Skepticism of CBO projections may appear to be warranted if we consider:
1. In March of 2010 the CBO said 21 million people would buy insurance on exchanges created as a result of the Affordable Care Act (ACA) by 2016. In fact, only 11.5 million people purchased coverage by the end of 2016
2. That same report said there would be 30 million fewer uninsured people in the US. The reality is the CBO reduced that number to 22 million in March of 2016.
Why is there such a gap between the CBO estimates and reality?
First, I think we need to understand the basis of the CBO’s baseline projections. CBO projections are based on the “assumption that current laws governing federal revenues and spending will generally remain unchanged.” Here is how the CBO further explains their projections:
The baseline projections are not intended to be a prediction of budgetary outcomes. Rather, the projections reflect CBO’s best judgment about how the economy and the budget will evolve under existing laws. That approach allows the baseline to serve as a neutral benchmark against which Members of Congress can measure the effects of proposed legislation.
This means projections will be modified when there are relevant legislative changes. One wonders how anyone could be in the ballpark in any projection or forecast if we couple possible legislative changes with unpredictable human behavior.
Let’s consider again the CBO’s estimates of how many people would purchase insurance on the new insurance exchanges. There were several mitigating factors here:
1. The CBO thought more employers would stop offering health insurance to their employees
2. More people were eligible for Medicaid than the CBO expected. This means the CBO underestimated how many people would get coverage through Medicaid.
3. The Supreme Court ruled in 2013 that the Medicaid expansion in the ACA was optional for states
The CBO changed its estimates after the 2013 Supreme Court decision. The revised forecast said 89% of Americans would have insurance by 2016 and, in fact , 89.7% of Americans under 65 had insurance in 2016.
A formal study was done by the Commonwealth Fund and it concluded:
Simulations of the effects of health insurance reforms have received considerable attention over the past two decades, leading to substantial improvements in modeling. The CBO, and several private forecasters, were fairly accurate in their predictions of the likely coverage and cost implications of the ACA. A few forecasters—notably the CMS—assumed much higher rates of responsiveness to subsidies and coverage expansions, and these models generated the least accurate predictions. CMS estimates of participation in subsidized coverage, Medicaid enrollment, and total marketplace spending were 2.7, 2.0, and 2.9 times, respectively, higher than actual figures.
I think we can trust the findings of the CBO as long we keep in mind the reasonable limitations mentioned in this series.