Ed Dolan's Econ Blog

Why such a Large Downward Revision of Q1 GDP Growth?

Today’s GDP report from the Bureau of Economic Analysis revised the estimate for Q1 growth sharply downward, from 2.4 percent to 1.8 percent. As the following chart shows, the revision makes the rebound from the slowdown at the end of last year look less robust.

The early estimates for each quarter are based on incomplete data supplemented by extrapolations from historical trends, and revisions are common. The revision announced today was larger than usual, however. The average revision (without regard for sign) is 0.2 percentage points from the second estimate to the third. This time it was 0.6 points. The change from April’s advance estimate of 2.5 percent was even larger. How can we explain it?

One way is to look at the contributions to GDP growth of each sector of the economy. The following table compares the data from the advance estimate to those of the third estimate. It shows that the downward revision affected every part of the economy. Growth of consumer spending, which almost always makes the largest single contribution, was revised downward from 2.24 percentage points to 1.83 percentage points. The contribution from investment growth fell from 1.56 points to .96 points, with decreases in both fixed and inventory investment. Shrinkage of the government sector, which has had a negative impact on growth throughout the recovery, was more rapid than previously reported. Every sector of government—federal defense and nondefense, state and local—made a negative contribution to growth in Q1.

Net exports made a smaller negative contribution to growth than previously reported, -.09 percentage points rather than -0.5 points, but when we look at the details, we cannot really call that good news. One disappointing development was a revision of export growth from positive to negative. As the next chart shows, that meant that exports fell in Q1 for the second consecutive quarter. Up to Q4 2012, exports had been one of the strongest drivers of the recovery. The slowdown in exports is largely due to weaknesses in almost all major U.S. trading partners. Furthermore, the contribution of imports to GDP growth turned from negative to positive in Q1. That, too, is not entirely good news, when we consider that imports are entered into the national accounts with a negative sign. A positive number in the reported contribution to growth thus means that imports were decreasing, reflecting the general weakness of demand both for imported consumer goods and imported capital goods.

The less impressive growth report for Q1 2012 throws new uncertainty into the Fed’s recent hints that it will begin to taper off its program of large-scale security purchases later this year. That prospect was based, in part, on relatively good job market performance since January. However, weak GDP growth suggests that improvement of the labor market might not continue. Much will depend on next month’s job market data and at the advance estimate for Q2 GDP, both of which will be available by the time of the Fed’s next policy meeting, scheduled for the end of July. If those data are weak, the Fed is likely to put its planned tapering on hold.

7 Responses to “Why such a Large Downward Revision of Q1 GDP Growth?”

rjsigmundJune 26th, 2013 at 8:34 pm

the big jump in farm inventories mostly looks like a seasonally adjusted aberration (ie winter inventories normal vs fall inventories weaker due to the drought)

still, it's awful large & my ag correspondents have been unable to explain it adequately

TomJune 26th, 2013 at 9:19 pm

It was indeed a pretty steep downward revision of merchandise exports, which probably reflects actual data from the fairly grim world of trade that it has been out there lately versus early overoptimistic modeling. There was also an even bigger and very curious drop in the estimate of personal consumption of "other services".

windrivenJune 27th, 2013 at 11:42 am

The third chart troubled me most. Employment is still weak at home which suggests that domestic consumption will remain lukewarm and with weak exports where is GDP growth supposed to come from?

Further troubling is that simple sum Q3+Q4 2009 growth was 5.4%, Q3+Q4 2010 was 5.0%, Q3+Q4 2011 was 5.4% but Q3+Q4 2012 was only 3.5%. Now we have a soft start to 2013.

It is hard for me to see many factors pushing 2013 toward a strong finish.

Ed Dolan EdDolanJune 27th, 2013 at 3:32 pm

I agree with you that it is hard to see factors pushing this year toward a strong finish. Be careful about adding the growth rates for consecutive quarters, though. All of these figures are quarterly data stated at annual rates, e.g., if the economy grew all year at the Q1 rate of 1.8%, total growth for the year would be 1.8%. not 7.2%. Instead of adding quarterly rates, you should average them (make it a geometric mean if you want to be fussy about it).

windrivenJune 27th, 2013 at 4:25 pm

Yes, of course. I use the simple sum as a sort of shorthand that amplifies the changes and makes blackboard of the mind year-to-year comparisons easy. I use Q3+Q4 because that sum recognizes a lot of consumer spending for back to school and Christmas. But in truth, Q3+Q4 2012 was so stark that amplification isn't necessary to pull signal from noise.

Another disturbing development is the NSA rolling UE claims slope seems to be flattening out after three years of at least modest decline. Flatlining at 300k suggests troubling UE rate for the foreseeable future.

ThomasGrennesJune 28th, 2013 at 11:15 am

Revisions of quarterly and monthly data are frequent and sometimes large. Samuel Rines points out in today's Wall Street Journal ( that GDP for the last quarter of 2008 was initially reported as -3.8% (annualized), but in the sixth revison it was reported as -8.9%. Others have shown that the noise/information ratio is large for the early reports. Rines suggests labeling the early reports as "best guesses". Given the size of revisions, they probably get more attention than they deserve.

Ed Dolan EdDolanJune 28th, 2013 at 11:33 am

Q4 2013 is an example where even the sign of the advance estimate was wrong: -0.1 percent, later corrected to 0.4 percent. Years ago, the BEA used to release an even earlier "flash" estimate based on even less complete data, but it turned out to have so little useful information that they dropped it.