Rejigging America’s GDP calculation requires dubious math, but it always did

We may earn a commission from links on this page.

Calculating a country’s gross domestic product is already an arcane business. So it’s little wonder that a few eyebrows went up yesterday on word that the US—specifically the Bureau of Economic Analysis (BEA), which does the country’s GDP estimates—plans to start counting a bunch of intangibles as part of GDP. These will include research and development and the creation of movies and other “artistic originals,” potentially raising GDP by as much as 3%, under a set of international standards that have been in the works for years.

Individual companies have a hard enough time putting a value on such things. Intangible assets are known to be a weak spot in corporate accounting; they’re hard to value objectively, and companies have some leeway to fudge figures if they really want to. Can economists really hope to do it on a larger scale?

“It’s too easy to get wrong,” says Jack Ciesielski, editor of the Analyst’s Accounting Observer, a respected newsletter. “There are a lot of identifiable tangibles where we haven’t done a great job of getting the economic measurement right.”

One potential complication: It’s easy to add up how much companies spend on R&D. It’s harder to know how much it actually produces in terms of revenues—sometimes years or decades later. That’s why corporate accounting treats R&D as an expense, rather than capitalizing it like more concrete forms of investment.

Moreover, some kinds of R&D spending really amount to what economists call rent-seeking behavior, not genuine innovation. Imagine one pharmaceutical company spending millions (or even billions) to circumvent a competitor’s patent. Even assuming the resulting product gets to market, it doesn’t really so much add anything to national output as cannibalize the competitor’s sales. But the R&D spending would, under the new rules, be counted in GDP.

“We know that goes on—is that 10% of research? Is it 30%?” says economist Dean Baker, co-director of the Center for Economic and Policy Research. “If that ends up being a lot of what we’re counting, then we’re overstating GDP.”

Another complication: If a company spends in the US to develop some intellectual property, and then transfers it to an Irish subsidiary that generates most or all the revenue it produces, should that R&D contribute to GDP in the US, or Ireland? Multinational companies are forever moving intellectual property across borders to maximize tax savings and other benefits, leading to income-shifting gambits like the Dutch Sandwich or Double Irish problem.

As for those “entertainment, literary, and other artistic originals,” which will also be counted toward GDP as fixed investments: It’s hard enough for movie companies to match marketing costs to revenue, Ciesielski notes—just think Ishtar, or John Carter.

Then there’s the sheer practical headache of it all, as The Guardian’s Heidi Moore laid out in a critical piece. The BEA will be revising GDP back to the beginning of economic time in the US, or 1929. The prospect of rewriting more than 90 years of economic history is daunting.

The BEA responds

The BEA’s director, Steve Landefeld, acknowledges that many of these are serious concerns. In fact, they’re “exactly the questions we’ve been working on over the years,” he told Quartz.

His answer to the skeptics largely boils down to this: Putting a number on intangibles won’t be be an exact science, but when it comes to calculating investment for GDP purposes, not much is.

For instance, a hundred dollars spent on a drill press may not prove to be a good investment in the long run, but it still is counted in GDP. Similarly, the new plan largely just adds up spending on R&D.

“Whatever you spent, you assume that’s what it’s worth,” Landefeld said. “It’s arbitrary, but we and every other country in the world do it with everything we count.”

Keeping the math relatively simple, he argues, also protects against subjective factors—such as assumptions about return on investment—that could allow the measure to be skewed. “It could be wrong, but it’s not subjective,” he says.

The overall system also attempts to take cross-border issues into account. Assets transferred to Ireland, and royalties sent back to the US, are supposed to be captured by balance-of-payment calculations. The calculation may prove flawed, but it’s at least an approximation, Landefeld argues.

As for tossing out decades of economic research based on the old data, Landefeld says—don’t. The agency has revised the GDP calculation to varying degrees every five years, typically expanding what the measure encompasses. This latest expansion may go deeper than some, but it isn’t expected to make a substantial difference in long-term growth measure or “the picture of the business cycle,” Landefeld says. The agency has earlier said, he notes, that “economic policy decisions should not need to be reconsidered in the light of revisions to GDP estimates.”

In the end, Baker says, countries have a relatively simple choice: Do their best to calculate the effect of these intangibles, or leave out a significant chunk of economic activity.

“Are we better off trying to imperfectly measure what the enduring value is from R&D, knowing we can’t do that perfectly, or just throw up our hands and say we’re going to leave it out of the accounts?” Baker said. “I can’t say whether they’ve done it right, but trying to do it is good.”