The Atlanta Fed's macroblog provides commentary and analysis on economic topics including monetary policy, macroeconomic developments, inflation, labor economics, and financial issues.

Authors for macroblog are Dave Altig, John Robertson, and other Atlanta Fed economists and researchers.

« Hard Landing? Still Waiting | Main | One Reason To Actually Worry About A Flat Yield Curve »

January 29, 2006

Fourth Quarter GDP: Read It And Weep

There was plenty to be said about Friday's advance release of 4th quarter GDP growth, and you can bet that it was said by someone. John Irons proposed that the news was "not very good."  SCSU Scholars concurred: It "ain't good."  The Skeptical Speculator described the outcome as "weak."  Andrew Samwick offered the phrase "lackluster."  William Polley weighed in with "disappointing." Brad DeLong noticed "lots of bad news." Jim Hamilton put the statistics in the "gloomy" category.  Kash called it a "terrible report."  BizzyBlog: "It stinks." And Barry Ritholtz thinks it is "amazing" that folks haven't figured out that -- this is his assessment -- the economic slowdown is "obvious not just in hindsight but for the past 6 months (at least)".

The Big Picture also has the Wall Street Journal's parade of punditry. Though there were exceptions, the general reaction was pretty sanguine, with the majority of commentators suggesting significant bounce back in the current quarter.  Tim Iacono noted that John Snow shares that majority opinion.  But Michael Shedlock asks "Exactly what data are you looking at Mr. Snow?"  Gerald Prante suggests one possibility, focusing on possible distortions in both the third quarter and fourth quarter stats, courtesy of the continuing aftermath of Katrina and Rita (a possibility also suggested by Andrew Samwick).  Brad Setser hopes the optimists "are not banking on a mirage" (of accelerating export growth, in particular).

For my money, it was a pretty disturbing, pretty confusing, even if ultimately inconclusive report.  The Capital Spectator identified what was on my mind:

The question is whether the advance GDP report constitutes reality, a line of inquiry that's found much attention today in the wake of the economic news. Knowing full well that each and every GDP report is revised, some are holding out the hope that today's 1.1% fourth-quarter rise will evolve into something more encouraging when the government dispenses the so-called preliminary report and then the final one.

Mark Thoma takes a look at the record, although he constructs the comparison the hard way, by examining archived vintage data sets (about which he expressed some skepticism).  Here's the record, directly from the Bureau of Economic Analysis.






Here's a close-up look of the past few years:




Even if the final estimate for the fourth quarter past matches the largest of these errors -- the upward revision in the third quarter of 2003 - we would end up at just 2% annualized growth for the quarter, still well below expectations.  The average positive error of about 0.54 percentage points would be well below that.  And, as noted by Jim Hamilton in his Econbrowser post, the fact that inventory growth was actually a positive in the latest report makes a positive revision far from certain.

In his post on the topic of the day, William Polley revealed that he "expect[s] that there might be a small revision upward in the next month or two."  The record suggests we shouldn't hope for too much more.

Related odds and ends:  General Glut claims we are seeing  the beginning of the end of the debt-driven consumption boom.  Menzie Chinn sees the beginning of the beginning of the much anticipated U.S. current account reversal.  The Nattering Naybob opines on why the market seemed to take the GDP report so well: "The slow down in Q4 GDP calmed inflation fears and sent the market on a tear."

A nice article on GDP revisions is Dennis Fixler and Bruce Grimm's aptly titled "Reliability of the NIPA Estimates of U.S. Economic Activity," which appeared in the February 2005 issue of the BEA's Survey of Current Business (brought to my attention and yours courtesy of the candle-burning Shadya Yazback).

UPDATE: Here, for the interested, is the data for the pictures above.  The file also includes the latest revisions for each quarter, as well as the record of revisions for the components of GDP:
Download GDP_quarterly_revisions_no_macros-1.xls

The original data can be found here.

January 29, 2006 | Permalink


TrackBack URL for this entry:

Listed below are links to blogs that reference Fourth Quarter GDP: Read It And Weep :


I was not surprised.

Posted by: anon | January 29, 2006 at 09:16 PM

The key is whare the GDP growth was coming from in the prior quarters -- the big spike in Q3 2003 was the massive tax cuts, and since then, the steady 3-4% GDP numbers have been driven primarily by home equity extraction.

As that slows, so too will the economy . . .

Posted by: Barry Ritholtz | January 29, 2006 at 10:22 PM

Looking at that first graph it appears that final numbers were often higher during strong/accelerating growth and lower during weak/decelerating growth.

Posted by: iasius | January 30, 2006 at 04:45 AM

Just fiddled around a bit with the data and there's an interesting pattern.
When growth decelerates (gdp growth is lower than the previous quarter) there are about as many positive as negative revisions.
However when growth accelerates there are almost no negative revisions.

If I were a betting man, I'd stay away from this one.

Accelerating growth: 13 positive revisions, 2 negative.
Decelerating: 8 positive, 9 negative revisions.

Posted by: iasius | January 30, 2006 at 05:16 AM

I took a look at the data from the first quarter of 1997. While I realize that this is a small sample, I found interesting that the GDP advanced 50% of the time from quarter to quarter. Also, if I am correct, when the advance GDP was up on a quarterly basis, the final data was higher than the preliminary 65% of the time.

Posted by: SamK | January 30, 2006 at 08:59 AM

Post a comment

Comments are moderated and will not appear until the moderator has approved them.

If you have a TypeKey or TypePad account, please Sign in

Google Search

Recent Posts



Powered by TypePad