National Health Spending Estimates Under Medicare for All

In 2018, RAND conducted a study for the New York State Health Foundation (NYSHealth) to understand the impacts of the comprehensive single-payer health plan (the New York State Health Act—NYHA) being considered by the New York state legislature (Liu et al., 2018). The RAND study assessed how the plan would affect several outcomes, including health care spending. In 2016, co-author Jodi Liu also analyzed a previous proposal by Sen. Bernie Sanders for a national single-payer plan (S. 1782, 113^th Congress) in the course of completing her dissertation at the Pardee RAND Graduate School (Liu, 2016).

In this research report, we extrapolate from our previous single-payer research, including the work mentioned above, to estimate the effects of a national single-payer health plan (often referred to as Medicare for All) that would provide comprehensive health care coverage to the population nationwide, including long-term care benefits and no cost sharing. The approach is similar to national single-payer health care proposals that have been discussed in Congress, including a recent plan sponsored by Rep. Pramila Jayapal (Medicare for All Congressional Caucus, 2019). We did not model a Medicare for All plan using a microsimulation approach; rather, we estimated aggregate changes to health spending that might occur under the plan by applying adjustments based on our previous work.

Top-Line Findings

We estimate that total health expenditures under a Medicare for All plan that provides comprehensive coverage and long-term care benefits would be $3.89 trillion in 2019 (assuming such a plan was in place for all of the year), or a 1.8 percent increase relative to expenditures under current law. This estimate accounts for a variety of factors including increased demand for health services, changes in payment and prices, and lower administrative costs. We also include a supply constraint that results in unmet demand equal to 50 percent of the new demand. If there were no supply constraint, we estimate that total health expenditures would increase by 9.8 percent to $4.20 trillion.

While the 1.8 percent increase is a relatively small change in national spending, the federal government’s health care spending would increase substantially, rising from $1.09 trillion to $3.50 trillion, an increase of 221 percent.

Spending Category	National Health Expenditure Accounts, 2019 (Status Quo)	RAND Medicare for All Estimate, 2019
Out of pocket	396.9	66.1
Private health insurance	1,278.2	0
Medicare	800.1	0
Medicaid	623.4	0
Other health insurance programs*	148.8	81.0
Other third-party payers**	575.5	506.2
Medicare for All	0	3,238.7
Total National Health Expenditure (in billions)	3,823.1	3,891.9

*Other health insurance programs include CHIP, the Department of Defense, and the Department of Veterans Affairs (VA). We assume that the VA is preserved under the single-payer system.

** Other third-party payers included here are worksite health care; other private revenues; the Indian Health Service; workers’ compensation; general assistance; maternal and child health; vocational rehabilitation; the Substance Abuse and Mental Health Services Administration; other state, local, and federal programs; school health; public health activities; and investment in research, structures, and equipment. We assume that the Indian Health Service, workers’ compensation, private revenues, worksite health care, vocational rehabilitation, school health, government public health activities, and investment would be maintained outside of the single-payer system.

NOTES: Totals may not sum due to rounding; NHEA = National Health Expenditure Accounts.

	National Health Expenditure Accounts, 2019 (Status Quo)	RAND Medicare for All Estimate, 2019 (Federal government does not recoup any state and lcoal spending)
		Federal government*	1,090.0	3,498.7
State and local government**	638.0	0
Business, households, and other private***	2,095.0	393.2
Total National Health Expenditure Projection (in billions)	3,823.1	3,891.9

Federal health spending in the status quo includes spending on Medicaid, Medicare, other programs (maternal and child health, CHIP, vocational rehabilitation, Substance Abuse and Mental Health Services Administration, Indian Health Service, federal workers' compensation, other federal programs, public health activities, Department of Defense, Department of Veterans Affairs, research, structures and equipment, and Marketplace premium tax credits), employer contributions to private health insurance premiums, and employer payroll taxes paid to the Medicare hospital insurance trust fund.

State and local health spending in the status quo includes spending on Medicaid, other programs (state phase-down payments, maternal and child health, public and general assistance, CHIP, vocational rehabilitation, other state and local programs, public health activities, research, and structures and equipment), employer contributions to private health insurance premiums, and employer payroll taxes paid to the Medicare hospital insurance trust fund.

Business, households, and other private spending under Medicare for All includes out-of-pocket spending on services and products not covered by the Medicare for All plan, workers’ compensation reimbursements to the Medicare for All plan, other private revenue (e.g., philanthropy, institutions’ gift shops, cafeterias, parking lots), worksite health care, school health, and private investment in research, structures, and equipment. Tax payments to finance the spending by the federal government are not shown.

NOTES: Totals may not sum due to rounding; NHEA = National Health Expenditure Accounts.

Approach

To estimate aggregate spending at the national level, we started with aggregate health spending as projected by the Centers for Medicare and Medicaid Services, Office of the Actuary (CMS/OACT) for 2019. Those estimates indicate that total U.S. health care spending is projected to be $3,823.1 billion in 2019. Taking the CMS/OACT estimate as a starting point, we made a number of additional adjustments to account for the effects of moving to a Medicare for All plan, described below.

Setting Aside Spending Categories That Would Not Be Affected

We assume the Department of Veterans Affairs and the Indian Health Service would continue to operate independently of a Medicare for All plan, and that the plan would require reimbursement from workers’ compensation carriers. We assume other categories of health spending and offsets that would be outside of the single-payer system include private revenues (e.g., from philanthropy, hospital gift shops, cafeterias, parking), worksite health care, vocational rehabilitation, school health, government public health activities, and investments in research, structures, and equipment. We identified these categories of spending in the tables reported by CMS/OACT and set aside $587.2 billion in spending that we assume would remain unchanged under a Medicare for All plan.

Estimating Changes in Demand for Medical Services

Under a national single-payer system, more people would be covered by insurance, and coverage would be more comprehensive than a typical employer plan or current Medicare coverage—requiring no copays, deductibles, or other cost sharing. The increased availability and generosity of benefits would in all likelihood lead to greater health care consumption. We made separate adjustments to account for changes in spending among people who would become newly insured under a Medicare for All plan, and for increases in spending among previously insured people who would face reduced cost sharing under the plan. To incorporate these adjustments, we assigned 7.2 percent of out-of-pocket spending to currently uninsured individuals, based on estimates published by Catlin et al. (2015). We allocated the remaining 92.8 percent of out-of-pocket spending to those with Medicare and private insurance, and we assume that enrollees with other sources of coverage (i.e., Medicaid, the Children’s Health Insurance Program [CHIP], and other health insurance and third-party payers) would face no out-of-pocket costs¹.

After allocating out-of-pocket spending across categories, we made adjustments to spending by payer to account for increased demand. For the uninsured category, we increased spending by a factor of 25 percent, the change in spending estimated in the Oregon Health Study (Finkelstein et al., 2012). For spending among the currently insured population, we assume that for every 1 percent reduction in out-of-pocket costs, individuals would increase demand by 0.2 percent (Keeler and Rolph, 1988). We applied these elasticities by payer to capture the fact that, on average, payers have different cost sharing in the status quo.

Using this approach, we estimate that under a Medicare for All plan, demand for services currently paid by Medicare would increase by 2.2 percent, demand for services currently paid by commercial insurance would increase by 2.6 percent, and demand for services currently paid for by uninsured individuals would increase by 25 percent². We assume that demand for services currently covered by Medicaid and other public insurance programs (such as CHIP) would remain unchanged because these programs have little to no cost sharing in the status quo.

Estimating Changes in Demand for Long-Term Care

We assume the plan would cover benefits for long-term care services and supports with no cost sharing. As with the demand for medical services, this would almost certainly lead to greater consumption of services. We made adjustments to account for changes in spending based on estimates derived from per capita long-term care costs and assumptions about the demand for noninstitutional long-term services and supports (e.g., home- and community-based services) and institutional care (e.g., nursing home care). Hurd et al. (2013) estimated the cost of long-term care for people over age 70 with dementia, finding that Americans spend $27,789 annually per person on informal home care (with caregiving time valued by replacement cost), compared with $5,678 for formal home care and $13,876 for nursing home care (net of Medicare and out-of-pocket spending). Following the assumptions that we used in our work on the NYHA (Liu et al., 2018), we assume that half of all informal home care costs are redirected into formal care. Of the half of informal care that shifts to formal care, we assume that 90 percent shifts to formal home care and 10 percent shifts to nursing home care. Applying these assumptions to the Hurd et al. cost estimates (with adjustments for Medicare and out-of-pocket spending), we derive a 200 percent increase in formal home care cost, and a 10 percent increase in nursing home costs.

Estimating Changes in Payment for Services

The Jayapal plan would set a global budget for health services that would be negotiated annually with CMS regional offices. We do not yet know how this budget would be set, nor the specifics of how payment would change. In our estimates, we assume that the CMS regional offices would establish a uniform set of payment rates for comparable services and that, in the initial years of operation, payment levels for medical services would resemble average payment under the status quo. We derived this all-payer average rate using these estimates from the prior literature:

• Commercial insurers pay 167 percent of the Medicare rate for hospital services and 125 percent of the Medicare rate for physician services (American Hospital Association, 2018; Centers for Medicare and Medicaid Services, 2012).
• Medicare and Medicaid pay roughly the same rates for hospital care, and Medicaid reimburses physicians at 72 percent of the Medicare payment, on average (Kaiser Family Foundation, 2016).

Given these assumptions, we estimate that the all-payer average rate under the Medicare for All plan in 2019 would be 124 percent of current Medicare rates for hospital payments and 107 percent of current Medicare rates for physician payment. In the absence of better data, we assume that reimbursement for other services, such as dental services, home health, and nursing care would resemble reimbursement for physician services. Across all provider types, these estimates yield a blended average payment rate of 109 percent of current Medicare rates³. In general, these assumptions imply that payment would increase for services currently covered by public payers and decrease for services currently covered by private payers.

Accounting for Supply Constraints for Services

Although, on average, we assume payment levels would be the same as they are in the status quo in 2019, providers’ willingness and ability to provide health care services—including the additional care required by the newly insured and those benefiting from lower cost sharing—would most likely be limited. The extent and distribution of unmet care would depend on providers’ payer mix under current law and their responses to Medicare for All payment levels. For example, some providers may elect to not participate in a Medicare for All plan (and instead enter in private contracts with individuals, an arrangement permitted in some single-payer bills), providers may alter when they retire, and potential medical students and trainees could change their career choices. As a result, some patients might experience longer wait times for care or face unmet needs.

RAND’s Health Care Payment and Delivery Simulation Model (PADSIM) accounts for how providers’ supply of services might respond to changes in payment (White et al., 2016). In our analysis of the NYHA, which included similar assumptions to those described above regarding how payment would change under a single-payer plan, PADSIM estimated that the supply of physician and hospital services would adjust such that unmet demand would approximately equal 50 percent of the new demand. We apply this 50 percent factor to our utilization estimates for 2019 to account for supply constraints in the current analysis. In a sensitivity analysis, we considered how the results would change if this supply constraint were alleviated.

Estimating Changes in Prices for Drugs and Devices

U.S. payers are currently charged substantially more for drugs than payers in other countries (Danzon and Furukawa, 2008), which may be partially attributable to the fact that Medicare is currently barred from negotiating drug prices on behalf of enrollees. Within the United States, drug prices vary by payer; currently, drugs obtained by the Medicaid program are governed by a “best price” rule that requires manufactures to sell drugs to the program at the lowest price offered to any payer (Baghdadi, 2017). Given this drug pricing rule, Medicaid drug prices have been estimated to be roughly 66 percent of Medicare prices and 54 percent of commercial prices (Roehrig, 2018; Gagnon and Wolfe, 2015; Government Accountability Office, 2014; Cook, 2013).

Consistent with our previous research on the NYHA, we assume that under a Medicare for All plan, the government would be able to negotiate drug and device prices that are 10 percent below current Medicare prices in the initial year; we do not assume they would be able to be reduced further, such as to the same level as current prices under the Medicaid program. For reference, estimates from a Congressional Budget Office analysis of a budget option for Medicare Part D Low-Income Subsidy rebates, which would be similar to the Medicaid 23.1 percent rebates, indicate that spending would decrease by 22 percent in 2019 (Congressional Budget Office, 2016; Congressional Budget Office, 2017). Although, it is possible that, over time, negotiated prices could reach levels comparable with those currently paid by Medicaid (or to levels found in other Organisation for Economic Cooperation and Development [OECD] countries), we assume it would be difficult for a national single-payer system to reach Medicaid price levels for the entire U.S. population initially.

Estimating Changes in Administrative Costs

The costs of administering a single-payer plan might be lower than current administrative costs across the health care system given such factors as economies of scale, the elimination of marketing expenses and the need to set premiums, and a reduced need to manage actuarial risk. According to the estimates from CMS/OACT, health plan administrative spending is lower in public health insurance programs than it is for private sources of coverage—for example, administrative spending in 2017 was 6.9 percent of personal health care spending on Medicare (for administration of traditional Medicare and privately managed Medicare Advantage plans) and 13.9 percent of personal health care spending for private health insurance. We assume that a Medicare for All plan would have administrative spending at 6.0 percent of personal health care spending, which is approximately 1 percent lower than the blended average of traditional Medicare and Medicare Advantage administrative rates. As a result of this assumption, we estimate that average health plan administrative spending would fall from 7.9 percent of total spending in the status quo to 5.3 percent of total spending under Medicare for All⁴.

Additionally, provider administrative spending may fall under a single-payer system because of reduced billing and claims processing costs. We assume that billing and insurance-related administrative costs account for 13 percent of physician expenditures, 8.5 percent of hospital expenditures, and 10 percent of other expenditures in the status quo (Jiwani et al., 2014). Based on an analysis by Pozen and Cutler (2010) that compared provider administrative spending in the Canadian and U.S. health care systems, we estimate that these administrative expenses would decline by 32.6 percent, falling to 5.6 percent of total spending under a Medicare for All plan.

Accounting for Out-of-Pocket Costs

Although most recent Medicare for All bills would provide comprehensive coverage with no cost sharing, we assume that an additional 2 percent of current spending would remain outside of the system. This accounts for over-the-counter medications that would not be covered, as well as elective services such as cosmetic surgeries, infertility treatment, and adult orthodontics. This uncovered spending accounts for $66.6 billion, bringing the total amount of spending that would remain unchanged to $653.7 billion.

Calculating Changes in Government and Private Spending

We estimate sources of payment in the status quo using National Health Expenditure Accounts (NHEA) Projected Table 16, which presents estimated spending by the federal government, state and local governments, and private businesses and households. We assume that the federal government would bear the entire cost of the Medicare for All plan. To estimate how other sources of payment change under the bill, we allocate the remaining $653.7 billion in funds that are not covered by the Medicare for All plan across sponsors by assuming that private households bear the entire cost of out-of-pocket payments, and that the federal government bears the entire cost of the Department of Veterans Affairs and the Indian Health Service. We allocate the costs of government public health activities and investment in research, structures, and equipment based on historical patterns. In sensitivity analyses, we consider the possibility that the federal government can recoup state and local financing for the Medicaid program, estimated at $244 billion in 2019, to help finance the Medicare for All plan.

Limitations

We extrapolated these estimates based on findings from our previous analyses of single-payer plans without rerunning our simulation approach. It is possible that a more nuanced analysis of the bill would yield different conclusions. For example, the provider payment response that we have estimated is derived from our analysis of the NYHA, and the response at the national level may be different given differences in average payment rates in the status quo. On the one hand, New York has more physicians per capita than the rest of the country (Kaiser Family Foundation, 2019); larger supply constraints may exist for the wider United States than we estimated in the NYHA study. On the other hand, the all-payer blended average payment rate that we estimated for New York was lower relative to Medicare than U.S. blended rates applied in this analysis, a factor that could lessen the effect of the supply constraint at the national level.

An additional point of uncertainty is how service utilization among the currently uninsured population would respond to the single-payer plan. We have estimated this change by applying the change in utilization found in the Oregon Health Insurance Experiment to the portion of out-of-pocket spending that is attributed to uninsured individuals. To the extent that health care utilization among the uninsured population is covered by other sources of payment, such as charity care, this approach may underestimate the change in utilization. However, the Oregon study estimated the change in utilization among a specific population of low-income individuals who attempted to gain access to Medicaid through a lottery. These individuals are demographically different from the general U.S. population, and the fact that they signed up for the Medicaid lottery could indicate that they have higher demand for health care than the general population (Finkelstein et al., 2012).

In addition, there are many uncertainties regarding how a Medicare for All plan would be implemented. The Jayapal plan would create a regional governance structure to set a global budget for health care services; this budget could have significant implications for the cost of the bill. In our analysis, we assumed a provider payment schedule that effectively compensates providers at 124 percent of Medicare rates for hospital services and at 107 percent of Medicare rates for physician and other services. However, in practice, a global budget might be set very differently, affecting both the total health spending and—potentially—providers’ willingness to supply services. The effects on drug spending are also unclear. The plan also requires the Secretary of Health and Human Services to negotiate drug prices with manufacturers; while it is likely that this negotiation would lead to reduced spending, the magnitude of the reduction could be larger or smaller than we have estimated.

There is also significant uncertainty regarding how much administrative spending might change under a single-payer plan. Our estimates assume the overall spending on health plan administration falls to rates that are slightly below the rate currently found in the federal Medicare program, including both traditional Medicare and Medicare Advantage plans, which are run by private insurers. On the one hand, savings could be even greater if the national single-payer plan is able to achieve administrative spending levels comparable with traditional Medicare, which are lower than rates found in Medicare Advantage. On the other hand, administrative spending might be higher than we have estimated if the Medicare for All plan requires new governance structures, such as the regional governance structure included in the Jayapal plan, and if these structures add complexity to the system. Similarly, we assume that provider administrative rates would fall to reflect changes in claims processing and billing costs. The magnitude of the reduction we applied to provider costs is based on a study by Pozen and Cutler (2010), which compared U.S. costs with those in Canada. We have not accounted for the possibility that reductions in claims processing reduces the need for other provider costs (for example, a reduced need for office space, leading to savings on building costs), which could further reduce provider administration costs overall. However, it is also possible that providers could face new quality or administrative reporting requirements under the single-payer system, which could add to spending. Utilization management procedures might also change under the Medicare for All plan, with uncertain effects for both provider and health plan administrative costs.

Finally, our estimates assume a hypothetical scenario in which the plan is fully implemented in 2019. This is an unrealistic assumption: 2019 has already begun, and the plan could not be implemented retroactively. Moreover, the bill would involve a massive reorientation of the U.S. health system and would need to be phased in over several years. The Jayapal bill includes a two-year transition period; however, a longer time may be required to enable consumers, providers, and regulators to fully adjust to this substantial change.

Detailed Results

Table 1 provides a breakdown of our estimates of the changes in 2019 spending under a Medicare for All plan—again, assuming such a plan were in place for all of 2019. Overall, we estimate that health care spending would increase by 1.8 percent relative to expenditures under current law, from $3,823.1 billion to $3,891.9 billion.

Spending Category	National Health Expenditure Accounts, 2019 (Status Quo)	RAND Medicare for All Estimate, 2019
Out of pocket	396.9	66.1
Private health insurance	1,278.2	0
Medicare	800.1	0
Medicaid	623.4	0
Other health insurance programs*	148.8	81.0
Other third-party payers**	575.5	506.2
Medicare for All	0	3,238.7
Total National Health Expenditure	3,823.1	3,891.9

*Other health insurance programs include CHIP, the Department of Defense, and the Department of Veterans Affairs (VA). We assume that the VA is preserved under the single-payer system.

** Other third-party payers included here are worksite health care; other private revenues; the Indian Health Service; workers’ compensation; general assistance; maternal and child health; vocational rehabilitation; the Substance Abuse and Mental Health Services Administration; other state, local, and federal programs; school health; public health activities; and investment in research, structures, and equipment. We assume that the Indian Health Service, workers’ compensation, private revenues, worksite health care, vocational rehabilitation, school health, government public health activities, and investment would be maintained outside of the single-payer system.

NOTE: Totals may not sum due to rounding.

In Table 2, we present our estimates of spending within various categories under a Medicare for All plan. Based on these assumptions, we estimate that total prescription drug spending would fall from $360.4 billion to $321.2 billion, a reduction of 10.9 percent. For overall personal health care spending, we estimate that the net change due to utilization is from $2,962.8 billion in the status quo to $3,190.5 billion under a Medicare for All plan, or a 7.7 percent increase⁵. Administrative spending falls from $580.8 billion to $422.1 billion, a reduction of 27.3 percent. About 61 percent of this reduction comes from reduction in health plan administrative spending, and the remainder comes from a reduction in provider administrative spending. Together, health plan and provider administrative spending account for roughly 15.2 percent of total health spending in the status quo and 10.8 percent of total spending under a Medicare for All plan.

	National Health Expenditure Accounts, 2019 (Status Quo)	RAND Medicare for All Estimate, 2019
Personal Health Care	2,962.8	3,190.5
Hospital care	1,148.0	1,151.2
Physician services	667.8	673.1
Prescription drugs	360.4	321.2
Durable medical equipment	60.8	58.7
Other professional services and dental services	221.3	228.5
Home health and other health, residential, and personal care	275.0	519.6
Nursing care and continuing retirement support facilities	160.2	167.5
Other non-durable medical products	69.3	70.7
Administrative Costs	580.8	422.1
Health plan administration*	301.4	204.4
Provider administration**	279.4	217.7
Public Health Activities and Investments	279.1	279.1
Total National Health Expenditure Projection	3,823.1	3,891.9

*Health plan administration includes government administration and the net cost of health insurance.

** In the National Health Expenditure Accounts, provider administrative expenditures are included in expenditures for hospital care, physician services, other professional services, dental services, and nursing care and continuing retirement support facilities. In this table, we include provider administrative expenditures as a separate line item from the health care service categories.

NOTES: We assume that the Department of Veterans Affairs, Indian Health Service, workers’ compensation, private revenues, worksite health care, vocational rehabilitation, school health, government public health activities, and investment would be maintained outside of the single-payer system. Totals may not sum due to rounding.

Table 3 shows our estimated changes in spending by the federal government, state and local government, and private sources. We estimate that federal health care spending would increase by 221 percent under Medicare for All relative to current law, from $1,090.0 billion to $3,498.7 billion. Private sources of spending would decrease by 81 percent, from $2,095.0 billion to $393.2 billion. However, these declines in health care payments by private sources would be replaced by new taxes, which would be needed to fund the increases in federal spending. If the federal government can recoup state Medicaid spending under current law and redirect the funds to the Medicare for All plan, then federal spending would increase by 199 percent.

	National Health Expenditure Accounts, 2019 (Status Quo)	RAND Medicare for All Estimate, 2019
		Assuming federal government does not recoup any state and local spending	Assuming federal government recoups state and local Medicaid spending
Federal government*	1,090.0	3,498.7	3,254.7
State and local government**	638.0	0	244.0
Business, household, and other private***	2,095.0	393.2	393.2
Total National Health Expenditure Projection	3,823.1	3,891.9	3,891.9

*Federal health spending in the status quo includes spending on Medicaid, Medicare, other programs (maternal and child health, CHIP, vocational rehabilitation, Substance Abuse and Mental Health Services Administration, Indian Health Service, federal workers' compensation, other federal programs, public health activities, Department of Defense, Department of Veterans Affairs, research, structures and equipment, and Marketplace premium tax credits), employer contributions to private health insurance premiums, and employer payroll taxes paid to Medicare hospital insurance trust fund.

** State and local health spending in the status quo includes spending on Medicaid, other programs (state phase-down payments, maternal and child health, public and general assistance, CHIP, vocational rehabilitation, other state and local programs, public health activities, research, and structures and equipment), employer contributions to private health insurance premiums, and employer payroll taxes paid to Medicare hospital insurance trust fund.

*** Business, household, and other private spending in the status quo includes employer contributions to private health insurance premiums, Medicare hospital insurance payroll taxes, self-employment payroll taxes, temporary disability insurance, workers’ compensation, and worksite health care; household private health insurance premiums, Medicare payroll taxes and premiums, and out-of-pocket spending; and other private revenues (e.g., philanthropy, non-operating revenue, investment income, and privately funded structures and equipment). Business, household, and other private spending under Medicare for All includes workers’ compensation reimbursements to the Medicare for All plan; other private revenue; worksite health care; school health; private investment in research, structures, and equipment; and out-of-pocket spending on services and products not covered by the Medicare for All plan. Tax payments to finance the spending by the federal government are not shown.

NOTE: Totals may not sum due to rounding.

The results above assume that overall demand for health care services that is unmet equals 50 percent of the new demand for services, potentially leading to longer wait times or decreased access to care. In Table 4, we consider the effects on spending if this supply constraint is lifted and people are able to get all the care that they demand. In this case, total health spending increases to $4,197.1 billion, a 9.8 percent increase over national spending in the status quo⁶. These results illustrate an important trade-off between cost and access—higher levels of total spending would be needed to ensure that all demand for health care is met. Although our baseline results suggest that the plan could be implemented with a 1.8 percent increase in total spending, this result is heavily dependent on the assumption that unfulfilled demand for health care services is approximately half of the new demand. If all demand is fully met, spending increases substantially.

Spending Category	National Health Expenditure Accounts, 2019 (Status Quo)	RAND Medicare for All Estimate with No Supply Constraints, 2019
Out of pocket	396.9	72.2
Private health insurance	1,278.2	0
Medicare	800.1	0
Medicaid	623.4	0
Other health insurance programs*	148.8	81.0
Other third-party payers**	575.5	506.2
Medicare for All	0	3,537.8
Total National Health Expenditure Projection	3,823.1	4,197.1

*Other health insurance programs include CHIP, the Department of Defense, and the Department of Veterans Affairs (VA). We assume that the VA is preserved under the single-payer system.

** Other third-party payers here include worksite health care; other private revenues; the Indian Health Service; workers’ compensation; general assistance; maternal and child health; vocational rehabilitation; the Substance Abuse and Mental Health Services Administration; other state, local, and federal programs; school health; public health activity; and investment in research, structures, and equipment.

NOTE: Totals may not sum due to rounding.

Notes

• [1] For other nondurable medical products (nonprescription drugs such as over-the-counter medications and medical sundries), we allocate 7.2 percent of spending to the uninsured population and the remaining to all payers in proportion to total expenditures by each payer. In the status quo, the vast majority of nondurable medical products are paid for out of pocket, with the exception of some expenditures paid by Medicare. We assume that nondurable medical products currently covered by Medicare would be covered for everyone under a Medicare for All plan, and the remainder would continue to be paid for out of pocket.
• [2] These calculations exclude changes in demand for long-term care (including home health and nursing home care), which are handled differently as described later in this report.
• [3] Across hospitals and physicians, we estimate a blended payment rate of 116 percent of current Medicare rates.
• [4] An alternative denominator for this calculation would be all personal health care spending plus administrative spending (including health plan and provider administration), which is smaller than total health spending. With this alternative denominator, the administrative rate would fall from 8.5 percent in the status quo to 5.7 percent under a Medicare for All plan.
• [5] Personal health care includes hospital care; physician services; prescription drugs; durable medical equipment; other professional services; dental services; home health care; other health, residential, and personal care; nursing care and continuing retirement support facilities; and other nondurable medical products. The net change due to utilization reflects changes in demand, payment rates and prices, and supply constraints; we exclude provider administrative spending here.
• [6] Without supply constraints, we estimate that spending due to the net change in utilization of personal health care is $3,458.1 billion under a Medicare for All plan, or a 16.7 percent increase from the status quo of $2,962.8 billion.

References

• American Hospital Association, Trendwatch Chartbook 2018, Appendix 4: Supplementary Data Tables, Trends in Hospital Financing, June 2018. As of March 29, 2019: https://www.aha.org/system/files/2018-06/2018-AHA-TrendWatch-Chartbook-Appendix-4.pdf.
• Baghdadi, Ramsey, “Health Policy Brief: Medicaid Best Price,” Health Affairs, August 10, 2017, doi: 10.1377/hpb20171008.000173.
• Catlin, Mary K., John A. Poisal, and Cathy A. Cowan, “Out-of-Pocket Health Care Expenditures, by Insurance Status, 2007-10,” Health Affairs, Vol. 34, No. 1, 2015, pp. 111–116, doi: 10.1377/hlthaff.2014.0422.
• Centers for Medicare and Medicaid Services, Annual Report of the Board of Trustees of the Federal Hospital Insurance and Federal Supplementary Medical Insurance Trust Funds, 2012. As of March 29, 2019: https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/ReportsTrustFunds/downloads/tr2012.pdf.
• Congressional Budget Office, “Options for Reducing the Deficit: 2017 to 2026—Require Manufacturers to Pay a Minimum Rebate on Drugs Covered Under Part D of Medicare for Low-Income Beneficiaries,” December 8, 2016. As of June 27, 2018: https://www.cbo.gov/budget-options/2016/52239
• Congressional Budget Office, “Medicare—Congressional Budget Office’s June 2017 Baseline,” June 29, 2017.As of June 27, 2018: https://www.cbo.gov/sites/default/files/recurringdata/51302-2017-06-medicare.pdf
• Cook, Anna, “Costs Under Medicare’s Prescription Drug Benefit and a Comparison with the Cost of Drugs Under Medicaid Fee-for-Service,” briefing slides, Congressional Budget Office, June 23, 2013. As of March 29, 2019: https://www.cbo.gov/sites/default/files/presentation/44366academyhealthpresentationcook0.pdf.
• Danzon, Patricia M., and Michael F. Furukawa, “International Prices and Availability of Pharmaceuticals in 2005,” Health Affairs, Vol. 27, No. 1, 2008, pp. 221–233, doi:10.1377/hlthaff.27.1.221.
• Finkelstein, Amy, Sarah Taubman, Bill Wright, Mira Bernstein, Jonathan Gruber, Joseph P. Newhouse, Heidi Allen, Katherine Baicker, and the Oregon Health Study Group, "The Oregon Health Insurance Experiment: Evidence from the First Year," Quarterly Journal of Economics, Aug 2012; 127(3): 1057-1106.
• Gagnon, Marc-Andre, and Sidney Wolfe, Mirror, Mirror on the Wall: Medicare Part D Pays Needlessly High Brand-Name Drug Prices Compared with Other OECD Countries and with U.S. Government Programs, July 23, 2015. As of June 27, 2018: https://carleton.ca/sppa/wp-content/uploads/Mirror-Mirror-Medicare-Part-D-Released.pdf.
• Government Accountability Office, Prescription Drugs: Comparison of DOD, Medicaid, and Medicare Part D Retail Reimbursement Prices, Washington, D.C., June 2014. As of June 27, 2018: https://www.gao.gov/assets/670/664521.pdf.
• Hurd, Michael D., Paco Martorell, Adeline Delavande, Kathleen L. Mullen, and Kenneth M. Langa, “Monetary Costs of Dementia in the United States,” New England Journal of Medicine, Vol. 368, 2013, pp. 1326–1334.
• Jiwani, Aliya, David Himmelstein, Steffie Woolhandler, and James G. Kahn, “Billing and Insurance-Related Administrative Costs in United States’ Health Care: Synthesis of Micro-Costing Evidence,” Health Services Research, Vol. 14, 2014, doi:10.1186/s12913-014-0556-7.
• Kaiser Family Foundation, Medicaid-to-Medicare Fee Index, 2016. As of March 29, 2019: https://www.kff.org/medicaid/state-indicator/medicaid-to-medicare-fee-index/.
• Kaiser Family Foundation, Professional Active Physicians, 2019. As of March 29, 2019: https://www.kff.org/other/state-indicator/total-active-physicians/.
• Keeler, Emmett B., and John E. Rolph, “The Demand for Episodes of Treatment in the Health Insurance Experiment,” Journal of Health Economics, Vol. 7, No. 4, 1988, pp. 337–367.
• Liu, Jodi L., Chapin White, Sarah A. Nowak, Asa Wilks, Jamie Ryan, and Christine Eibner, An Assessment of the New York Health Act, Santa Monica, Calif.: RAND Corporation, RR-2424-NYSHF, 2018. As of March 29, 2019: https://www.rand.org/pubs/research_reports/RR2424.html.
• Liu, Jodi L., Exploring Single-Payer Alternatives for Heath Care Reform, Santa Monica, Calif.: RAND Corporation, RGSD-375, 2016. As of March 29, 2019: https://www.rand.org/content/dam/rand/pubs/rgs_dissertations/RGSD300/RGSD375/RAND_RGSD375.pdf.
• Medicare for All Congressional Caucus, Medicare for All Act of 2019 Section-by-Section, February 25, 2019. As of March 29, 2019: https://jayapal.house.gov/wp-content/uploads/2019/02/Medicare-for-All-Act-of-2019-Section-by-Section-Summary.pdf.
• Pozen, Alexis, and David M. Cutler, “Medical Spending Differences in the United States and Canada: The Role of Prices, Procedures, and Administrative Expenses,” Inquiry: Journal of Health Care Organization, Provision, and Financing, Vol. 47, No. 2, 2010, pp. 124–134, doi:10.5034/inquiryjrnl_47.02.124.
• Roehrig, Charles, “Rebates, Coupons, PBMs, and the Cost of the Prescription Drug Benefit,” Health Affairs Blog, April 26, 2018. As of June 27, 2018: https://www.healthaffairs.org/do/10.1377/hblog20180424.17957/full/.
• White, Chapin, Jodi L. Liu, Mikhail Zaydman, Sarah A. Nowak, and Peter S. Hussey, The RAND Health Care Payment and Delivery Simulation Model (PADSIM): Concepts,Methods, and Examples, Santa Monica, Calif.: RAND Corporation, RR-1428-RC, 2016. As of June 27, 2018: https://www.rand.org/pubs/research_reports/RR1428.html.

Related Content

• 

  Report

  An Assessment of the New York Health Act: A Single-Payer Option for New York State

  Aug 1, 2018

• 

  Research Brief

  Estimating the Effects of a Single-Payer Proposal in New York State

  Aug 1, 2018

Topics

• Health Care Financing
• Health Care Reform
• Health Insurance
• United States