Democrats and Republicans have begun arguing over whether the U.S. economy is in a recession ahead of a key data release on Thursday.

But the official pronouncement will ultimately come down to a little-known group of economists selected by the National Bureau of Economic Research called the “Business Cycle Dating Committee,” which stubbornly takes its time and tries to wall itself off from political interference or attempts to spin its findings.

The stakes for the group are high, in part because of the extraordinarily unusual economic conditions two years after the last recession, early in the coronavirus pandemic. The economy contracted in the first quarter of the year, and many Republicans say a recession is here already, with Thursday’s release expected by many analysts to show a second consecutive quarter of negative growth. But from President Joe Biden down, administration officials are instead pointing to other indicators showing the economy remains strong and insisting the committee would be wrong to declare a recession.

The political haggling is not supposed to matter to the eight economists who rule on when recessions start. Their decision is almost certainly months away, if it comes at all: The committee typically waits long after a recession has begun to declare it, only acting when the evidence has become overwhelming, sometimes even after the recession is already over. That puts the pressure bearing down on the organization from the outside — to promptly render a verdict on one of the most important matters facing economic policymakers — directly at odds with its mission to provide unassailable empirical decisions.

As a result, what seems like a straightforward question — is the U.S. economy in a recession? — is in part decided on a subjective basis at a later date, sometimes when it no longer appears pertinent, by experts in closed-door meetings of a privately selected committee.

“By far, the most important thing to try to convey is that the committee is not trying to do real-time dating of whether we’re in a recession,” said MIT economics professor James Poterba, the NBER president and a member of the committee, in an interview. “There’s often enormous amount of interest in that question and what many people are hoping for, but the committee’s task is to create a consistent historical record of the turning points — the peaks and the troughs in the U.S. economy.”

The group’s calculus could become increasingly fraught in the following months, amid puzzling economic conditions that defy easy characterization. The political consequences for the committee could be significant, as the Biden administration faces growing public anger over high inflation and its economic stewardship. Congressional Republicans will also be eager to seize on a ruling that the economy is in a recession, trying to capitalize on voter discontent ahead of this fall’s midterm elections.

Asked about the upcoming economic numbers, Biden on Monday disputed the notion that a recession was imminent. That is part of a broader administration campaign in recent weeks to rebut GOP claims that a recession has already begun. Top economic officials, including Treasury Secretary Janet Yellen and White House National Economic Council Director Brian Deese, appeared on cable TV news Sunday and Monday to reiterate their view that the U.S. economy is not technically in recession — and wouldn’t be even if the GDP numbers show a second consecutive quarter of contraction.

This strategy has its risks, though, because if the United States does enter a recession later, their current assurances will look misguided — particularly after the administration already incorrectly dismissed the threat of inflation last year.

“We’re not going to be in a recession, in my view. The [unemployment] rate is still one of the lowest we’ve had in history,” Biden said on Monday. “My hope is we go from this rapid growth to steady growth.”

At the core of the challenge facing the committee of economists is that it relies on more than a half-dozen criteria to measure when a recession has begun. The general impression many Americans — and some commentators — have is that a recession is defined as two consecutive quarters of negative economic growth. But that is not how the NBER, or most economists, think of it. Instead, the committee weighs factors such as payroll levels, retail sales, industrial production and personal income when making a comprehensive assessment about whether the economy is in recession. The committee stresses on its website that “there is no fixed rule about what measures contribute information to the process or how they are weighted in our decisions.”

As Deese told CNN: “In terms of the technical definition, it’s not a recession — the technical definition considers a much broader spectrum of data points.”

Traditionally, all these various economic metrics move in tandem, making the committee’s job easier. Usually, when growth declines, so do employment, consumer activity and the other measures of economic health. But the economy since the start of the pandemic has confounded prior models and may do so again. Economic growth may end up declining for two consecutive quarters — although the first quarter in the United States was negative due largely to technical factors, such as a temporary uptick in overall imports — even as unemployment remains among the lowest levels in American history. Consumer spending has remained strong, too, in a way at odds with a typical recession. If unemployment remains low even as growth contracts, the NBER economists could face a vexing challenge in deciding how to categorize the situation.

Predicting the committee’s decision is made more difficult by the way in which it operates. As part of the NBER, the Business Cycle Dating Committee is run by a private nonprofit group — not the federal government or a state statistical agency. Its membership is selected by the president of the NBER “in consultation” with the committee chair, according to Poterba.

The committee’s meetings are not publicized. They’re held in a closed-door conference room on the third floor of the Cambridge, Mass., office building where NBER is headquartered. They don’t meet on a fixed schedule: Board Chairman and Stanford economist Bob Hall is responsible for calling the meetings. During long periods of consistent economic growth, the board can go years without having anything to discuss, and therefore, it might hold no meetings. It wouldn’t even confirm when past meetings have happened.

“The committee does not announce its meeting schedule, and that’s something we don’t talk about,” Poterba said.

Its last public pronouncement came on July 19, 2021 — when the committee declared that there had been a recession between February and April 2020, the shortest one in U.S. history.

The eight economists on the committee are among the most respected in their field. Some have served in Democratic administrations, but past members have also included GOP appointees. In addition to Poterba and Hall, the members are Christina Romer and David Romer of the University of California at Berkeley; James Stock, of Harvard; Robert Gordon, of Northwestern; Valerie Ramey, of the University of California at San Diego; and Mark Watson, of Princeton.

The NBER has its roots in the period following World War I, after a Columbia-trained economist who worked for labor and trade organizations and the chief statistician at AT&T formed a new organization after realizing they had little shared empirical data with which to conduct policy debates. In the early 1960s, the Commerce Department began publishing a digest on business conditions that cited the NBER’s work on the ups and downs of the business cycle, giving it a kind of federal imprimatur, according to Poterba.

Poterba stressed that the board is aware of public hunger for guidance about a recession but does not let it dictate their decisions. Economic data is often later revised, and the committee is careful not to announce a verdict that is contingent on data that could later be changed.

“The NBER is really trying to provide guideposts for researchers; it’s not trying to provide short-term political talking points for either party,” said Steve Miran, who served as a senior official in the Treasury Department under Donald Trump’s administration and is the co-founder of Amberwave Partners, an investment fund. “We’d all like it to be binary — 0 to 1, recession or not — but the truth is it’s much more of continuum. It requires interpretation of the length, depth and speed of the contraction, as well as which sectors of the economy are contracting and why. ... And that requires an element of judgment.”

Still, that does not mean the board’s members are always in agreement. Harvard economist Jeffrey Frankel, who served as a member of the committee for roughly 25 years, said there is usually unanimity about the broad questions of whether a recession is beginning or ending, but differences can emerge about precisely what month a recession began or ended.

“There are times when the right answer is not clear, and the kind of thing there may be disagreement about is someone wanting more data — like revisions to the [gross domestic product], for instance — and someone else saying, ‘It’s already been 11 months, and if we wait longer people will think the news will be too stale,’ “ Frankel said. “That tension is always an issue.”