The jobs report released this morning was shockingly good — which immediately made some people question whether it was too good to be true. The most high-profile skeptic was left-leaning economist and New York Times columnist Paul Krugman, who wondered on Twitter whether the Trump administration had “gotten to” the Bureau of Labor Statistics (BLS), which produces the jobs report. (He apologized a few hours later.) But while it’s hardly the first time conspiracy theories about the report have circulated, it would be a big mistake to assume that the White House had a hand in making the statistics look rosier than they really are.
The same was true when Republicans questioned whether the Obama administration had cooked the numbers. Donald Trump himself said the unemployment rate was “phony” when he was campaigning for president, and that the real number was much higher. And in fall 2012, former General Electric CEO Jack Welch caused a stir when he suggested that the Obama administration had fudged the jobs report for political benefit.
Those conspiracy theories don’t engage with the facts, though. The BLS is insulated on many levels from political interference. And if there were some kind of meddling, economists outside the government would quickly pick up on it and sound the alarm. That doesn’t mean the numbers won’t change — some likely will be revised quite substantially. But that’s not a sign of tampering.
“This [pandemic] is a really difficult environment for producing economic statistics, and we know there are going to be errors and revisions of some kind, which the statisticians in the government are working hard to address,” said Tara Sinclair, an economist at the George Washington University. “But that is a very different thing than choosing to report numbers because they look good for your boss, or your boss’s boss. The way our statistical agencies are set up, that really cannot happen.”
For one thing, a successfully rigged jobs report would have to involve a conspiracy of several dozen government statisticians who crunch the numbers from the two surveys of households and businesses that make up the report, and who have access to the underlying data. Nearly all of them are career bureaucrats who have been with the agency through multiple administrations. Moreover, the process for producing the report is highly automated, which makes it harder to tamper with. The process is also insulated from the political appointees who might, in theory, have more of an incentive to produce a result that’s more favorable to the president.
Even the people in charge of the bureau don’t know what’s coming far in advance. Erica Groshen, who served as the commissioner of the BLS from 2013 to 2017, told me that she wouldn’t receive the final headline numbers from the jobs report until a few days before it was released. It would have been very difficult (if not impossible) for even someone in her position of power to tweak the numbers at that point, she said. “I got copies of the releases and I could ask for extra information, but I did not have access to the underlying files,” she said. “I wouldn’t have known how to tamper with them, and it would have been noticed if I had someone do it for me.”
This jobs report — while it certainly contradicted most economic forecasts — also didn’t raise any red flags for Groshen or Sinclair that might indicate the numbers were manipulated. For one thing, Groshen said, it came out on time, with no sign that there was unrest or discomfort from BLS employees. And Sinclair pointed out that simply defying expectations doesn’t make a result wrong. Economists quickly dove into the data and found several plausible explanations for why jobs had been added and the unemployment rate had fallen. And the separate surveys of the households and businesses found similar results — another indication that the findings, while surprising, are most likely accurately reflecting our economic reality.
Economists might be a little embarrassed that their predictions were so off the mark. But that doesn’t mean there was a vast conspiracy to doctor the data. After all, economists don’t have a stellar track record of predicting recessions — which is why, Groshen said, it’s so important to have high-quality economic data to rely on. “Sometimes our impressions of what is happening in the economy just aren’t right,” Groshen said. “That’s why we need the gold-standard data with consistency in how the information is collected and processed — because we don’t always know the answers ahead of time.”
CORRECTION (June 5, 2020, 4:47 p.m.): An earlier version of this article incorrectly stated that Jack Welch was the former CEO of General Motors. He was the former CEO of General Electric.
