From the moment that the Supreme Court handed down its decision on the Affordable Care Act, Democratic pundits by and large dismissed the threat that states, emboldened by the ability to reject the Medicaid expansion without losing access to federal funds for the existing program, would choose to opt out. “Eventually,” the punditry said, all states would accept the expansion. After all, it was such a “good deal.”

If “eventually” covers past 2022, i.e. ten years from now, perhaps they’re right. But that’s going to be far too late for millions of low-income citizens. And so the Congressional Budget Office’s updated estimate for the ACA should serve as a warning that the Supreme Court ruling significantly deteriorated the coverage provisions in the law.

The CBO’s estimates are a guess, based on their expectations of how many states will comply with the expansion over time. Their fairly conservative estimate suggests that the coverage expansion will be reduced by 10%, and that expansion on the basis of Medicaid will be reduced by 37.5%.

In this update of figures published in March 2012, CBO and JCT now estimate that fewer people will be covered by the Medicaid program, more people will obtain health insurance through the newly established exchanges, and more people will be uninsured. The magnitude of those changes varies from year to year. In 2022, for example, Medicaid and the Children’s Health Insurance Program (CHIP) are expected to cover about 6 million fewer people than previously estimated, about 3 million more people will be enrolled in exchanges, and about 3 million more people will be uninsured.

Looking at their figures, CBO said that 1/3 of the potential new Medicaid beneficiaries, from expanding coverage to 133% of the federal poverty level (this jumps around; CBO puts it at 138% FPL, but I’ve read 133% FPL in the text of the law, so I’m going with that), will live in states that accept the expansion. 1/2 of those potential new eligibles will see only a partial expansion, to either 100% FPL or less. Finally, 1/6 of the new eligibles will live in states where no coverage expansion is enacted.

This is a total guess, which CBO admits represents a middle range of outcomes. It could go far better or far worse than this. But as a baseline, it’s not a bad way to look at it. As a result, 6 million of the 16 million potential eligibles for Medicaid would not get their coverage. By 2022, about half of those people would be eligible to receive coverage on the insurance exchanges, with subsidies, if they make between 100% and 133% of FPL. Those under the poverty line fall into a no man’s land, where they are not going to be eligible for Medicaid, not eligible for exchange subsidies, and under a hardship exemption not responsible for paying a tax penalty on the individual mandate. They’ll just be uninsured. And by 2022, CBO estimates there will be 3 million of them.

This actually doesn’t make perfect sense, as later in the document CBO writes that “roughly two-thirds of the people previously estimated to become eligible for Medicaid as a result of the ACA will have income too low to qualify for exchange subsidies.” If that’s the case, then 4 million uninsured should result. And that’s what CBO estimates in 2014. But it changes that to 3 million by 2022, and the reasons behind that shift are obscure.

This has the related result of making the bill cheaper. Medicaid is cheaper than the exchange subsidies, so for 3 million, the federal government will pay more. But for the other 3 million who will now receive no coverage, the government will pay nothing, saving them lots of money. In fact, by 2022, the government will have spent $210 billion more on the exchanges, but $289 billion less on outlays for Medicaid (and to a lesser extent, the Children’s Health Insurance Program). Add in $1 billion more expected in mandate penalties (for those exchange eligibles who forsake coverage) and $4 billion in business penalties, and the bill is not $84 billion cheaper in the ten-year window. But that comes at the cost of a lack of coverage for 3 million Americans.

The report also notes that coverage on the exchange will have higher out-of-pocket costs than through Medicaid, so some people shut out of Medicaid will decide they cannot afford the exchanges, and opt out of coverage altogether. In addition, in states that do not adopt the Medicaid expansion, they will see less of a “woodworking” effect, where current eligibles not enrolled in Medicaid find out about the program and sign up.

This data should act as a big flashing warning sign for Democrats. Millions of people will be consigned to the fate of the uninsured, barred from any acceptable health coverage, because of the Supreme Court’s actions and expected state intransigence. And the consequences could range far worse than this estimate. There’s no way that anyone associated with the Democratic Party can look at these numbers and rest on their laurels, secure in the knowledge that the hospital industry will somehow come to the rescue and force governors to adopt the Medicaid expansion. CBO doesn’t believe that, and they’re probably right. If you think that forcing adoption of the expansion is a winnable fight, you need to get in that fight today.

More from Brian Beutler, who notes that Harry Reid’s immediate reaction, through a spokesman, to the report was to say that “The Affordable Care Act saves lots of money.” Ah, liberalism.