A complicated calculation.

Photographer: Dan Kitwood/Getty Images

Don't Count on That U.S. Economy Data Quite Yet

Justin Fox is a Bloomberg View columnist. He was the editorial director of Harvard Business Review and wrote for Time, Fortune and American Banker. He is the author of “The Myth of the Rational Market.”
Read More.
a | A

This morning's U.S. gross domestic product report was a disappointment, with inflation-adjusted growth for the second quarter coming in at 1.2 percent, well below the 2.5 percent median forecast of the economists surveyed by Bloomberg.

I will leave to others the work of explaining what the various parts of the GDP report say about the economy. I'm interested more in the question of whether that 1.2 percent figure will bear any resemblance to what, in the fullness of time, the Commerce Department's Bureau of Economic Analysis will conclude that second-quarter 2016 GDP growth actually was.

GDP numbers get revised, again and again and again. Today's number was the "advance estimate" of second-quarter GDP. Next month we'll get the "preliminary estimate" and the month after that the "final estimate." Then there will be annual revisions, and eventually benchmark revisions that can go back decades.  The GDP numbers from 2002, for example, have been changed nine times since the release of the initial estimates, most recently in 2014. Sometimes these revisions reflect newly available data, sometimes (mainly with the benchmark revisions) they reflect changing views about what should be counted and how.

Occasionally these revisions can be pretty dramatic. I have written before about the first quarter of 2008, for which the advance estimate was that GDP had grown 0.6 percent. According to the most recent revision, it actually shrank 2.7 percent. For the fourth quarter of that year, the GDP decline has increased to 8.2 percent in the latest revision from 3.8 percent in the advance estimate. But this was during the onset of the worst recession in 75 years. It's not surprising that an advance estimate based in part on extrapolation from the past would underestimate its severity.

Out of curiosity, I put together a chart of all the advance estimates of GDP growth since 2002 and what the numbers have subsequently been revised to.

Gross Domestic Product, Before and After
Percent change in real GDP, annualized and seasonally adjusted
Source: Bureau of Economic Analysis

Looked at this way, the discrepancies don't seem so bad. Yes, there were some big misses around the recession, and the revised GDP numbers are more volatile than the advance estimates.  But overall, the two lines are telling more or less the same story. On average, the advance estimates were 0.3 percentage points higher than the revised numbers, but there have actually been more upward revisions since 2002 (28) than downward ones (26).

What does this tell us about second-quarter GDP growth? That it almost certainly won't end up at exactly 1.2 percent, but that the overall message of the GDP report -- the economy has slowed down since about a year ago, but hasn't stalled -- will probably hold up.

  1. For whatever it's worth, Bloomberg Intelligence economists Michael McDonough and Yelena Shulyatyeva put out a note about an hour before the GDP release warning that the new Census Bureau Advance Economic Indicators Report, released for the first time Thursday, seemed to indicate slower-than-2.5-percent GDP growth.

  2. I started in 2002 because that's how far back this handy BEA spreadsheet of GDP revisions (if you click, it will download the Excel file) goes.

  3. Standard deviation of 2.46 vs. 2.06.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

To contact the author of this story:
Justin Fox at justinfox@bloomberg.net

To contact the editor responsible for this story:
Susan Warren at susanwarren@bloomberg.net