CMS Actuaries Project Health Spending Reaches 20.6% of GDP by 2034 as Medicare Growth Outpaces Every Other Payer

National health expenditures compound off a 2024 base of $5.3 trillion; gross domestic product compounds off a base roughly five and a half times larger. When the numerator grows 5.4% a year and the denominator grows 4.1%, the gap between them does not close, it widens every year, and ten years of that arithmetic is what pushed the CMS Office of the Actuary to project health spending at 20.6% of GDP by 2034, up from 18.0% in 2024, in the National Health Expenditure projections it published in Health Affairs on June 24, 2026.

The Arithmetic Behind the Widening Wedge

By 2034, CMS's actuaries project national health spending will total nearly $9.0 trillion, up from the $5.3 trillion recorded in 2024 (CMS Office of the Actuary via Health Affairs, June 2026). Over the 2025-34 window, the agency models average NHE growth of 5.4% a year against average GDP growth of 4.1%, a spread of roughly 130 basis points that compounds rather than resets each year. That compounding is the entire mechanism behind the GDP-share climb: a constant annual gap of even one percentage point, held for a decade, moves the ratio by several points, and CMS's own trajectory shows the share crossing 19% before 2030 on its way to 20.6% by 2034.

The projection is not a single smooth curve. CMS's actuaries describe three forces operating in sequence: elevated utilization of medical services and goods that persists through 2026, major legislative changes to insurance coverage and spending that play out from 2026 through 2028, and a longer demographic shift of the population toward public programs, mainly Medicare, that continues through the full window. The first force pushes growth up early. The second, dominated by the One Big Beautiful Bill Act's Medicaid and marketplace subsidy provisions, pulls growth down in the middle years. The third reasserts itself in the second half of the decade as the workforce ages into Medicare regardless of what happens to any single statute.

Medicare's 7.7% Growth Path and Why the Middle Years Look Different

Medicare is the fastest-growing major payer in the projection, expanding at an average 7.7% a year from 2025 through 2034, a rate CMS attributes to both faster per-enrollee cost growth and faster enrollment growth as the population ages (CMS NHE Fact Sheet, June 2026). That average, though, masks a mid-decade deceleration. CMS's actuaries model average per-enrollee Medicare spending trend across 2027-28 at roughly 6.0%, softer than the growth rates surrounding it, before the full-window pace reasserts itself later in the projection. The near-term driver inside that number is retail prescription drugs: CMS models Medicare Part D spending growth at 14.7% in 2026, slowing to an average of 8.4% across 2027-28, a deceleration the agency links partly to moderating GLP-1 utilization growth and to the Inflation Reduction Act's drug-pricing provisions taking fuller effect.

Retail prescription drugs overall are the fastest-growing spending category in the projection, and that growth is front-loaded into 2025 and 2026 rather than spread evenly across the decade. For plan sponsors and PBMs building multi-year trend assumptions, the practical read is that the drug-spend acceleration the industry has been pricing for GLP-1s and specialty pharmacy is not a permanent new baseline in CMS's own model; it is a two-year spike that the agency expects to fade as utilization growth normalizes and IRA negotiated pricing phases in more fully.

Payer / Category	Avg. Annual Growth, 2025-2034
Medicare	7.7%
Medicaid	5.0%
Private health insurance	5.0%
Out-of-pocket spending	4.7%
National health expenditures (all payers)	5.4%
Gross domestic product	4.1%

Source: CMS Office of the Actuary, National Health Expenditure Projections 2025-2034 (Health Affairs, June 2026).

The First Modeled Medicaid Enrollment Decline Under OBBBA

This is the first NHE release to incorporate the Medicaid provisions of the One Big Beautiful Bill Act, and it shows up directly in the enrollment path. CMS projects the insured share of the population falling from 91.8% in 2024 to 90.5% in 2034, reversing a run that had reached a historic high of 92.5% as recently as 2023 (CMS Office of the Actuary via Health Affairs, June 2026; PubMed abstract, June 2026). The dip is not smooth: CMS models the insured share bottoming near 90.4% around 2028, driven by the phased rollout of Medicaid work requirements and the scheduled expiration of enhanced ACA marketplace subsidies at the end of 2026.

Independent estimates of the same policy give pricing actuaries a range to bracket CMS's internal figure against. The Congressional Budget Office estimated in July 2025 that OBBBA would increase the number of uninsured Americans by 10 million by 2034, split roughly 7.5 million from Medicaid policy changes and 2.1 million from marketplace subsidy expiration. Of the Medicaid portion, CBO attributed 5.3 million to the new work and community engagement requirement alone, before accounting for interaction effects with other provisions. CMS's own regulatory impact analysis for the work requirement rule, summarized by the Healthcare Financial Management Association, projects enrollment falling by 2.3 million in the rule's first full year and by 3.1 million to 3.3 million annually from the following year onward, relative to a no-OBBBA baseline. A separate Urban Institute analysis using a comparable methodology to RAND's puts the 2034 Medicaid enrollment reduction at approximately 7.6 million, within a few hundred thousand of CBO's figure. Three independently built estimates converging within a tight band gives plan actuaries an unusually well-bracketed range to reserve and price against, rather than the single-point estimates that typically accompany a major coverage policy change.

The mechanism behind the decline is not simply fewer people on the rolls; it is a compositional shift in who remains. CMS's actuaries have already documented this dynamic in actual 2024 data: Medicaid enrollment fell 7.9% to 84.5 million that year as the FFCRA continuous-enrollment provision expired, and per-enrollee Medicaid spending grew 15.2% in the same year, a jump CMS attributes to younger, healthier enrollees churning off the program faster than the older, sicker population that remains (Peterson-KFF Health System Tracker summary of CMS data, 2026). OBBBA's work requirements and twice-yearly redeterminations are modeled to run the same churn dynamic a second time, on a population that has already been through one unwinding cycle. That is the reason aggregate Medicaid spending growth in the new projection, 5.0% a year, sits well below Medicare's 7.7% even as the remaining Medicaid population gets measurably sicker on average: falling enrollment offsets rising per-capita acuity in the aggregate figure, and the two effects move in opposite directions inside a single line item.

Marketplace Subsidy Expiration and the Adverse-Selection Path to 2028

The insured-share trough CMS models for 2028, 90.4%, is not driven by Medicaid alone. Enhanced ACA marketplace subsidies are scheduled to expire at the end of 2026, and CMS's actuaries fold the resulting direct-purchase coverage decline into the same 2026-28 legislative-impact window that carries the Medicaid work-requirement effect. The mechanism differs from Medicaid's churn dynamic in a way that matters for rate filers: Medicaid disenrollment is largely procedural, driven by paperwork and reporting lapses that skew toward younger, healthier enrollees, while marketplace disenrollment following a subsidy cliff is a price-driven decision, and price sensitivity is not evenly distributed across a risk pool. Healthier enrollees with more elastic demand for coverage are more likely to drop out when their net premium jumps, leaving a marketplace risk pool that skews sicker on average even before accounting for any Medicaid crossover effect.

For actuaries certifying 2027 ACA rate filings, that means the same compositional problem showing up in Medicaid capitation work reappears on the individual market side, on a faster timeline. A base period built on 2025-26 marketplace experience, collected while enhanced subsidies were still in effect, will systematically understate the morbidity of the 2027 pool once the price-sensitive, healthier enrollees have exited. CMS's own projection of private health insurance spending growth averaging 5.0% a year over 2025-34, a figure close to Medicaid's aggregate rate for a similar reason, masks the same offsetting dynamic: falling marketplace enrollment pulls the aggregate figure down even as per-enrollee morbidity trend among those who remain runs meaningfully higher. A rate filing that applies a flat 5% trend to a marketplace block without adjusting for the subsidy-driven exit of healthier lives is pricing last year's risk pool, not the one the policy will produce.

What the Payer Mix Means for Medicare Advantage Bid Trend

CMS's own trend assumptions do not directly set Medicare Advantage bid pricing parameters, which flow through the separate Rate Announcement and Advance Notice process, but the NHE projection is the closest thing the industry has to CMS's institutional view of where per-capita Medicare cost growth is heading, and MA bid actuaries use it as an external check on internally developed trend. A 7.7% average payer-level growth rate, decelerating to roughly 6.0% on a per-enrollee basis through 2027-28 before demographic pressure reasserts itself, is a materially different signal than a flat 7-plus percent trend carried uniformly across the bid year. Plans that built 2027 and 2028 bids assuming the higher end of recent Medicare trend without reflecting the mid-decade dip risk overstating expected cost and underpricing the value they need to deliver through supplemental benefits to stay competitive on star ratings and enrollment growth, both of which this site has tracked through the 2027 MA rate cycle.

The composition of Medicare's growth matters as much as its level. Because CMS attributes the 7.7% average partly to enrollment growth as the population ages into the program and partly to per-enrollee cost growth, MA plans need to separate the two in their own bid trend work rather than treating the payer-level figure as a per-member trend input. Enrollment growth raises total MA revenue and total risk-adjusted spend without necessarily raising per-member cost; per-enrollee cost growth is the actual trend factor that belongs in a bid pricing model. Conflating the two, a mistake this site has flagged in prior coverage of the 2027 MA rate reversal, overstates the trend a plan needs to price for and can leave money on the table in a competitive bid environment where CMS's own benchmark growth already reflects the demographic tailwind.

Medicaid Capitation Repricing Under a Shrinking, Sicker Pool

The consequence for Medicaid managed care actuaries runs through the same compositional logic that played out in 2024, at larger scale and on a longer timeline. A 5.0% average aggregate Medicaid growth rate sitting alongside a multi-million-enrollee decline means per-capita capitation trend, the number that actually drives a state's rate certification, has to run well above the aggregate figure once the OBBBA-driven churn accelerates in 2027 and 2028. Base-period claims data collected before the work requirements take effect will understate the acuity of the population that remains once healthier, more mobile enrollees are disenrolled for missed reporting or lapsed six-month redeterminations, the same dynamic this site examined in the 2026-2027 Medicaid rate guide's certification requirements.

The practical implication for a rate-setting actuary is that a capitation rate built on a trailing twelve-to-eighteen-month base period, standard practice under most state rate development guides, will be pricing an enrollee mix that no longer exists by the time the rate takes effect, if the base period predates the bulk of the work-requirement disenrollment. States and their contracted actuaries have limited tools to correct for this in real time: an explicit population composition adjustment, disclosed with its own uncertainty range, is the closest substitute for claims experience that has not yet accumulated. CMS's separate proposal to cap state-directed payments compounds the repricing pressure on the same managed care organizations absorbing the composition shift, tightening the margin available to absorb a rate that turns out to be too low once the healthier enrollees have already left the program.

The Uncompensated-Care Path Back to Employer Plans

A rising uninsured share does not remove demand for care, it relocates who pays for it. Hospitals and health systems absorbing a larger uncompensated-care load from a population that lost Medicaid or marketplace coverage typically respond by shifting costs onto payers with negotiating leverage they can still exercise, historically the commercial book. CMS's own projection of private health insurance spending growth, 5.0% a year on average, sits close to the levels the agency modeled in prior cycles when the insured share was rising rather than falling, which understates the case for a cost-shift premium building into that figure as the uninsured population grows from roughly 8.2% of the population in 2024 toward the high single digits later in the window. Group underwriters and stop-loss actuaries pricing multi-year commercial contracts through 2028 are pricing into a hospital cost-shifting environment that CMS's aggregate private-insurance trend line does not explicitly decompose, and that the agency's methodology paper treats as an assumption embedded in the hospital spending category rather than a separately quantified commercial add-on.

The timing compounds the exposure rather than diluting it. The same 2027-28 window in which CMS models the insured-share trough is the window in which Medicare Part D drug trend is decelerating and Medicaid per-capita costs are climbing, which means a commercial actuary reading CMS's aggregate 5.4% NHE growth figure as a stable, blendable planning number is averaging across payer categories that are moving in different directions for different reasons in the same two years. Treating 2027-28 as a trend-flat period because the aggregate figure looks unremarkable would miss the compositional shift running underneath it on both the public and private sides of the ledger.

What the Ten-Year Window Implies for Trend-Setting

The CMS projection's most durable value to a pricing actuary is not any single-year number, it is the shape of the decade: front-loaded utilization growth through 2026, a legislatively driven trough through 2027-28 concentrated in Medicaid and marketplace coverage, and a demographic reassertion of Medicare-driven growth for the remainder of the window that no policy change currently on the books is projected to offset. Building a flat trend assumption across that shape, whether for a Medicare Advantage bid, a Medicaid capitation rate, or an ACA marketplace filing, this site's coverage of HHS's 2027 marketplace repricing rules and its analysis of GLP-1-driven employer trend for 2027 bids both make the same point from different angles, understates the volatility actuaries need to reserve and price against inside any single projection year, even when the ten-year average lands close to what CMS models.