Here’s why the U.S. revises GDP estimates — over and over

Within the U.S. Department of Commerce, there’s an office called the Bureau of Economic Analysis that’s responsible for estimating the nation’s gross domestic product.

GDP is supposed to represent the sum of all things produced in the economy. The bureau makes these GDP estimates for the year, and it does it for each quarter of the year. And it revises those estimates continually.

Today, the bureau announced its first revision of GDP for the fourth quarter of last year, saying it grew 2.7%. That’s less than its previous estimate – 2.9% for the same period.

What do these revisions mean?

That revision of two-tenths of a percentage point for what’s called headline GDP isn’t huge. But EY-Parthenon Chief Economist Gregory Daco said the components that make up GDP made some big moves in this first update.

“Consumer spending was revised down from an elevated reading to a much lower reading by 0.7 percentage points,” he said.

That’s notable because about two-thirds of GDP comes from consumer spending.

“That was partially offset by upward revisions, which were actually quite significant for business investment,” Daco explained.

All those changes reflect how the Bureau of Economic Analysis estimates GDP.

The BEA maintains a 454-page handbook of data points that go into the final number, said Austin Clemens of the Washington Center for Equitable Growth.

“You have things like casino gambling. They have a deal with the National Indian Gaming Commission, where they report some of their receipts to BEA,” Clemens said.

Others include college sports sales data from the NCAA and information about paid streaming subscriptions from the Recording Industry Association of America.

With inputs coming from so many sources, Clemens said, “a lot of it is coming to you in different kinds of formats, or varying degrees of quality.” 

One single survey might come back from, say, retailers, over the course of months, according to Jonathan Pingle, chief U.S. economist at UBS.

“For the first month, there’s only a few thousand respondents to that survey. A month later, they have over 10,000 respondents to that survey,” Pingle said.

The BEA makes revisions to include that new information, said Moody’s Analytics Chief Economist Mark Zandi.

He said GDP estimates, like so much of the data that economists look at, aren’t precise.

“It’s not like you’re reading an X-ray of the economy,” Zandi said. “It’s like you’re reading an impressionist painting of the economy.”

The revision released Thursday may be updated eight or nine more times, each version getting a little closer to the true picture.