How Will Candidate Plans Affect Total Health Costs?

We recently updated our estimates of the budgetary impacts of health care reform plans put forward by 2020 presidential candidates Joe Biden, Bernie Sanders, and Elizabeth Warren; and former candidate Pete Buttigieg. We found that all four plans would significantly increase federal spending on health care (though by different amounts), and all but Mayor Buttigieg's would likely add to federal budget deficits (also by different amounts) over the course of a decade.

While these proposals' effects on the federal budget are important to consider when evaluating their costs and benefits, so too are their effects on national health expenditures (NHE) – the total amount of health spending from both the public and private sectors. NHE represent an all-encompassing measure of health care spending in America, including spending by the federal government, state and local governments, households, and employers. Each plan would increase government spending on health care but would also result in lower private spending by households and employers.

Our analysis below was produced prior to Mayor Buttigieg dropping out of the race on March 1, and we continue to show the effects of his plan – which we estimate would have reduced health costs, lowered deficits, and expanded coverage.

Two of the plans we evaluated would likely reduce NHE: the Biden and Buttigieg plans. According to our central estimates, the Biden plan would reduce NHE by 1 percent over the next decade relative to current law, while the Buttigieg plan would reduce NHE by 2 percent. The Sanders and Warren plans, on the other hand, would increase NHE by 6 percent and 3 percent, respectively.

Share this graphic on Twitter and Facebook.

By their nature, these estimates are uncertain, which is why our analyses include ranges. Under our low-cost estimate, the Biden, Buttigieg, and Warren plans would reduce NHE by 3 percent, 5 percent, and 4 percent, respectively, while the Sanders plan would increase them by 2 percent. Under our high-cost estimate, the Biden and Buttigieg plans would each increase NHE by 1 percent, while the Sanders and Warren plans would increase NHE by 14 percent, and 11 percent, respectively.

Estimating Medicare for All Plans

Theoretically, Medicare for All could increase or decrease NHE depending on whether or not private spending would fall by more than public spending would rise under such a plan. Expanding coverage to more people and including more services (especially long-term care services) would result in higher total expenditures. So too would eliminating deductibles, copayments, and other cost sharing because the lack of point-of-service payments would lead to higher utilization of care. On the other hand, under Medicare for All, the government could use its regulatory and monopsony buying power to reduce the cost of medical services and prescription drugs, reducing total expenditures. Medicare for All would also likely involve lower administrative costs than the current system, further reducing expenditures.

While the ultimate effect of Medicare for all on NHE is theoretically ambiguous, it is our assessment that a comprehensive cost sharing-free plan that provides for universal long-term care and pays Medicare rates would likely increase costs, particularly over the first decade.¹ Some estimates do find costs would decline significantly, but those estimates generally do not account for the expansion of long-term care benefits under Medicare for All proposals and most rely on a variety of heroic assumptions. For example, a recent study by Alison Galvani, Alyssa Parpia, Eric Foster, Burton Singer, and Meagan Fitzpatrick not only fails to appropriately account for long-term care, but assumes – counter to a substantial body of empirical literature – that eliminating cost sharing would have no effect on the use of medical services for those who currently have adequate insurance. The study also assumes both extremely ambitious reductions in administrative costs and in fraud. These assumptions are unrealistic on their own and particularly when taken together, as adequate administrative spending is necessary to achieve fraud reduction. These and other flaws are explained in footnote 16 of our Primary Care paper and in pieces by Politifact, Vox, and others. (On Twitter, health expert Adrianna McIntyre said, "I wonder – quite sincerely – how this article got past peer review.")

Our estimates of NHE under the Sanders and Warren plans are derived from an estimate of one version of a Medicare for All plan, which was analyzed by numerous authors and written up by Linda Blumberg, John Holahan, and Michael Simpson at the Urban Institute. These scholars estimate NHE will total $52 trillion from 2020 through 2029 under current law and $59 trillion under Medicare for All, a 13 percent increase. From 2021 through 2030, we project NHE would total roughly $54.5 trillion under current law and $61.5 trillion under Medicare for All.

To arrive at our central estimate of the Sanders plan, we assume administrative costs would be 4.5 percent of overall costs as opposed to 6 percent ($750 billion savings) as Urban assumes, and hospital payments would be set at current Medicare rates as opposed to 115 percent of Medicare rates ($1.8 trillion savings). We also incorporate a plan to reduce brand-name drug prices by 50 percent as opposed to 25 percent ($850 billion savings). These and smaller assumptions result in total NHE of roughly $58 trillion from 2021 to 2030, a 6 percent increase from baseline levels of $54.5 trillion.

Our central estimate of the Warren plan also assumes administrative costs at 4.5 percent as opposed to 6 percent ($750 billion savings) and a substantial further reduction in brand-name drug prices ($850 billion savings). It assumes hospital payments at 110 percent of Medicare rather than 115 percent ($600 billion savings), the enactment of further provider payment reduction and reform measures ($2.5 trillion savings), and the enforcement of a spending cap through further provider cuts ($500 billion savings). These and smaller assumptions result in total NHE of over $56 trillion from 2021 to 2030, a 3 percent increase from baseline levels.

Note: Numbers in table may not sum due to rounding

While these estimates show costs increasing, cost reduction is also possible under the right set of circumstances and assumptions. Our low-cost estimate of the Warren plan, for example, finds NHE would fall by about 4 percent. Relative to our central estimate, this low-cost estimate assumes much lower administrative costs (2.3 percent as opposed to 4.5 percent), deeper cuts to drug prices, a smaller effect on utilization, and an effective use of global budgets and other tools to keep cost growth in line with the overall rate of economic growth. Taken together, we find NHE would total roughly $52.5 trillion over a decade under these assumptions.

Estimating ACA Expansion and Public Option Plans

Like Medicare for All, proposals to expand the Affordable Care Act (ACA) and offer a public option involve a theoretically ambiguous effect on NHE, especially when those plans make other changes that affect overall health care costs.

On one hand, policies that result in an increase in the number of people with insurance, lead to coverage of more services, or reduce cost sharing would all increase costs by expanding the use of health services. On the other hand, public options that pay providers at-or-near-Medicare rates would reduce overall health care costs by providing cheaper care to those they cover and increasing competitive pressure on private plans. Proposals to limit drug prices and private sector provider payments or reduce Medicare costs would also generally lower overall health spending, while proposals to reduce Medicare cost sharing and expand long-term care coverage would increase overall spending.

Estimating the net impact of these types of plans on NHE is much more complicated than estimating the same impact of Medicare for All plans because it requires understanding the effect of policies on private-sector spending (under Medicare for All, nearly all spending becomes public-sector spending). For this reason, our estimates of the Biden and Buttigieg plans are more uncertain than those of the Sanders and Warren plans, and our low- and high-cost estimates are thus more meaningful.

In our central estimates for both the Biden and Buttigieg plans, we assume the core policies of expanding the ACA and offering a public option would have a minimal impact on overall health expenditures. This is consistent with a number of Urban Institute estimates of various plans. We assumed the core plans would increase expenditures in our high-cost estimates and reduce them in our low-cost estimates.

Outside of the coverage expansion proposals, we estimate that Vice President Biden’s plan includes about $550 billion of additional federal spending and $450 billion of federal spending reductions. Though these policies would increase federal spending on net, we estimate they would reduce NHE by roughly $400 billion, with especially large non-federal effects from drug pricing and surprise billing reforms. This amounts to a reduction in total NHE of less than 1 percent.

Mayor Buttigieg’s plan includes roughly $1.25 trillion of spending increases outside his core proposal, along with $1.2 trillion of spending reductions. In total, we estimate this would result in a nearly $1.3 trillion reduction in NHE (more savings to the private sector than costs to the public sector) – driven by drug pricing and surprise billing reforms, reductions in Medicare provider payments, and a cap on private-sector out-of-network payments. This equates to a reduction in NHE of over 2 percent.²

Accounting for uncertainty in these estimates, we find the Biden plan could reduce NHE by as much as 3 percent and increase them by more than 1 percent under our low- and high-cost estimates, respectively. Likewise, we find the Buttigieg plan could reduce NHE by as much as 5 percent and increase them by less than 1 percent.

Read our full analysis of all four plans in Primary Care: Estimating Democratic Candidates’ Health Plans.

1 For example, Liu and Eibner (RAND) find Medicare for All would increase NHE by 2 to 10 percent, while Blumberg, et al. (Urban) find it would increase costs by 13 percent. A study from Blahous (Mercatus) does find Medicare for All would reduce NHE by 2 to 4 percent under his "lower-bound" estimate. However, he argues that under a more realistic estimate costs would rise; and his estimates also exclude the cost of universal long-term care.



2 While our estimates of the Buttigieg plan find NHE reductions roughly three times as large as for Biden, all figures are rounded to the nearest percent – reflecting the uncertainty of the estimate