurement often cited by critics should be dismissed. The GDP price index rose only 1.2% at an annual rate in the second quarter. Since that rate is so much lower than even the most casual observation of price increases would suggest, it has been suggested that the GDP price index must be understated. And if that is true, then applying a much higher price increase to nominal GDP growth of 4.6% would deliver a much lower growth rate than 3.3%.

This line of reasoning is mistaken. The GDP price index measures the changes in the price of the goods and services produced by U.S.-located households and businesses. Imported oil, for example, is not produced in the United States, so is not covered in the GDP price index. The price index for gross domestic purchases, in contrast, covers all goods and services purchased by U.S.-located households and businesses, so does include imported oil. This price index rose by 4.2% in the second quarter, a much more rapid increase than the GDP price index.

It is quite possible for rapid increases in the price of imported goods such as oil to cause the GDP price index to decline. If higher import costs cannot be passed on to consumers (e.g., gasoline refining margins drop), the price of goods to the consumer may go up but the price to the U.S. producer (which covers only the refining margin) goes down. In this case, the squeezed margin shows up in a lower GDP price index.

The first estimate of profits for the second quarter, also published today, is consistent with the margin-squeeze story. Pre-tax profits from current production were down 7.0% year-on-year, their worst quarter since the third quarter of 2001. Like the headline GDP figure, profits would have been worse but for help from overseas. Profits earned at home were down 16.9%, while profits earned overseas were up 16.5%, boosted by robust foreign growth and favorable currency translation effects from the weak dollar.

The GDP report and the profits report both emphasize just how important foreign demand has been in limiting the extent of the slowdown in the U.S. economy. With the rest of the world now slowing and the dollar off its lows, the U.S. will be more reliant on domestic demand in coming quarters. Since consumer spending is slowing down and the credit crunch is tightening its grip, it is hard to foresee another quarter with such a robust GDP headline for some time.

We continue to expect a W-shaped pattern to this business cycle. The evidence for the third quarter so far suggests that consumer spending growth has come to a halt and may even have turned negative. We expect GDP growth to slow in the third quarter to around 1.5%, but then to dip to zero or below in the fourth as the credit crunch bites harder. This growth profile should keep the Fed on hold until 2009.