How Accurate is the CBO?


No one gets too upset if weather forecasts are often wrong. Everyone realizes they’re just educated guesses. So why does the media think the Congressional Budget Office (CBO) is any better?

I recently received an email from a regular reader of this blog who expressed his concerns about the latest CBO predictions for the American Health Care Act (AHCA). Clearly he put a lot of faith in the CBO and the accuracy of its predictions. How accurate are those predictions?

Predictions v. Reality

The Wall Street Journal editorial board recently put this issue into perspective. On the same day that the CBO was releasing its latest predictions there were two reports of actual facts that were much more important yet were overlooked.

  • HHS released that average premiums in the individual market have increased 105% since 2013 – in the 39 states where ObamaCare exchanges are federally run.
  • BCBS of Kansas City withdrew its ObamaCare plans for 2018 from Kansas and Missouri. They cited losses of $100 million they called “unsustainable for our company.” That leaves 77 of Missouri’s 114 counties, including St. Louis, with one insurer and 25 counties with no coverage options. Premiums have increased 145% on average over four years.


These statistics are real and affect real Americans. They are not predictions from the CBO that may or may not happen. Yet these significant reports were completely overshadowed in the media by the predictions of the CBO.

The predictions of the CBO were actually encouraging on the whole. The CBO confirmed that the AHCA is a major fiscal dividend, which cuts taxes by $992 Billion, cuts spending by $1.1 Trillion, and lowers the federal deficit by $119 Billion over 10 years. When compared to an earlier version of the AHCA it will leave the number of insured “slightly higher” and the cost of premiums “slightly lower.”

These numbers were largely ignored by the media. They focused on the CBO prediction that said 14 million fewer people would be insured in 2018 relative to the ObamaCare status quo and this would rise to 23 million by 2026.

What do these numbers really mean?

Democrats have defined these predictions to mean that people will “lose coverage.” But that isn’t what the CBO predicted. They estimated that 14 million on Medicaid today would roll off the Medicaid program as it shifted to block grants – or a 17% drop in enrollment after the ObamaCare expansion. The intent here is to focus Medicaid on the poor and disabled, as it was originally intended, rather than on able-bodied, working–age adults.

The remaining people who would “lose coverage” (according to Democrats) would actually be people choosing to drop their coverage “because the penalty for not having insurance would be eliminated” in the words of the CBO. In other words, people would refuse to purchase insurance they didn’t want when the threat of a government tax is eliminated. ObamaCare insurance isn’t worth the price.

How accurate is the CBO?

This brings me back to the original question. The CBO has made many predictions about ObamaCare since its inception in 2010. Below is a graph that depicts the accuracy of these predictions:


Clearly the CBO has been overly optimistic when making ObamaCare predictions in the past. This raises the question of why they are being overly pessimistic about the AHCA – and whether or not they are truly non-partisan as advertised. In truth, the CBO is quite poor when trying to predict market-based reforms, such as the 2003 Medicare Part D drug benefit whose costs the CBO badly overestimated.

Are there errors in the CBO methodology that impact their predictions?

Seth Chandler, writing in Forbes, points out some significant errors. In March, 2016, the CBO predicted there would be 18 million enrolled in ObamaCare in 2018. Despite the actual numbers enrolled falling far short every year of their predictions (see graph above) the CBO used the same predictions in May, 2017, when they were projecting ObamaCare enrollment for 2018 through 2026. This assumption greatly influences the comparisons between the status quo (ObamaCare) and the proposed new legislation (AHCA).

In other words, the CBO is using faulty data about the number of insured today under ObamaCare to compare and project the number of insured under the AHCA. Despite the impending implosion of ObamaCare, with prices skyrocketing and insurers pulling out of the exchanges, they are assuming there will be eight million more insured next year than this year!

The old saying about computers is “garbage in – garbage out.” The same could be said of the CBO predictions predicated on 2018 ObamaCare enrollment projections that only the foolish would believe. Chandler explains:

“The problem is that the failure of the CBO to update its enrollment projections means that its enrollment projections for the ACA and its cost projections for the ACA, which depend on enrollment, are going to be wrong.”

Why does this matter?

The Wall Street Journal puts it this way:

“Headlines aside, the CBO report matters because it is the fiscal template for Senate negotiations and what policies can be included under the budget “reconciliation” procedure that requires 51 votes. But Senators shouldn’t allow the budget scorekeeper’s opinions about the future to dictate policy or political decisions. Incentives and private competition can produce better outcomes than CBO’s model foresees.”


We must focus on the real world facts that show ObamaCare has been a colossal failure in holding down prices and providing real improvement in healthcare outcomes. These are facts the CBO never saw coming.

No comments yet. You should be kind and add one!