Powered by 
Last month we highlighted Social Security’s cumulative effect on federal deficits in the next ten years. That often surprises people who think Social Security is “paid for” through payroll taxes.
What will surprise them even more: the much larger Medicare effect on federal deficits over the next decade.
You might ask yourself: don’t people pay for Medicare with payroll taxes when they work or with monthly premiums when they retire?
They do. And payroll taxes do cover almost all of Part A’s expenditures. But it is written into law that three-quarters of Parts B and D are explicitly subsidized by taxpayers. The Medicare Trustees report calls that money “government contributions.” It’s also often called general fund revenue transfers, since the actual cash is borrowed from taxpayer revenue. (Part A covers inpatient hospital stays. Part B covers doctor visits, outpatient care, and durable medical equipment. Part D covers prescription drugs.)
Those subsidies, and the interest on the borrowing they require, account for nearly half of everything the federal government will borrow over the next ten years.
The plot(s)
Our two plots show how much Medicare adds to the federal deficit each year and what percent of the federal deficit comes from Medicare’s general fund revenue transfers.
Total federal deficits are expected to be $1.85 trillion in 2026 and $2.96 trillion in 2035, an increase of 60 percent. Medicare’s effect on the deficit is projected to be $651 billion in 2026 and $1.73 trillion in 2035.
That’s why Medicare’s share of the federal deficit rises from 35 percent to 58 percent in just ten years. It passes half of the entire federal deficit in 2031.
How do you get these numbers? It’s relatively simple.
First, you project what spending will be for Parts A, B, and D. Then you subtract everything Medicare raises on its own. That includes payroll taxes, beneficiary premiums, the tax on Social Security benefits, and a few other small revenue sources. What you’re left with is the amount the general fund has to cover. That’s estimated to be $651 billion this calendar year and $1.34 trillion in ten years.
And since all this extra spending comes from federal borrowing, you then have to add in the interest costs. Granted, I’m starting from zero and not building in previous borrowing effects, so even these numbers are conservative. So, while 2026 has zero dollars of interest, the borrowing cost is up to $384 billion a year by 2035.
The point
Why are the numbers rising so fast?
Yes, more baby boomers are retiring, but that’s actually a small portion of the cost.
Instead, it’s driven by the per-person spending, mostly in Part B. When it costs more to cover either Part B or Part D, the government (via taxpayers) has to pay for three-quarters of the increase.
Part B spending per person is on track to increase between 5 percent and 8 percent a year, from $9,776 in 2026 to $17,618 in 2035. Once you combine that 80 percent per-person increase with more enrollees, Part B spending doubles.
Part D spending is also going up but is supposed to settle into a roughly 3 percent increase year-over-year once the Inflation Reduction Act’s provisions all take effect.
Part A is almost entirely paid for with payroll tax revenue and revenue from the taxation of benefits. In the future that will not be the case.
When splitting out by program, about 7 percent of the increase in deficits comes from Part A, 71 percent comes from Part B, and 22 percent comes from Part D.
By the way, ordinary consumer inflation isn’t really a driver. Medicare’s per-person costs have grown faster than inflation for decades, even as its payment rates have been held down by law. What grows is the utilization and mix of care. Each person gets more services over time, and the newest ones are often the most expensive. Remember, Medicare is an entitlement program without a true benefit cap. Medicare covers a broad set of services and traditional fee-for-service Medicare mostly pays for as much of them as beneficiaries use. Put it together and you see why per-person spending far outpaces inflation.
What do the numbers tell us?
That Medicare is heavily subsidized, and the more health care costs grow, the more it is going to drive federal borrowing.
Notes on calculations
Medicare data are from the 2026 Medicare Trustees Report, Tables III.B4, III.C4, and OACT supplementary tables.
Federal deficits are from CBO’s February 2026 Budget and Economic Outlook, Table 1-2, adjusted to exclude the effects of payment timing shifts.
Medicare outlays and revenues are in calendar years. Deficits are in fiscal years.
The interest calculation starts from a zero balance in 2026, so it counts only the borrowing this decade itself adds and nothing inherited from Medicare’s past. Federal deficits, by contrast, include interest on all prior borrowing. The early-year shares are understated for that reason.
Projections assume Part A benefits are paid in full after the trust fund depletes in 2033.
All figures are nominal.
Tom Church is a policy fellow at the Hoover Institution. He studies health care policy, entitlement reform, and the federal budget. Along with Daniel Heil and Lanhee Chen, he is a co-author of “Choices for All,” a set of commonsense health care reforms. He participates in Hoover’s Healthcare Policy Working Group, Fiscal Policy Initiative, and Tennenbaum Program for Fact-Based Policy.