Blog

In February, Republicans announced their intention to cut $2 trillion in spending over the next 10 years. $880 billion of that is to come from the Committee on Energy and Commerce (See Title II), which oversees Medicare, Medicaid, national energy policy, environmental protection, and several others.

Just doing basic searches on Medicaid cuts, trying to find information, all of the top stories include a warning about how many people are going to be hurt by this, but interestingly, there are no mentions of how much Medicaid has grown in the past ten years, even beyond its ACA expansion, how fast it's projected to grow, or how much waste could be cut. Every story just assumes, without including details, that every penny of Medicaid goes to children, poor adults, or disabled adults, but given the recent expansions, this seems unlikely.

The Recent Run-Up in Medicaid

Since 2000, federal Medicaid spending has been trending upward. In 2000, after 8 years of a Democratic president, federal Medicaid spending was just 1.1% of GDP, and it has nearly doubled since then. Some of that surely is the increasing cost of medical care, but it's also driven by enrollment growth due to expanded eligibility and state incentives to get matching federal funding.

Federal Medicaid Spending as a % of GDP

Source: Author Calculation based on NHE (Actual Federal Medicaid Expenditures), FRED (Actual GDP), CBO (Projected Medicaid and GDP)

In the last twenty-five years, the largest, single-year increase in federal Medicaid spending occurred in 2020, the year of Covid. This, in and of itself, is unsurprising, but what's harder to understand is why it has barely fallen back to its pre-Covid rate, even four years later, while the economy has boomed.

In the last twenty-five years, the largest, single-year increase in federal Medicaid spending occurred in 2020, the year of Covid. This, in and of itself, is unsurprising, but what's harder to understand is why it has barely fallen back to its pre-Covid rate, even four years later, while the economy has boomed.

Waste, Fraud, and Abuse

CMS periodically estimates the total "improper payments" in Medicare and Medicaid. The most recent estimate for Medicaid is that 5.1% of total Medicaid payments are "improper." While that sounds like fraud, in fact, the majority of these improper payments are simply insufficiently documented legitimate payments. They also estimated how much of the improper payments were actually illegitimate and found that 20% of them are truly improper and should not have been paid out.

Applying this rate of illegitimate payments to the CBO's ten-year forecast produces an estimate of $91.2 billion over ten years that can be saved without harming Medicaid recipients.

Applying this rate of illegitimate payments to the CBO's ten-year forecast produces an estimate of $91.2 billion over ten years that can be saved without harming Medicaid recipients.

Locking in Normal Growth

One approach Republicans might consider would be to lock in overall Medicaid growth to match its 2016-2019 rate. This was a time of a stable economy, after the major ACA expansion and before further expansions under Covid-era policies and the Biden administration. It represents a steady-state era for Medicaid.

During this time, Medicaid grew at 3.1% per year. This would represent smaller growth than the CBO estimated (between 4 and 5% per year).

Federal Medicaid Spending Growth by Year 2000-2024

Source: Author Calculation based on NHE (Actual Federal Medicaid Expenditures), CBO (Projected Medicaid)

Holding growth in Medicaid to the 2016-2019 average over the next ten years would save $778 billion. Considering that 2016-2019 was a time without major cuts to Medicaid benefits or eligibility, it would seem possible to lock in that growth rate without harming any beneficiaries.

Federal Medicaid Spending if Growth Held to 2016-2019 Rate

Source: Author Calculation based on NHE (Federal Medicaid Expenditures), CBO (Projected Medicaid)

Locking in Enrollment Levels

Another path Republicans might take is setting an enrollment target for Medicaid, which is supposed to assist the poorest Americans, yet enrolls close to 1 in 4. According to poverty statistics, before accounting for transfers 11.1% of Americans are impoverished. Medicaid, however, covered 27.4% of the US population in 2023.1

From 2000 to 2023, enrollment in Medicaid has grown from 12% of the population to 27% of the population, more than doubling, even though the federal poverty rate has barely moved, going from 11.3 to 12.5%. Some of that, obviously, was intentional, as the ACA expanded Medicaid, but in 2019, before the pandemic hit, and after state expansions had taken place, Medicaid penetration was only 23.0%. Since Covid, it has risen several percentage points, even without expanding Medicaid in large states like Texas and Florida.

% of Total US Population enrolled in Medicaid by Year

Source: Author Calculation based on NHE (Total Medicaid Enrollment), CBO (Projected Medicaid Enrollment). Note: The 2025 CBO projection for Medicaid spending is substantially higher than the 2024 CBO projection, but they have not yet released the breakdown. The dotted line represents a higher level of enrollment which could explain the higher CBO estimates.

Setting a maximum Medicaid enrollment of 23.0%, the highest level it reached before Covid and after the ACA expansion, would save $1.4 trillion over the next ten years.

Capping Medicaid enrollment at its pre-Covid peak would save $1.4 trillion over the next ten years.

Federal Medicaid Spending if Enrollment Held at Pre-Covid Maximum

Source: Author Calculation based on NHE (Total Medicaid Enrollment), CBO (Projected Medicaid Enrollment)

Conclusion

Rolling the Medicaid program back to the pre-Covid norm, when there were no significant criticisms of its inadequacy, would save more than Republicans are pushing for.

Medicaid has expanded in scope and scale over the past twenty-five years and has grown to a point where some pruning can be accomplished without the devastating results that many warn will occur. Simply rolling the program back to its scale from prior to the pandemic would provide more money than Congress is looking for and would return Medicaid to funding and enrollment levels that, at the time, drew little to no criticism.

Notes

For a counterargument, Kaiser Family Foundation claims it's impossible.
For some supportive analysis and a more detailed proposal for Medicaid, Manhattan Institute

Stray Thoughts

One of the explanations CBO gave for increased expected Medicaid spending from 2026-2035 was new spending on GLP-1s, which reduce obesity. Theoretically, though, the reduction in obesity should also lead to improved health and reduced spending. I'm curious if that is included in their estimates.

Footnotes

1This number has fallen in 2024 as the Medicaid expansions from the Covid-era are being unwound. The rate today is between 25 and 26%.

In light of the Trump administration and DOGE’s antipathy towards foreign aid and NIH, and their significant budget cuts, the most moral and productive thing that can be done, for those who strongly disagree with these decisions and are concerned with the hardships they’ll cause worldwide is to donate money directly to the affected programs.

Below, I’ve identified the entire costs of these programs, but also try to break them out individually. I’ve also calculated how much each person would need to donate to make up for the cuts and, where I could, pointed to some organizations that have a history of addressing the same issues.

I encourage you to share any other organizations you come across that could fill in the gap. Also, if you have any links to news articles that identify some of the most effective programs within these agencies, please leave them in comments, and I’ll see if I can add them in to the estimates.

US Total Expenditures on Foreign Aid and Research

• The entire USAID budget peaked in 2023 at about $50 billion.

US Aid Budget by Year

• While USAID is the largest component, the entire foreign aid budget is $71.9 billion.
• The NIH budget is $47 billion.

Some Background Math

For tax-year 2022, the most recent-available, there were 161,336,659 total tax returns filed. Of these, 55 million were married filing jointly, 4 million were married filing separately, 21 million were heads of household, and 81 million were individuals. Since some of the 55 million married filing jointly were dual earners, I determined the total number of tax-payers by assuming 48.9% of the married couples were dual earners. That comes out to 187,155,980 earners paying federal income taxes, if my math is right. As an alternative number, there were 231,529,762 Americans over the age of 25 in 2023.

Filling the Gap

The table below shows how much each adult or taxpayer would need to contribute to completely replace the programs. For example, if NIH were completely eviscerated, because the federal government spent $47 billion on it, each taxpayer would need to throw in $251.13 to make up for the loss.

Obviously, reality is much more complicated. Perhaps you want to make up for your NIH-skeptical friends and double your contribution. And, at this point, it’s not even clear that NIH spending will be cut, without Congressional approval. Also, maybe you want to focus your charity on the most effective aspects of these programs and not everything.

Individual Components

USAID

On March 10th, Secretary Rubio announced the review of USAID programs was done and they were cancelling “83% of programs at USAID.” Because “programs” could encompass small projects and large, it’s unclear how to translate that to a dollar figure. Fortunately, the Center for Global Development estimates the dollar figure to be around 34% of the annual spending. Of course, the normal caveat remains, that this money, as of now, still must be spent on foreign aid, it’s just Rubio and the new Trump-designated team that will direct it.

In addition, one might suspect that the programs cut would be the least beneficial and they maintained the most effective ones (the ones that would be the most difficult to defend cutting to voters), but nevertheless, 34% of USAID amounts to $15 billion.

The other major USAID program is PEPFAR, the US President’s Emergency Plan for AIDS Relief. Created by George W. Bush, PEPFAR provides relief to people living with AIDS around the world and is credited with saving more than 25 million lives.

The FY 2024 funding was $6.5 billion. Now, while that funding hasn’t been cut, it’s not getting to the people who need it, so the best use of dollars would probably address some of that gap.

Some organizations that overlap with PEPFAR

• The Global Fund (to fight AIDS, Tuberculosis, and Malaria)
• AIDS Healthcare Foundation
• UNAIDS
• AIDS Foundation South Africa
• Elton John AIDS Foundation - has a “Rocket Response Fund” as a direct result of US cuts

Notes

• Some interesting information on foreign aid
• Some good info on USAID

While a majority of Americans support government funded scientific research, and many support foreign assistance, people often argue that foreign aid and government-funded research are miniscule parts of the federal budget so don't matter, but when you How much does the typical taxpayer contribute to foreign aid and NIH? The answer will probably surprise you. Lower than you might think.

Firstly, as of 2024, US government revenues were 17.1% of GDP, while expenditures were 23.4% of GDP. This means, in 2024, revenues made up only of government revenues only cover 72.9% of government expenditures, so we can say that only 72.9% is paid for by taxes, the rest is paid for by debt-purchasers.

Of all government revenues, income taxes make up about 49%. The next largest portion, 35%, comes from payroll taxes, but since payroll taxes are meant for Social Security and Medicare, they are not used for foreign aid or NIH. Because Social Security and Medicare make up a significant portion of US spending, and they're paid for, in part, by separate payroll taxes, these programs should probably not be included in these calculations.

As a result, individual taxpayers pay only 56% of the cost of NIH and foreign aid. However, even this doesn't tell the full story. Because the federal income tax is progressive, not everyone who pays federal income tax pays for the same share of government.

Proportion of Income Taxes Paid and Discretionary Spending by Income

Annual Income Range	Proportion of Total Income Taxes Paid	Contribution to Discretionary Spending (%)
$0 - $25,000	0.4%	0.2%
$25,000 - $50,000	3.5%	1.9%
$50,000 - $75,000	5.4%	3.0%
$75,000 - $100,000	5.8%	3.2%
$100,000 - $200,000	19.5%	10.9%
$200,000 - $500,000	22.2%	12.4%
$500,000 - $1,000,000	11.5%	6.5%
$1,000,000+	31.8%	17.8%

Note: Percentages are author's calculation based on Table 3.5 (2022 version) from IRS Statistics site - https://www.irs.gov/statistics/soi-tax-stats-individual-statistical-tables-by-tax-rate-and-income-percentile

Proportion of Discretionary Spending Paid by tax filers, based on income

Author's calculation

Accounting for all of this, taxpayers contribute much less to NIH and foreign aid than they might first think. The table below lays out the per tax-filer income tax contribution to these programs based on income level.

Total Contribution to Various Discretionary Programs, based on income

Income Range	Total Foreign Aid	USAID	NIH
$0 - $25,000	$3.01	$1.85	$1.97
$25,000 - $50,000	$37.57	$23.10	$24.56
$50,000 - $75,000	$90.60	$55.70	$59.22
$75,000 - $100,000	$153.50	$94.37	$100.34
$100,000 - $200,000	$303.02	$186.28	$198.08
$200,000 - $500,000	$891.78	$548.21	$582.94
$500,000 - $1,000,000	$2,777.15	$1,707.23	$1,815.38
$1,000,000+	$15,938.36	$9,797.99	$10,418.68

Total program costs can be found at previous post on subject. https://chrisoldman784482.substack.com/p/making-up-for-doge

As background given the Trump EPA's reconsideration of Biden's Social Cost of Carbon update, here are some good explainers on the topic.

• Maxim Lott - January 2023 - What's the real cost of CO2?
• Alexander Frei, Kevin Dayaratna and Austin Gae Heritage - January 2025 Calculating the “Social Cost of Carbon” with the GIVE Model: An EPA Model Not Ready for Prime Time
• James Broughel (CEI) - September 2024 - The Social Cost Of Carbon Is A Flawed Metric For Policy Decisions
• Travis Fischer CATO - February 2024 - The Political Economy of EPA’s Updated Social Cost of Carbon
• Mario Lewis, Jr. (CEI) - June 2023 - Elephant in the elevator: How government manipulates the social cost of carbon to justify regulations
• Kevin Dayaratna, Patrick Michaels, Marlo Lewis (Heritage) - June 2021 - Biden Administration Is Skewing Carbon Numbers to Push Regulations
• Patrick Michaels CATO - February 2017 - At What Cost? Examining The Social Cost of Carbon
• Patrick J. Michaels and Paul C. “Chip” Knappenberger - CATO - June 2016 - You Ought to Have a Look: The Hows and Whys of the Social Cost of Carbon
• Jason Scott Johnston CATO - Spring 2016 - The Social Cost of Carbon

In what should come as no surprise, Republicans are warming to the expanded ACA that was designed, implemented, and already extended once by Biden along with the Democratic-controlled Congress. I recently wrote about how our elected officials have lost the ability to exhibit the responsibility and fiscal restraint that existed just a few decades ago, and this is yet another example.

These subsidies shouldn't be expanded for six reasons:

• The ACA was designed to increase subsidies as premiums rose and set definitions of affordability that ensured low-income Americans could afford to purchase insurance no matter the price.
• The ACA focused on people above the poverty line, and paid subsidies to people earning much more than the poverty level, even beyond the median household income.
• There is no reason that the affordability thresholds set by the ACA were too high and need to be reduced.
• Expanded subsidies have not led to uniform gains in coverage at all affected incomes, and the apparent gains are questionable.
• Increased subsidies will generally increase premiums, which will end up hurting taxpayers and people ineligible for subsidies while benefitting insurance companies.
• As always, given debt levels, spending more nonexistent money to make, at best, minor enhancements to a program with many other problems is unjustified.

ACA Already Set Subsidies to Make Health Care Affordable

One of the major features of the Affordable Care Act was the subsidies it paid to people earning too much to qualify for Medicaid, to be used to purchase insurance on the exchanges. The subsidies were designed on a sliding scale, to bridge the gap between what the ACA determined people could afford, and the current cost of insurance. The lower a person's income, the higher the subsidy. The higher their income, up to 4 times the federal poverty level, the lower the subsidy.

According to the original legislation, which Obama signed and every Democrat lauded for a decade, every American, regardless of their income level was expected to pay at least a little bit for their health insurance. For people making less than 133% of the Federal Poverty Level (FPL), Obama and Democrats believed contributing at most 2% of their monthly income was affordable. For a single-person right at the poverty line, that would amount to $20/month. For a family of four, less than $40/month. The government/taxpayers would pay the rest via a subsidy.

What often gets lost in discussions about ACA subsidies is that while these subsidies go to individuals and families who are low-income, they are not going to anyone living in poverty, by the government's definition. "% FPL" is a term used frequently in policy circles, but it's important to remember what it signifies--how much more income a person makes compared to someone living in poverty. The ACA, as originally legislated, focused its subsidies on Americans earning above the poverty level but no more than four times the poverty level.

In 2014, the median household income was $53,657 (2014 $), but subsidies were available to households up to 4 times the federal poverty level, which would amount to $71,040 for a household of 2.5 people (the national average). The diagram below shows how the subsidy availability compares to the income scale. In 2014, 300% FPL was pretty close to the median household income, meaning that subsidies were available well into the middle class of Americans.

One of the smart elements of the ACA was the definition of affordability was in terms of percentage of income. This produced several benefits. For one, it would prevent medical inflation from affecting the subsidy. For instance, if health care costs doubled, and premiums consequently doubled, while incomes remained stagnant, then the subsidy would cover the entirety of the premium inflation.

Defining affordability as the maximum percentage of income people should be expected to pay for health insurance meant that annual updates would be unnecessary and affordability would incorporate both the increase in health care premiums and changes in income. There would be little need to change it going forward. Beyond that, Democrats have been celebrating the slow growth of ACA premiums.

However, in 2021, for no particular reason, Biden and Democrats modified the thresholds so that, at every income level, the expected contribution of people to their health care premiums would fall, for most people by more than 50%, and for a large proportion, so that they were no longer expected to contribute anything. The expanded subsidies were only supposed to last through 2022, presumably covering the pandemic, but in the Inflation Reduction Act, the Democrats extended through 2025.

Affordability Redefined - Changes in Expected Contribution by Income Level

The table above shows the amount that people at various income levels were expected to contribute under the original ACA and under the expanded subsidies. President Obama and the Democrats of 2010 considered $25/month affordable for people just above the poverty threshold, and a gradually increasing amount for people making several multiples of that threshold. President Biden and the Democrats of 2021 substantially redefined affordability. In fact, Biden cut the expected contributions more if you were higher income than if you were lower income. While that may not be clear based on percentages, reducing contributions 4% for a higher income is a greater dollar reduction than for a lower income.

Also, bear in mind that these expected contributions are based on a Silver plan, which has higher premiums and lower out of pocket costs than a Bronze plan. Given Bronze plans are generally cheaper, in most cases, they would already be available for zero premium dollars for people between 100 and 200% of the federal poverty level.1

Like much of the other spending in the ARPA (for instance union bailouts, school slush funds, and an additional stimulus check), the expanded subsidies were not really necessary and were insufficiently debated and discussed, but got smuggled through because on their face, they seemed reasonable given the pandemic.

Subsidies Available to Way More Than Most Needy

Subsidies Raise Prices and are Captured by Corporations not Consumers

History has shown that subsidizing consumers often leads to increases in the list prices of the product. Sometimes, the increase in price completely offsets the subsidy meaning the money intended to help consumers ends up having no effect on their well-being at all, but goes to the company selling the product instead. This relationship has been borne out in education, housing, farm rental, child care, and solar panels.

Generally, this effect is more pronounced when supply of a good is limited, such as with housing or child care. In this case, while insurance products themselves are effectively unlimited, the premiums eventually go to a product that is limited - health care. To the extent that coverage leads to unnecessary utilization, it will lead to price increases for providers and premium increases for everyone.

Gains in Insured Population Are Questionable and Uneven

Since the enhanced subsidies, total ACA enrollment has nearly doubled, but looking more closely, gains have not been uniform across all the income levels. 60% of the growth in enrollment came from the 100-150% FPL and 71% of the growth came from people earning less than 200% of the FPL. On a per-beneficiary basis, these were the smallest changes in subsidy, but for these people their benchmark premium fell to zero. If the goal is to maximize enrollment, then dollars are best spent on the lowest income participants. This has the secondary benefit of being more in line with a system that focuses its social safety net on those most in need.

71% of the growth in enrollment came from the people who received the smallest boost in subsidies.

The Paragon Health Institute conducted its own review of the gains and compared to the number of people who would be eligible for subsidies at these income levels. They found that for many states, there was greater growth in enrollment than there were people in those states. They also found that reported incomes more often than would be expected, exactly matched the income thresholds that would warrant the subsidies. Consequently, they concluded that a 5 million of the newly insured were misstating their income to receive the subsidies. Before extending these subsidies, the government should redouble efforts to make sure only the people who deserve them are receiving them.

Further, there's reason to question the CMS enrollment numbers themselves. According to those numbers, ACA enrollment from people earning between 100 and 199 of the federal poverty level has more than doubled, adding millions of new beneficiaries in this group. The National Health Insurance Survey, though, shows no such growth in private ACA enrollment through 2023, the most recent available.

Insurance Coverage Levels for 0-64 year olds between 100-199% of FPL

Source: Lynn A. Blewett, Julia A. Rivera Drew, Miriam L. King, Kari C.W. Williams, Daniel Backman, Annie Chen, and Stephanie Richards. IPUMS Health Surveys: National Health Interview Survey, Version 7.4 [dataset]. Minneapolis, MN: IPUMS, 2024. https://doi.org/10.18128/D070.V7.4

The Ever-Present Debt

Lastly, of course, is the same argument against all new spending--there is no money to spend. Deficits are increasing more every year and will continue to for decades. New spending today will mean some combination of reduced services or higher taxes for the next generation. As with all spending, if you're in favor of this, if you think it is important and substantially beneficial, it should be no problem to name something that is less important, that is less beneficial, that is a worse use of money. Unless that can be named, and money shifted, then new spending should not be added.

Put slightly differently, and others don't report it this way, but every deficit dollar we spend today, is actually a commitment to spend more than double that over the next twenty years. It's a commitment to spend 7x that over the next 50 years, because every dollar we spend we have to borrow, and we never pay it back. So year after year, we're paying interest. Imagine you took out a 30k loan for a car that you didn't have to payoff for fifty years, and you never paid back the principle. You'd end up spending $213,000 over the fifty years and still owe more. The CBO predicts a baseline interest rate of 4% going forward, but there are many reasons to believe that that interest rate will increase as our debt increases and the United States falls deeper and deeper into debt.

As worrisome as the graph above looks, the reality will probably be worse because of the likelihood of higher interest rates as the debt mounts but also because CBO forecasts don't account for surges in spending due to recessions or spendthrift leadership.

Don't Extend the ACA Subsidies

There are plenty of reasons to be circumspect about the ACA subsidy expansion. Unfortunately, it's a lot easier to avoid careful review and not worry about the budget, the effectiveness of your spending, or fraud but bask in the adulation you get from blindly spending other people's money on a problem people wish was fixed.

Notes

1On average, Bronze plans' premiums are around $100 lower than Silver plans' premiums, so if your expected contribution is below $100 for the second-lowest cost silver plan, the subsidy you receive will be greater than the cost of the bronze plan.