Himmelstein DU, Woolhandler S. Cost control in a parallel universe: Medicare spending in the United States and Canada.. Arch Intern Med.. Published online October 29, 2012. doi:10.1001/2013.jamainternmed.272.
eAppendix. Methods: cost control in a parallel universe: Medicare spending in the U.S. and Canada
eFigure: Change since 1971 in inflation-adjusted Medicare spending per person over age 64, U.S. and Canada
eReferences: Bibliographic sources
Himmelstein DU, Woolhandler S. Cost Control in a Parallel Universe: Medicare Spending in the United States and Canada. Arch Intern Med. 2012;172(22):1764–1766. doi:10.1001/2013.jamainternmed.272
As the United States was implementing Medicare in 1966, Canada was phasing in its own Medicare program, which covered all Canadians under provincially administered plans. While these provincial plans varied, all incorporated significant payment reforms—global budgeting of hospitals and stringent capital expenditure controls—and banned copayments and deductibles.
Before the mid-1960s, the 2 nations' health care financing systems were similar, and health care costs were comparable.1 Since then, overall US costs have grown more rapidly, but no study has compared spending for the elderly—the populations covered by Medicare in both nations.
We obtained official figures for Medicare spending for persons older than 64 years in Canada and the United States for 1971 (when Canadian Medicare became fully operational) through 2009. Since available Canadian data for 1971 through 1979 are less detailed, we focus principally on changes since 1980.
We adjusted Canadian figures for minor changes in government accounting. To avoid distorting time trends, we excluded Medicare Part D (which began in 2006).
We calculated percentage changes in inflation-adjusted spending per elderly enrollee and compared actual US Medicare expenditures in each year since 1980 (and 1971) with the projected level of expenditure had US Medicare spending increased at Canada's rate. See the eAppendix for further details.
US Medicare spending per elderly enrollee rose from $1215 in 1980 to $9446 in 2009 (an inflation-adjusted 198.7% increase). The comparable increase for Canada was 73.0% (from $2141 to $9292). Canada's higher base-year spending reflects its more comprehensive benefits, covering about 80% of seniors' total health costs, vs about 50% in US Medicare.
The Table lists actual US Medicare spending from 1980 through 2009 and projected spending and savings had US costs risen at the lower Canadian rate. Projected savings totaled $154.2 billion in 2009 and $2.156 trillion for 1980 through 2009.
Medicare hospital spending per elderly enrollee grew 44.7% in Canada vs 81.9% in the United States. Physician spending grew 100.7% in Canada vs 274.3% in the United States. Hospitals' share of total Medicare spending fell from 49.6% to 41.5% in Canada and from 68.4% to 41.5% in the United States. Spending for other services (eg, home, hospice, and skilled nursing facility care) rose from 3.9% to 23.6% of spending in the United States and from 39.7% to 44.3% in Canada.
For the 1971-2009 period, US costs rose 374.1% vs 126.3% for Canada, and estimated foregone savings were $2.9024 trillion (eFigure).
Medicare spending has grown nearly 3 times faster in the United States than in Canada since 1980. Had US Medicare costs risen at Canadian rates, rather than a deficit of $17.1 billion in 2009, the Medicare Hospital Trust Fund would have realized a $32.3 billion surplus. Savings on Medicare Part B would have been even larger. By 2009, the $2.156 trillion in excess spending attributable to US Medicare's faster growth was equivalent to more than one-sixth of the national debt.
Several features of Canada's program help constrain costs. First, the single-payer system has simplified administration, holding administrative costs to 16.7% of overall spending vs 31.0% in the United States.2 Although US Medicare's internal overhead costs are low, it remains one among many payers. Hence providers' administrative costs are inflated by having to deal with a multitude of payers and track eligibility, attribute costs, and bill for individual patients and services.
Second, Canadian hospitals receive prospectively determined global operating budgets, removing incentives to provide unnecessary care while simplifying billing and administration. However, unlike accountable care organization payment schemes in the United States, capital costs are not folded into the global budgets but distributed separately through an explicit health-planning process. Canadian hospitals cannot use operating surpluses to fund new buildings or equipment but must request separate capital appropriations. Hence, they cannot expand by overproviding lucrative services, gaming the payment system through upcoding, avoiding unprofitable patients, or cost shifting.
Third, 51% of Canada's physicians are primary care practitioners vs 32% in the United States.3 Primary care–centered health systems are generally thriftier.4 Canada's outpatient fee schedules are also less technology skewed than in the United States.
Fourth, Canada's provincial plans have used their concentrated purchasing power to limit drug and device prices.
Finally, litigation and malpractice costs have remained relatively low in Canada.
Life expectancy at age 65 years is longer and has grown faster in Canada than in the United States since 1980 (and 1971),5 offering reassurance that cost control has not compromised quality. A meta-analysis suggests that clinical outcomes are, if anything, better for Canadians than for insured Americans.6
To some, US Medicare's grim financial health suggests an even grimmer conclusion: it can no longer keep its promise of all needed care for the elderly population.7 Some would replace it with vouchers that seniors could use to purchase private coverage. Others suggest upending the current payment system by inverting volume-based incentives, offering instead profits to organizations that limit utilization. Yet the efficacy of these drastic solutions remains unproven.8 Canada's road-tested cost-containment methods offer an alternative.
Correspondence: Dr Woolhandler, School of Urban Public Health at Hunter College, City University of New York, 255 W 90th St, Apt 12A, New York, NY 10024-1152 (firstname.lastname@example.org).
Published Online: October 29, 2012. doi:10.1001/2013.jamainternmed.272
Conflict of Interest Disclosures: None reported.
Author Contributions:Study concept and design: Himmelstein and Woolhandler. Acquisition of data: Himmelstein and Woolhandler. Analysis and interpretation of data: Himmelstein and Woolhandler. Drafting of the manuscript: Himmelstein and Woolhandler. Critical revision of the manuscript for important intellectual content: Himmelstein and Woolhandler. Statistical analysis: Himmelstein and Woolhandler. Administrative, technical, and material support: Himmelstein and Woolhandler. Study supervision: Woolhandler.