In the political hullabaloo over efforts to shift costs of health care to someone else, the argument for keeping Obamacare’s compulsory insurance and ever‐​expanding Medicaid enrollment relies naïvely on notoriously comical Congressional Budget Office (CBO) 10‐​year “projections.”

CBO claims the initial House Republican plan would eventually cause 3 million to “lose” health insurance simply because they would no longer be fined up to 2.5% of income for not buying a policy designed by and for politicians. This not a loss, but a gain – in freedom of choice.

CBO claims the GOP plan would “lose” another 14 million by not expanding Medicaid enrollment as rapidly as Obamacare hopes to. The federal government pays about 57% of the cost of Medicaid for poor people, but 90–93% (until 2022) to the 31 states that provide Medicaid to those earning up to 138% of the poverty line. That has added 17 million to the Medicaid rolls, and enriched big health insurers and Kaiser Permanente.

Since expanded Medicaid added only 17 million, how could the Republican plan’s modest frugality possibly subtract 14 million? Only because CBO assumes that if nothing changes then more and more states will leap on this stalled bandwagon, thus shoving millions more on Medicaid. This seems politically unlikely (few Republican governors have done so), but also unhealthy. All those new patients on Medicaid may have to drive very far to find an MD who accepts Medicaid. Having a third‐​rate health insurance card is not “health care.” It’s a card.

Some of the biggest public policy blunders of recent years resulted from taking CBO’s crystal ball far too seriously. Higher 1991 tax rates on high incomes and luxuries worsened the recession and lost revenue, for example, yet seemed courageous by the low standards of faulty CBO bookkeeping.

An even better example of CBO dominating policy decisions was the foolishly phased‐​in and poorly‐​targeted 2001 Bush tax cuts.

In January 2001, CBO predicted that budget surpluses would add up to $5.6 trillion by 2011. That absurd CBO forecast made their (overestimated) $1.35 trillion 10‐​year revenue loss from look puny – merely nibbling away at 24% of projected surpluses. Little wonder the 2001 law was a grab‐​bag of feel‐​good giveaways (10% tax rate, refundable credits, etc.) thanks in large part to the CBO’s blurred vision of endless surpluses.

Why does this have anything to do with health insurance policy? Because CBO projections for the Affordable Care Act (ACA) have also been embarrassingly wrong. In March 2010 CBO estimated only 21 million would be uninsured by 2016, but the actual number turned out to be 27 million. CBO expected 21 million policies on the Exchanges by 2016, but the actual number turned out to be 12 million.

By using CBO projections as the baseline, as is being done with the GOP plan, it could be said that Obamacare “lost coverage” for 6 million people in 6 years –compared with 2010 CBO projections.

In short, the CBO has always grossly overestimated the success of Obamacare. And they still do. They appear to underestimate how much larger federal subsidies to Medicaid and the Exchanges “crowd out” other insurance – luring people out of unsubsidized insurance into subsidized insurance, or into seemingly “free” expanded Medicaid in 31 states.

The CBO’s rosy scenario about the future success of Obamacare exchanges, and about future growth of expanded Medicaid coverage, results in a grossly exaggerated estimate of the number of policies to be “lost” by reducing compulsion, mandates and taxpayer subsidies. Good policy requires more good judgment, not more bad estimates.