Policymaking in Washington can live and die by “the baseline” — in particular, the baseline set by the Congressional Budget Office.

During the debate over the 2017 Republican efforts to repeal the Affordable Care Act, Democrats often cited estimates that the repeal would result in 23 million fewer people having health insurance within a decade. That number came from the CBO’s estimate of the impact of the law versus what it predicted would happen under existing policy — i.e., “the baseline.” The big part of the change was because the law proposed to repeal the mandate that individuals must be insured.

The GOP tax bill included an effective repeal of the mandate — the penalties for failing to comply were dropped to zero. An article in the Washington Examiner caught our eye, which led us to study a report issued Feb. 20 by the Medicare Office of the Actuary on its methodology for 10-year projections on health-care spending.

There, we saw that the impact of eliminating the individual mandate was considered almost a nonevent. That sent us scurrying back to a May 2018 estimate by the CBO, where we learned the agency had revised its thinking on the impact of the mandate.

So what’s going on? Did the CBO blow it? Let’s take a look.

The Facts

This is not just an arcane discussion about numbers. Readers may recall that Democrats frequently attacked the repeal effort with these numbers (using inflated language that the GOP would “kick X millions of people off” insurance). Republicans fired back, saying numbers could not be trusted. The White House even released a video attacking the CBO, accusing it of “faulty assumptions and bad numbers,” including “faulty baseline estimates.”

It’s important to remember that an estimate made from a baseline is simply a snapshot in time — what the analysts believe will happen at the moment, based on the information available. Lawmakers must rely on those numbers for months — until new data is received and another estimate is made.

In the case of the initial estimate, the GOP leadership shot itself in the foot by insisting that the CBO (which takes its orders from Congress) rely on a baseline from March 2016, rather than a more recent January 2017 baseline. That’s because the GOP-led Congress in January was in such a rush to repeal Obamacare that it moved forward before the CBO could complete an updated baseline, as it does regularly during the year.

So budget instructions for the year were based on the March 2016 baseline, and the CBO was ordered to do its analysis of health-care proposals based on the old baseline.

The March 2016 baseline had indicated 18 million people would be on the Obamacare exchanges in 2018. But the January 2017 baseline scaled back the figure and estimated that 11 million people would participate in the exchanges in 2018, a huge drop.

But the GOP did not want the CBO to switch to the more recent baseline because it might have resulted in less budget savings — and thus made it harder to pass the health-care bill (which failed anyway).

At the time, based on limited data available about how the Affordable Care Act impacted insurance markets, the CBO assumed that without a mandate, many people, especially younger, healthier Americans, would simply choose to not buy health insurance.

That led the CBO to estimate that American Health Care Act would lead to 14 million fewer people insured than previously anticipated by 2018, the second year of the House GOP bill. Eight million of those 14 million would be people who had coverage through the individual insurance market; 4 million would be people with coverage under Medicaid; and 2 million would be people with coverage through their employers, which also would no longer be required to provide insurance. That 14 million figure grew to 23 million fewer people with insurance by 2026.

But the GOP push to repeal Obamacare collapsed in the Senate, and Washington moved on to the battle over the tax bill. When the individual mandate yet again appeared to be on the chopping block, CBO in November 2017 published a budget estimate of the savings — which included what the impact would be on people with health insurance.

Now the agency said 4 million fewer people would have insurance in the second year — not the 14 million from the estimate six months before. And the 10-year number was 13 million, not 23 million. (Note: the 23 million figure in May included the impact of a number of policy changes. A July report by the CBO more carefully defined the impact of individual mandate using the 2016 baseline: 15 million fewer insured in the second year, 16 million at the end of ten years.)

What happened? The agency finally was using a later baseline, from the middle of 2017, rather than the 2016 one. Moreover, the agency started rethinking whether removal of the individual mandate would be as dramatic as it once assumed, as the estimate incorporated the expectation that individuals’ and employers’ full reaction to the elimination of the individual mandate would phase in more slowly than the agencies previously projected.” (The “agencies” also included the Joint Committee on Taxation.)

The report added that the work was ongoing:The agencies have undertaken considerable work to revise their methods to estimate the effects of repealing the individual mandate. … Because that work is not complete and significant changes to the individual mandate are being considered as part of the budget reconciliation process, the agencies are publishing this update now without incorporating major changes to their analytical methods.” The report added that the preliminary indication was “the estimated effects on the budget and health insurance coverage would probably be smaller than the numbers reported in this document.”

The big problem the CBO faced was that it was trying to model human behavior. The initial assumption was that people were motivated to buy insurance because of the mandate — if an individual did not get insurance, he or she would face a fine. But there were other possibilities: Maybe all the news about Obamacare made people realize that relatively inexpensive insurance (after subsidies) was available in the individual market? Maybe people had developed a stronger belief that they needed to get insurance — and they understood that the Obamacare marketplace was an easy place to find it?

By May, the CBO had incorporated this analysis in its estimates. The impact of the mandate was downgraded even further, to just 2 million fewer insured in 2019 — and 8 million by the end of the decade. So that’s a 38 percent reduction from the November 2017 report and a drop of 50 percent from the report that caused such a stir a year earlier. (Note: we have updated this chart to reflect the July report that more closely modeled the impact of just the mandate.)

It’s important to understand that there were other things going on in the market that affected the CBO analysis. The Trump administration in late 2017 decided to stop making cost-sharing reduction (CSR) payments to insurers (We covered this here.) Insurance companies and states responded with a workaround that, ironically, helped boost premium subsidies and made plans more affordable — and increased participation in Obamacare exchanges.

The Medicare actuary believes the impact of the individual-mandate repeal is even less: “The elimination of the individual mandate in 2019 is assumed to result in a small enrollment decrease, primarily concentrated in 2019” — about 2.5 million — eventually rising to an increase of 4 million in the uninsured population by 2021.

The two groups of experts especially differ on the impact on Medicaid enrollment. The Medicare actuary says the impact is almost nil: “There is no assumed increase in the number of uninsured related to the repeal of the mandate as that coverage has no premium and little-to-no out-of-pocket costs.” The CBO still believes there will be some impact under the theory that the mandate led people to get insurance, only to discover they were eligible for Medicaid.

The Bottom Line

CBO forecasts do not always look great in the rearview mirror. The agency was certainly hamstrung at the start by the GOP’s demand that the CBO first estimate the health-care repeal with an old baseline. Ironically, that helped make the impact of the GOP plan look even worse. Over time, with more data about how Americans reacted to the individual mandate and fresher baselines, the CBO became more conservative in estimating its impact.

Whether it made a difference in the political battle is unknowable. But it would have made the Democratic talking points less scary.

(About our rating scale)

Send us facts to check by filling out this form