“Their sabotage of the Affordable Care Act has already cost more than 3 million Americans their health insurance.”
— Former president Barack Obama, in a speech at the University of Illinois in Urbana, Ill., Sept. 7, 2018
Obama’s speech was a long call to arms for Democrats to vote in the upcoming midterm election. But there was a particular number that caught our attention — that Republican efforts to dismantle Obama’s signature Affordable Care Act had “already” led to the loss of health insurance for more than 3 million Americans.
Is that really the case?
Obama did not offer a source for his statistic, but his office supplied it: a survey by Gallup-Sharecare released in January. That survey found that from the fourth quarter of 2016 to the fourth quarter of 2017, the percentage of U.S. adults without insurance rose from a record-low 10.9 percent to 12.2 percent. “That 1.3 point increase represents an estimated 3.2 million Americans who entered the ranks of the uninsured in 2017,” Gallup said in a news release.
But that does not put Obama in the clear. He could have also cited a report known as the National Health Interview Survey (NHIS), run by the Centers for Disease Control and Prevention. The NHIS found an increase of only 700,000 uninsured people from 2016 to 2017 — what the report called “not significantly different from 2016.” In 2017, 29.3 million (9.1 percent) people of all ages were uninsured at the time of the NHIS interview. (A NHIS report released in August, for the first three months of 2018, said 28.3 million people of all ages, or 8.8 percent, were uninsured at the time of interview.)
On top of that, five days after Obama spoke, the Census Bureau issued its own health insurance estimate, based on the Current Population survey. “The percentage of people without health insurance coverage for the entire calendar year was 8.8 percent, or 28.5 million people,” the agency said. “The uninsured rate and number of uninsured in 2017 were not statistically different from 2016,” when 400,000 more people had health insurance.
As Obama should know, all three surveys are well regarded. His White House in 2014 issued a news release touting the decline in the uninsured as shown in the NHIS, while also referencing the Gallup poll, the Census Bureau and other surveys. The differences in the results depend on factors such as when people were interviewed, how they were interviewed and so forth.
Obama’s office referred us to a quote by Matthew Buettgens, a senior research analyst with the Urban Institute health policy center, to defend the use of the Gallup survey. So we checked with him to see whether he thought Gallup was a superior report.
“It is correct that estimates from the National Health Interview Survey and the Current Population survey did not find significant increases in the number of uninsured in 2017, while Gallup did,” Buettgens said. “Both of these surveys have a larger sample size than Gallup and both are used extensively by researchers. I do not consider Gallup to be better.”
Obama’s office had no further comment, except to note that Obama couldn’t predict what the Census Bureau would conclude.
The Pinocchio Test
This shows the perils of cherry-picking data. Obama selected a number that put the impact of Trump’s moves on Obamacare in the worst possible light, even though there was another well-respected survey that showed virtually no statistically significant decline in the number of people with insurance. Now the Census Bureau has backed up the NHIS conclusion that in Trump’s first year, the impact was minimal.
The story may be different in 2018, given that Congress eliminated the individual mandate that required the purchase of health insurance or the payment of a fine. But Obama stated the decline as an established fact. He had a responsibility to note he was only citing one survey and that there was contradictory data. He earns Two Pinocchios.
Send us facts to check by filling out this form
Sign up for The Fact Checker weekly newsletter
The Fact Checker is a verified signatory to the International Fact-Checking Network code of principles