Spending growth
Is that geographic gap about prices, or about how much care patients actually use?
Health care took 18.0 percent of GDP in 2024, and CMS projects 20.6 percent by 2034. Spending has grown faster than wages in most recent years, so the share of the economy devoted to care keeps rising.
The problem
Health spending growth is a sustainability problem for hospitals, employers, public budgets, and households. The national landscape does not have one spending problem: it has price growth, utilization variation, payer-mix pressure, and regional practice differences that require different management levers.
The recommendation
Segment the spending problem before choosing reforms. The recommended approach is to distinguish price-driven from utilization-driven growth, benchmark high-cost markets, and focus payment reform where spending does not produce better access or outcomes.
The national curve
Is health spending outrunning the economy and the paychecks that fund it, and which payer is driving the growth?
Cumulative growth since 2013: health spending, GDP, and wages (2013 = 100)
Each line compounds annual growth rates from a shared 2013 base of 100. A higher line means faster cumulative growth.
Read it this way By 2024, health spending had compounded to 184.6 on a base of 100, well ahead of GDP's 173.3 and wages' 155.6, so the gap between the top and bottom lines is the share of the economy health care absorbs that wages have not kept pace with. The chart shows relative growth from a shared starting point only, not dollar amounts or any single year's actual spending level. Use this chart to determine whether the spending pressure is price, utilization, geography, or payer mix, and to connect that diagnosis to the right cost-control lever.
Caveat Index levels are derived at authoring time by chaining the CMS Table 1 annual growth rates. Wage growth uses the SSA National Average Wage Index, a different concept and data source included for a directional comparison.
⊞ data table⬇ CSV
| Year | Health spending | GDP | Wages |
|---|---|---|---|
| 2013 | 100 | 100 | 100 |
| 2014 | 105.3 | 104.3 | 103.6 |
| 2015 | 111.1 | 108.4 | 107.2 |
| 2016 | 116.2 | 111.4 | 108.4 |
| 2017 | 121.1 | 116.2 | 112.1 |
| 2018 | 127.1 | 122.4 | 116.2 |
| 2019 | 133 | 127.6 | 120.5 |
| 2020 | 146.9 | 126.6 | 123.9 |
| 2021 | 153 | 140.5 | 134.9 |
| 2022 | 160.3 | 154.3 | 142.1 |
| 2023 | 172.2 | 164.6 | 148.4 |
| 2024 | 184.6 | 173.3 | 155.6 |
CMS NHE Projections Table 1, National Health Expenditures and Selected Economic Indicators · 2013-2034 · source
Annual growth: health spending versus GDP versus wages
Nominal annual percent change. Wage growth is the SSA National Average Wage Index, a separate series shown for comparison.
Read it this way Look for years where the health spending bar clears both the GDP and wages bars, such as 2015, 2016, and 2024, since those are the years the spending gap widened. 2020 and 2021 are pandemic-distorted outliers where GDP swung more than health spending did, so a single year of health spending trailing GDP does not mean the multi-year trend has reversed. Use this chart to determine whether the spending pressure is price, utilization, geography, or payer mix, and to connect that diagnosis to the right cost-control lever.
Caveat Wage growth uses the SSA National Average Wage Index, a different concept and data source from NHE or GDP, included for a directional comparison rather than as a component of the accounts.
⊞ data table⬇ CSV
| Year | Health spending % | GDP % | Wages % |
|---|---|---|---|
| 2014 | 5.3 | 4.3 | 3.55 |
| 2015 | 5.5 | 3.9 | 3.48 |
| 2016 | 4.6 | 2.8 | 1.13 |
| 2017 | 4.2 | 4.3 | 3.45 |
| 2018 | 5 | 5.3 | 3.62 |
| 2019 | 4.6 | 4.3 | 3.75 |
| 2020 | 10.5 | -0.8 | 2.83 |
| 2021 | 4.1 | 11 | 8.89 |
| 2022 | 4.8 | 9.8 | 5.32 |
| 2023 | 7.4 | 6.7 | 4.43 |
| 2024 | 7.2 | 5.3 | 4.84 |
CMS NHE Projections Table 1, National Health Expenditures and Selected Economic Indicators · 2013-2034 · source
National health spending by source of funds, 2018 through 2034
Dollars in billions. Years 2025 onward are CMS projections.
Read it this way The height of each band is that payer's dollar spending, so the widening total area from 2018 through 2034 shows spending growing across every payer at once, not just one. This view is about dollar scale, not growth rate, and 2025 onward is a CMS projection, not measured spending. Use this chart to determine whether the spending pressure is price, utilization, geography, or payer mix, and to connect that diagnosis to the right cost-control lever.
Caveat Other combines CHIP, DoD, VA, worksite care, workers compensation, public health activity, and investment. 2025 through 2034 are projections.
⊞ data table⬇ CSV
| Year | Medicare | Medicaid | Private insurance | Out of pocket | Other |
|---|---|---|---|---|---|
| 2018 | 751.6 | 596.8 | 1142.8 | 403 | 743.5 |
| 2019 | 805.2 | 615.4 | 1179.3 | 420.4 | 784.8 |
| 2020 | 833.8 | 672.5 | 1184.3 | 412.3 | 1101.4 |
| 2021 | 894.6 | 738.1 | 1263.1 | 457.2 | 1023.9 |
| 2022 | 951.7 | 809.9 | 1358.6 | 491.9 | 974.5 |
| 2023 | 1037.3 | 873.7 | 1511.2 | 525.4 | 977.7 |
| 2024 | 1118 | 931.7 | 1644.6 | 556.6 | 1027.7 |
| 2025 | 1204 | 1009 | 1779 | 591.1 | 1078.6 |
| 2026 | 1314.6 | 1057.1 | 1891.1 | 627.9 | 1126.7 |
| 2027 | 1422.7 | 1077.6 | 1979.1 | 651.8 | 1170.4 |
| 2028 | 1544.2 | 1115.7 | 2067.2 | 681.1 | 1218.9 |
| 2029 | 1664.6 | 1172.1 | 2153.6 | 710 | 1266.3 |
| 2030 | 1783.7 | 1226.4 | 2248.3 | 742.1 | 1315.5 |
| 2031 | 1911.1 | 1290.6 | 2347.2 | 773.6 | 1366.5 |
| 2032 | 2035.8 | 1359 | 2451.8 | 808 | 1420.1 |
| 2033 | 2188.3 | 1433.8 | 2565.7 | 845.1 | 1478 |
| 2034 | 2354.3 | 1518.6 | 2678.4 | 883.2 | 1535.8 |
CMS NHE Projections Table 3, National Health Expenditures by Source of Funds · 2018-2034 · source
Projected growth by payer, 2024 to 2034
Percent increase in nominal spending from 2024 to 2034. Medicare grows fastest as the population ages.
Read it this way Medicare's projected 110.6 percent growth roughly doubles the 49 to 63 percent projected for every other payer, which is what an aging population does to a program tied to age eligibility. These are ten-year projections computed from CMS's first and last modeled years, not a measured trend, so read the ranking as directional. Use this chart to determine whether the spending pressure is price, utilization, geography, or payer mix, and to connect that diagnosis to the right cost-control lever.
Caveat Computed from the first (2024) and last (2034) rows of the CMS projection. Projected years are modeled estimates, revised annually.
⊞ data table⬇ CSV
| Payer | 2024 ($B) | 2034 ($B) | Growth % |
|---|---|---|---|
| Medicare | 1118 | 2354.3 | 110.6 |
| Medicaid | 931.7 | 1518.6 | 63 |
| Private insurance | 1644.6 | 2678.4 | 62.9 |
| Out of pocket | 556.6 | 883.2 | 58.7 |
| Other | 1027.7 | 1535.8 | 49.4 |
CMS NHE Projections Table 3, National Health Expenditures by Source of Funds · 2018-2034 · source
Same care, different price
How wide is the geographic spending gap, and is it driven by prices or by how much care is used? Medicare fee-for-service Parts A and B per beneficiary, 2022.
Medicare Parts A and B spending per beneficiary by state, 2022
Darker means higher per-beneficiary spending. Fee-for-service Medicare only, not all payers. DC is shown as a state tile.
Read it this way Darker states, led by New York at $15,008, spend well above the $12,023 national average per Medicare beneficiary. Find your state's shade to see roughly where it lands. This map reflects Medicare fee-for-service Parts A and B only, so it says nothing about what commercially insured patients pay in the same states. Use this chart to determine whether the spending pressure is price, utilization, geography, or payer mix, and to connect that diagnosis to the right cost-control lever.
Caveat This is Medicare fee-for-service Parts A and B per beneficiary in 2022, a subset of the population, not all-payer spending.
⊞ data table⬇ CSV
| State | Spending per beneficiary |
|---|---|
| New York | 15007.92 |
| California | 14511.77 |
| Connecticut | 14241.64 |
| New Jersey | 14162.99 |
| Maryland | 13758.56 |
| District of Columbia | 13753.56 |
| Massachusetts | 13090.3 |
| Texas | 12921.76 |
| Florida | 12860.11 |
| Nevada | 12544.73 |
| Louisiana | 12486.33 |
| Illinois | 12314.04 |
| Oklahoma | 12104.63 |
| Delaware | 12068.92 |
| Mississippi | 11978.22 |
| Pennsylvania | 11642.84 |
| Indiana | 11382.19 |
| Minnesota | 11378.74 |
| Michigan | 11356.16 |
| Georgia | 11341.91 |
| Rhode Island | 11334.85 |
| Alabama | 11300.4 |
| Nebraska | 11166.18 |
| Kansas | 11163.17 |
| Missouri | 11155.94 |
| Ohio | 11024.73 |
| Kentucky | 10879.36 |
| Alaska | 10833.2 |
| Arizona | 10823.71 |
| Utah | 10795.57 |
| South Carolina | 10766.52 |
| Arkansas | 10705.38 |
| Tennessee | 10604.23 |
| West Virginia | 10587.76 |
| New Hampshire | 10519.89 |
| Virginia | 10494 |
| North Carolina | 10489.59 |
| South Dakota | 10444.74 |
| Colorado | 10409.66 |
| North Dakota | 10392.04 |
| Vermont | 10228.34 |
| Wisconsin | 10205.24 |
| Iowa | 10178.28 |
| Maine | 10076.69 |
| Wyoming | 10048.01 |
| Washington | 9866.73 |
| New Mexico | 9594.89 |
| Oregon | 9573.37 |
| Idaho | 9442.81 |
| Hawaii | 9350.43 |
| Montana | 9027.17 |
CMS Medicare Geographic Variation Public Use File, by National, State and County · 2022 · source
Medicare Parts A and B spending per beneficiary, 51 states ranked, 2022
Each dot is one state. The line marks the national average. New York spends 1.66 times what Montana does.
Read it this way The dots sitting farthest right of the average line, led by New York at $15,008, spend the most per beneficiary. The 1.66-times gap between New York and Montana ($9,027) is the distance between the two ends of the strip. Rank order here does not explain why a state spends more, only that it does. Use this chart to determine whether the spending pressure is price, utilization, geography, or payer mix, and to connect that diagnosis to the right cost-control lever.
Caveat Medicare fee-for-service Parts A and B per beneficiary, 2022, a subset of all payers. Includes DC, so 51 rows.
⊞ data table⬇ CSV
| State | Spending per beneficiary |
|---|---|
| New York | 15007.92 |
| California | 14511.77 |
| Connecticut | 14241.64 |
| New Jersey | 14162.99 |
| Maryland | 13758.56 |
| District of Columbia | 13753.56 |
| Massachusetts | 13090.3 |
| Texas | 12921.76 |
| Florida | 12860.11 |
| Nevada | 12544.73 |
| Louisiana | 12486.33 |
| Illinois | 12314.04 |
| Oklahoma | 12104.63 |
| Delaware | 12068.92 |
| Mississippi | 11978.22 |
| Pennsylvania | 11642.84 |
| Indiana | 11382.19 |
| Minnesota | 11378.74 |
| Michigan | 11356.16 |
| Georgia | 11341.91 |
| Rhode Island | 11334.85 |
| Alabama | 11300.4 |
| Nebraska | 11166.18 |
| Kansas | 11163.17 |
| Missouri | 11155.94 |
| Ohio | 11024.73 |
| Kentucky | 10879.36 |
| Alaska | 10833.2 |
| Arizona | 10823.71 |
| Utah | 10795.57 |
| South Carolina | 10766.52 |
| Arkansas | 10705.38 |
| Tennessee | 10604.23 |
| West Virginia | 10587.76 |
| New Hampshire | 10519.89 |
| Virginia | 10494 |
| North Carolina | 10489.59 |
| South Dakota | 10444.74 |
| Colorado | 10409.66 |
| North Dakota | 10392.04 |
| Vermont | 10228.34 |
| Wisconsin | 10205.24 |
| Iowa | 10178.28 |
| Maine | 10076.69 |
| Wyoming | 10048.01 |
| Washington | 9866.73 |
| New Mexico | 9594.89 |
| Oregon | 9573.37 |
| Idaho | 9442.81 |
| Hawaii | 9350.43 |
| Montana | 9027.17 |
CMS Medicare Geographic Variation Public Use File, by National, State and County · 2022 · source
Actual versus price-standardized spending per beneficiary, by state, 2022
The horizontal axis removes price differences. States sitting high above their horizontal position spend more because local prices are higher, not because patients use more care.
Read it this way States sitting well above the diagonal, like New York and California, spend more mainly because local prices are higher: removing price differences on the horizontal axis pulls their point down closer to the pack. States that sit close to the diagonal are spending roughly what their price-standardized quantity of care would predict, meaning volume and mix, not price, explain their level. Use this chart to determine whether the spending pressure is price, utilization, geography, or payer mix, and to connect that diagnosis to the right cost-control lever.
Caveat Both measures are Medicare fee-for-service Parts A and B per beneficiary, 2022. Price-standardized spending removes geographic price differences to approximate the quantity of care.
⊞ data table⬇ CSV
| State | Actual per beneficiary | Price-standardized |
|---|---|---|
| NY | 15007.92 | 12143.9 |
| CA | 14511.77 | 11354.83 |
| CT | 14241.64 | 12020.39 |
| NJ | 14162.99 | 12049.7 |
| MD | 13758.56 | 11130.01 |
| DC | 13753.56 | 11131.51 |
| MA | 13090.3 | 10916.1 |
| TX | 12921.76 | 12522.07 |
| FL | 12860.11 | 12657.25 |
| NV | 12544.73 | 11216.06 |
| LA | 12486.33 | 13055.28 |
| IL | 12314.04 | 11536.34 |
| OK | 12104.63 | 12481.21 |
| DE | 12068.92 | 10904.4 |
| MS | 11978.22 | 12590.89 |
| PA | 11642.84 | 10774.65 |
| IN | 11382.19 | 11241.34 |
| MN | 11378.74 | 10431.29 |
| MI | 11356.16 | 10907.74 |
| GA | 11341.91 | 11128.87 |
| RI | 11334.85 | 10141.91 |
| AL | 11300.4 | 11882.97 |
| NE | 11166.18 | 10822.21 |
| KS | 11163.17 | 11281.19 |
| MO | 11155.94 | 11140.2 |
| OH | 11024.73 | 10792.07 |
| KY | 10879.36 | 11065.74 |
| AK | 10833.2 | 8390.88 |
| AZ | 10823.71 | 10331.79 |
| UT | 10795.57 | 10314.68 |
| SC | 10766.52 | 10726.27 |
| AR | 10705.38 | 11292.84 |
| TN | 10604.23 | 10905.48 |
| WV | 10587.76 | 10649.39 |
| NH | 10519.89 | 9403.34 |
| VA | 10494 | 10038.02 |
| NC | 10489.59 | 10159.19 |
| SD | 10444.74 | 9991.77 |
| CO | 10409.66 | 9808.51 |
| ND | 10392.04 | 9931.98 |
| VT | 10228.34 | 8567.43 |
| WI | 10205.24 | 9699.93 |
| IA | 10178.28 | 10038.6 |
| ME | 10076.69 | 9447.5 |
| WY | 10048.01 | 9032.07 |
| WA | 9866.73 | 8756.02 |
| NM | 9594.89 | 9153.59 |
| OR | 9573.37 | 8377.18 |
| ID | 9442.81 | 9213.02 |
| HI | 9350.43 | 7650.67 |
| MT | 9027.17 | 8554.16 |
CMS Medicare Geographic Variation Public Use File, by National, State and County · 2022 · source
Inpatient use versus price-standardized spending, by state, 2022
The dashed line is a least-squares fit. States with more hospital use tend to have higher price-neutral spending.
Read it this way States further right, like DC and New York with the most inpatient days per 1,000 beneficiaries, also tend to sit higher on price-standardized spending, and the dashed trend line traces that upward relationship. The caveat on this chart is explicit that the line is directional, not a causal estimate of how much utilization drives spending. Use this chart to determine whether the spending pressure is price, utilization, geography, or payer mix, and to connect that diagnosis to the right cost-control lever.
Caveat Inpatient days per 1,000 beneficiaries and price-standardized spending are both from the CMS Geographic Variation file, 2022. The trend line is directional, not a causal estimate.
⊞ data table⬇ CSV
| State | Inpatient days per 1,000 | Price-standardized |
|---|---|---|
| DC | 1846.9 | 11131.51 |
| NY | 1656.6 | 12143.9 |
| CT | 1547.9 | 12020.39 |
| MA | 1531.2 | 10916.1 |
| NJ | 1494.8 | 12049.7 |
| AL | 1427 | 11882.97 |
| WV | 1412.3 | 10649.39 |
| MI | 1398.1 | 10907.74 |
| MD | 1389 | 11130.01 |
| DE | 1376.6 | 10904.4 |
| MS | 1369.2 | 12590.89 |
| IL | 1366.8 | 11536.34 |
| LA | 1343.7 | 13055.28 |
| FL | 1342 | 12657.25 |
| KY | 1335.3 | 11065.74 |
| GA | 1311.3 | 11128.87 |
| PA | 1310 | 10774.65 |
| MO | 1298.8 | 11140.2 |
| NV | 1250.4 | 11216.06 |
| IN | 1230.3 | 11241.34 |
| MN | 1226.5 | 10431.29 |
| RI | 1224.8 | 10141.91 |
| OK | 1223 | 12481.21 |
| TN | 1219.3 | 10905.48 |
| OH | 1204.2 | 10792.07 |
| CA | 1200.2 | 11354.83 |
| NC | 1187.9 | 10159.19 |
| TX | 1187.3 | 12522.07 |
| AR | 1162.7 | 11292.84 |
| NH | 1125.4 | 9403.34 |
| ME | 1113.8 | 9447.5 |
| WI | 1113.4 | 9699.93 |
| VA | 1101.6 | 10038.02 |
| KS | 1084.1 | 11281.19 |
| VT | 1077.7 | 8567.43 |
| AK | 1069.7 | 8390.88 |
| ND | 1059.8 | 9931.98 |
| SC | 1054.9 | 10726.27 |
| NE | 1037.6 | 10822.21 |
| WA | 1027.4 | 8756.02 |
| SD | 992.8 | 9991.77 |
| IA | 961.8 | 10038.6 |
| OR | 945.2 | 8377.18 |
| NM | 930 | 9153.59 |
| AZ | 925.4 | 10331.79 |
| HI | 903.4 | 7650.67 |
| WY | 871.4 | 9032.07 |
| CO | 859.9 | 9808.51 |
| MT | 832.5 | 8554.16 |
| ID | 763.9 | 9213.02 |
| UT | 726.1 | 10314.68 |
CMS Medicare Geographic Variation Public Use File, by National, State and County · 2022 · source
Geography
The same question, state by state and then county by county. Pick a state in the filter above to drill into its counties.
Medicare per-capita spending (FFS, standardized)
County · Medicare FFS onlyFee-for-service only; standardized to remove local price differences. MA-heavy counties reflect the FFS remainder.
Each tile is a state. Pick a state in the Scope control above to drill into its counties.
CMS Medicare Geographic Variation PUF · 2023 · source
Why this matters
On the payer side, Medicare's projected 110.6 percent growth from 2024 to 2034 is nearly double every other payer's 49 to 63 percent, consistent with an aging population drawing on an age-eligible program. On the geographic side, price differences, not volume of care, explain most of the 1.66-times gap between the highest- and lowest-spending states: removing price differences pulls high-spend states like New York and California back toward the pack, while inpatient utilization shows only a directional relationship with price-standardized spending.
Recommended actions
- Track the NHE share of GDP as the primary affordability indicator rather than metrics that only shift spending between payers.
- Prioritize Medicare-specific cost containment given its disproportionate projected growth relative to every other payer.
- Examine price levels, not utilization, as the main lever in high-differential states such as New York, California, Connecticut, and New Jersey.
- Treat the 2025 through 2034 CMS figures as projections revised annually, not fixed targets, and revisit the curve each year.
- Pair utilization data with price-standardized data before attributing any state's spending gap to either cause alone.
The recommendation
Therefore, segment the spending problem before choosing reforms. The recommended approach is to distinguish price-driven from utilization-driven growth, benchmark high-cost markets, and focus payment reform where spending does not produce better access or outcomes.
Demographic slice none. NHE is a national aggregate.
Sources
- CMS Office of the Actuary, NHE Fact Sheet · 2024
- CMS NHE Projections Table 1, National Health Expenditures and Selected Economic Indicators · 2013-2034
- CMS NHE Projections Table 3, National Health Expenditures by Source of Funds · 2018-2034
- SSA Office of the Chief Actuary, National Average Wage Index · 2024
- CMS Medicare Geographic Variation Public Use File, by National, State and County · 2022