GDP in the Rear-View Mirror

appears smaller than it did before --  the BEA's "third estimate" of real GDP growth came in at -2.9% annual rate.  That's really bad, and a big revision from the "advance estimate" in April of 0.1% growth, and the "second estimate" in May of -1%.

One of the things I emphasize to my students are the limitations of GDP statistics.  One of the difficulties in using them is that they are subject to substantial revisions, that come in with considerable lags.  Policymakers - and anyone else trying to judge the state of the economy - are looking at noisy, backward-looking data. 

Here it is, just past the summer solstice, that we learn that GDP last winter (Jan. - Mar.) was dropping at its fastest rate since 2009 (the first quarter of 2011 is only one other quarter since the recession with declining GDP).  The rate of decline in the 1st quarter was worse than either of the two quarters with negative growth in the 2001 recession.

Although there are usually some changes, this particular revision was unusually large - the change from the initial to the third estimate was the largest since the BEA began releasing estimates this way in the mid-1980's.

The prevailing theory on why the first-quarter was so bad appears to be that it was mainly due to unusually severe weather; although the data are "seasonally adjusted" to account for the fact that some types of economic activity normally are lower in January and February - this winter may have been worse than most.

As Neil Irwin and CEA Chair Jason Furman both note, other indicators - like employment - looked ok during the same period.  Payroll growth averaged 190,000 during the first three months of the year.  That, as Justin Wolfers explains, means a large deviation from the historic relationship between unemployment and output growth known as "Okun's Law".  It also implies a big drop in productivity as we measure it.