Urban Institute response to attacks on their analysis of the Sanders Single Payer plan

[ehrnst]

The Sanders campaign and David Himmelstein and Steffie Woolhandler reacted with sharp criticisms to our recent report, The Sanders Single-Payer Health Care Plan: The Effect on National Health Expenditures and Federal and Private Spending (Holahan et al. 2016). The campaign argues that we understated reductions in the cost of prescription drugs, understated administrative cost savings, and ignored the availability of state and local funds to finance the plan. Himmelstein and Woolhandler (hereinafter referred to as HW) argue that our “ridiculous assumptions yield ridiculous estimates”; specifically, they argue that we overstated administrative costs, ignored administrative cost savings for providers, understated reductions in drug spending, and overstated utilization increases.1



Our analysis was based on detailed modeling of acute care for the nonelderly, acute care for the elderly, and long-term care services and supports. It is impossible to wholly impose a new health care system in the United States that changes the way all residents receive and finance their health care, even one that may be successful in another country, without disrupting many existing institutions, such as insurance companies, integrated health systems, hospitals, physicians, and pharmaceutical manufacturers. To be politically acceptable, compromises would have to be made, and those compromises are reflected in our assumptions.


In this brief, we discuss our key assumptions in these areas of disagreement and highlight ways in which we may have actually underestimated overall costs of the Sanders proposal. By and large our assumptions are laid out thoroughly in the original paper, but here we use them to address the specific statements made by the campaign and HW, and we provide additional reliable evidence to counter some of HW’s claims.

The increases in federal spending that we estimated ($32 trillion between 2017 and 2026) are so large because all current public and private spending would be transferred to the federal government, benefits would be expanded, and out-of-pocket costs to consumers would be eliminated.

Payment rates would have to be acceptable to providers. We assume a substantial reduction from current rates paid by private insurers and some increases over current rates paid by public programs. For example, the program would pay 25 percent less than current Medicare levels for prescription drugs, and physicians would be paid at Medicare rates. Both of these are increases relative to current Medicaid payment rates.

Utilization of health care services will increase if benefits are expanded and cost sharing is eliminated. Our estimates include modest increases in the use of services based on actuarial standards and the health economics literature. Contrary to HW’s claim in their article, health care use and spending for the elderly population did increase substantially once the Medicare program was implemented in 1965.

We assume administrative costs of 6 percent. A new system would have a host of important administrative functions necessary to effective operations, such as rate setting for many different providers of different types; quality control over care provision; development, review, and revision of regulations; provider oversight and enforcement of standards; bill payment to providers; and other functions. We base our administrative cost estimates on Medicare’s costs to administer the entire Medicare program. But even if we have modestly overestimated the appropriate administrative load, the difference in costs for the federal government would be only about 1 percent of total added federal spending per percentage-point reduction, a tiny fraction of the additional $32 trillion in federal funding that we estimate would be needed to fully finance the Sanders health plan.

We provide estimates of current state and local spending on health care through the Medicaid program and on payments for uncompensated care. Requiring state governments to give the amount they currently spend on Medicaid to the federal government to help finance the single-payer system is of very uncertain legality given the Supreme Court’s ruling in National Federation of Independent Business v. Sebelius.

We may have underestimated the costs of the Sanders plan in several important areas. These were described in the original paper and are summarized below.

https://www.urban.org/research/publication/response-criticisms-our-analysis-sanders-health-care-reform-plan

[TCJ70]

1. Hmmm...$3.2 Trillion per year? Less than we spend now and not everyone is covered...

...sounds like a deal to me.

http://www.pbs.org/newshour/rundown/new-peak-us-health-care-spending-10345-per-person/

[ehrnst]

2. You are still smarting about this, aren't you?

By all means, please share a sourcewatch page that copies and pastes from Wikipedia that "proves" they are in the pocket of charter schools and CIGNA....

[TCJ70]

6. Ummm...no. I didn't wade into the whole...

...sourcewatch garbage. What would I be smarting from? That fact that they had a typo that you said was impossible for them have?

I'm not even disputing the findings here (typos excluded). I'm only saying that this doesn't paint a bad picture. If anything, it indicates a cheaper, more efficient system with a single-payer environment.

[ehrnst]

7. So the fact that the analysis says that disruption to health care would be due to the short time

of implementation, and that the taxes needed to pay for the plan are underestimated (which was one of the things that killed the Vermont single payer program) indicate "Cheaper, more efficient" system to you?

Why?

[TCJ70]

8. They should put their worst case numbers out there...

...if they're really concerned about it. Why put out a number ($32 Trillion) if you "know" you're shorting the estimates? Just put out the numbers.

Disruption doesn't bother me because I expect it. That's not at all shocking. There's going to be disruption even if a public option is implemented so who cares?

[NurseJackie]

3. TMYK ★ ...

[ehrnst]

5. Confirmation bias.... it's a problem.

[ismnotwasm]

4. K&R

And this is what I keep saying right here

Our estimates include reasonable increases in the use of services (and therefore health care spending) based on actuarial standards and the health economics literature (Buettgens 2011). In addition, we incorporate an assumption that not all the increased demand for health care would be met under the Sanders plan, at least in the short run, because of constraints in the supply of health care providers. The supply constraints implicit in our estimates are consistent with those experienced by enrollees in the current-law Medicaid program. Because of Medicaid’s historically low payment rates to providers relative to private insurers and the Medicare program, empirical analyses have demonstrated that Medicaid beneficiaries use less care, particularly for specialists, than they would without such supply constraints. Although increasing the supply of physicians in the short run is challenging, changes to work hours, the use of physician extenders (such as nurse practitioners and physician assistants), and the hiring of more foreign-trained physicians can help. Moreover, the federal government could consider further subsidizing medical education to increase the supply of needed physicians over time; however, we have not estimated the cost of doing so.