Share

*Findings from the previous Polygraph newsletter were featured in this Forbes op-ed and this Yale debate. Thank you to my paid subscribers for making that research possible.

*Big thanks to Heath P. and Ethan R. for becoming paid subscribers! Join them and the other Polygraph VIPs to support my work.

Subscribe now

Situation

President Trump’s budget proposal for next year manages to be $361 billion larger than last year’s despite cutting $300 billion in social programs and other public investments. The main reason is the vast increase in military spending.

The $1.504 trillion military budget serves as the lifeblood for Trump’s next war(s) and his justification for slashing social programs. “We’re fighting wars,” Trump said earlier this month. “It’s not possible for us to take care of daycare, Medicaid, Medicare, all these individual things…we have to take care of one thing: military protection.” The White House tried and failed to message its way out of the resulting public backlash. A few days earlier, Secretary of State Marco Rubio said, “Imagine if instead of spending billions of dollars supporting terrorists or weapons, Iran had spent that money helping the people of Iran.”

US infrastructure for drinking water and wastewater received C– and D+ grades respectively from the American Society of Civil Engineers last year. Next year’s budget proposal slashes funding for water infrastructure. For the Pentagon, a river of money; for workers, a river of effluent.

What $1.5 trillion could fund instead

Federal agencies released detailed budget justifications early this month in time with the White House’s overarching budget, but the Pentagon’s was delayed until a few days ago. Why? Budgeting beyond WWII-level military spending when there’s nowhere near WWII-level military demand has created a glut, an overabundance, un embarras de richesses — a fact made clear in this actual headline in the Washington Post: “Trump aides struggle with how to spend $500 billion more on military.” A $1.5 trillion budget is more than the Pentagon knows what to do with.

One way to express the severity of this failure of governance is to illustrate what $1.5 trillion could accomplish when put toward other ends. Oddly, doing so makes the Pentagon’s “problem” somewhat relatable — $1.5 trillion really is a hell of a lot of money. The graph below shows what I did with the equivalent of Trump’s proposed 2027 military budget. Globally, I eliminated hunger and extreme poverty; domestically, I ended homelessness, covered every person without health insurance with Medicaid, implemented universal pre-K for three- and four-year olds, issued another round of $600 relief checks, and built enough solar infrastructure to power half of all households. I still have $77 billion left over.

What’s more, while the massive, trillion-dollar-plus annual military budgets like Trump proposed for 2027 are “the new normal,” the $596 billion for solar and $141 billion for relief checks are one-time costs. Only $690 billion is for yearly expenses (though the $93 billion for world hunger isn’t planned to be needed after 2030).

Methodology is below the chart.

^Alt text for screen readers: What $1.5 trillion for the Pentagon could fund instead. This column chart compares the $1.504 trillion requested military budget for 2027 against a set of alternatives. In billions: 93, End world hunger; 318, End global extreme poverty; 29, End US homelessness; 214, Medicaid for uninsured; 35, Universal US pre-K; 141, Send $600 checks to 235M Americans; 596 , Power half of US households with solar. Proposed and estimated costs are annual, except solar (overnight capital cost) and $600 checks (matching cost of January 2021 relief payments). Data: President’s Budget for fiscal year 2027, author analysis.

Methodology

Power half of US households with solar

There are 134.8 million US households (as of Dec 2025) and 50% of that is 67.4 million. On average, each household uses 10,500 kWh of electricity per year, creating an annual electricity demand of 707.7 billion kWh.

Assuming a 25% capacity factor, 1 kW of utility-scale solar produces an average of 2,190 kWh per year (0.25 times 8,760 hours in a year). Dividing 707.7 billion kWh by 2,190 kWh per kW equals 323.1 million kW, the estimated utility-scale solar capacity required to meet household demand.

According to the Department of Energy’s PV System Cost Model, the estimated cost per kW is $1,488 to $2,200 for MSP. I chose the midpoint — $1,844 — for this analysis. Multiplying 323.1 million kW by $1,844 equals $595.8 billion, the estimated one-time cost to build the solar capacity to power half of US households.

Note that this estimate includes energy storage, allowing solar power to be delivered at night rather than only matching aggregate electricity demand. DOE’s model calculates the cash cost of an installed system by factoring in the following expenses:

• Module – The cost to the installer of photovoltaic modules, as delivered.

• Inverter – The cost to the installer of equipment for converting direct current (dc) to alternating current (ac), as delivered.

• Energy Storage System (ESS) – The cost to the installer of adding an energy storage system, as delivered.

• Structural Balance of System (SBOS) – The cost to the installer of structural balance of system components, as delivered.

• Electrical Balance of System (EBOS) – The cost to the installer of electrical balance of system components, as delivered.

• Fieldwork – The cost to the installer of work performed at the installation site.

• Office work – The cost to the installer of work performed off-site.

• Other – Costs incurred by the project developer not included elsewhere.

Send $600 checks to 235M Americans

Division N of Title II of the Consolidated Appropriations Act, 2021 (P.L. 116-260), provided $600 “recovery rebates” to eligible taxpayers, providing $141 billion in tax relief, according to IRS data. There were 146.5 million second-round tax rebate payments, including 41.7 million to married couples filing a joint return. All told, the $600 checks reached about 235 million Americans, including dependents (eligible filers also received $600 per child).

Universal US pre-K

The simple annualized cost is $35.1 billion, based on Penn Wharton’s estimate (“We estimate that nationwide universal preschool education for three- and four-year-olds will cost $351 billion over the next 10 years.”)

Medicaid for uninsured

There are 27.1 million Americans without health insurance, according to the Census Bureau (2024). Average annual Medicaid spending per full-benefit enrollee is $7,909, according to KFF (2023). Multiplying these two numbers together gives you $214.4 billion.

End US homelessness

The National Alliance to End Homelessness (NAEH) estimates it would cost $15.1 billion to fully address sheltered homelessness through a Housing First policy. This policy is about getting people without a home into one as quickly as possible and for as long as needed. Optional support services are offered as well.

My estimate is nearly double NAEH’s for two reasons. First, the NAEH figure is also calibrated to 2022, and homelessness has rapidly increased since then. I accounted for the rise in homelessness using the latest population data from the Department of Housing and Urban Development (HUD). HUD’s point-in-time counts show sheltered homelessness rising from 348,630 in 2022 to 497,256 in 2024 — a 43% increase — yielding an adjusted cost estimate of roughly $21.5 billion to cover the current sheltered population. Second, the NAEH figure applies to sheltered homelessness but not unsheltered.

To account for unsheltered homelessness, I adjusted the $21.5 billion figure upward by 36%, matching the share of people experiencing homelessness who are unsheltered (274,224 out of 771,480). This yields an estimated $29.1 billion annual budget.

This is a rough estimate. Combining the strongest available cost model with the most recent population data requires bridging different measurement systems: the NAEH cost model is based on households counted over the course of the year while the HUD data counts individual people at a single point during a year. As a result, there are simplifying assumptions, like household homelessness increasing proportionately with individual homelessness and per-household costs remaining consistent across time and populations.

End global extreme poverty

A recent working paper estimates that the cost of ending extreme poverty through direct cash transfers is $318 billion, or roughly 0.49% of OECD GDP.

The US would not be expected to front the entire cost. If the US financial contribution matched its contribution to the UN regular budget (22% in 2025), the annual cost to the US would be $70 billion.

End world hunger

Ending world hunger by 2030 would cost $93 billion per year, according to UN estimates.

The US wouldn’t be expected to front the entire cost. If the US financial contribution matched its contribution to the World Food Programme (32% as of 2025), the annual cost to the US would be $29.5 billion through 2030.

SPECIAL THANKS TO: Abe B., Alan F., Alexander L., Alissa Q., Amin, Andrew R., AT., B. Kelly, Barbara B., Bart B., BeepBoop, Ben, Ben C.,* Bill S., Bob N., Brett S., Byron D., Carol V., Catherine L., Chris, Chris G., Claudia, Cole H., Coleman J., D. Kepler, Daniel M., Dave, David J., David S.,* David V.,* David M., Dharna N., Elizabeth R., Emily H.,* Errol S., Ethan R., Foundart, Fran Q., Francis M., Frank R., Fred R., Gary W., Gladwyn S., Graham P., Griffin R., Heath P., Hunter S., IBL, Irene B., Isaac, Isaac L., Jacob, James G., James H., James N., Jamie LR., Jcowens, Jeff, Jennifer, Jennifer J., Jessica S., Jerry S., Joe R., John, John, John A., John K., John M., Jonathan S., Joseph B., Joshua R., Julia G., Julian L., Katrina H., Keith B., Kesh L., Kheng L., Lea S., Leah A., Leila CL., Lenore B., Linda B., Linda H., Lindsay, Lindsay S.,* Lora L., Lou B., Mapraputa, Marie R., Mark L., Mark G., Marvin B., Mary Z., Marty, Matthew H.,* Megan., Melanie B., Michael S., Mitchell P., Nick B., Noah K., Norbert H., Omar A., Omar D.,* Peter M., Phil, Philip L., Ron C., Rosemary K., Sari G., Scarlet, Scott H., Silversurfer, Soh, Springseep, Stan C., TBE, Teddie G., Theresa A., Themadking, Tim C., Timbuk T., Tony L., Tony T., Tyler M., Victor S., Viviane A., Wayne H., William H.,* William P.

* = founding member

-Stephen (Follow me on Instagram, Twitter, and Bluesky)

Subscribe now

Share

Share

ICYMI: US spends $2.1 billion a day on the Iran War

Thank you, Lou B. and Viviane A., for becoming paid subscribers! Join Lou, Viviane, and the other Polygraph VIPs thanked at the bottom of each note to support my work.

Latest newsletter for VIPs: Trump proposes largest military budget ever

Subscribe now

Trump’s budget proposal, in brief

The White House’s recent budget proposal outlines how the president wants to spend the public’s money next year. The gratuitous cuts to civilian agencies suggest the goal is smaller government. But that’s not the case — next year’s budget would still be $361 billion larger than last year’s. Trump’s concern isn’t the size of government; it’s what the government does and who it does it for.

So what does Trump want the government to do in 2027? More war, more (militarized) policing, and less of everything else. Based on his budget request, Trump sees little point in government outside performing those first two functions. War and policing receive record funding; funding for all other government functions is cut by $300 billion. (I will refer to the latter spending category as “public investment” for the rest of the article.1) Slashing funding for the former and surging spending for the latter is how you budget for the 1%.

Trump’s budget proposal, a breakdown

War & policing

Of the $2.2 trillion Trump requested, 80% — $1.8 trillion — is tied to war and policing.

Most of that is the $1.5 trillion military budget, which would be the largest ever. Adjusted for inflation, Trump wants to spend $260 billion more on the military in 2027 than the US did in 1945. It isn’t World War II-level spending; it’s parody-level.

Another $282 billion comes from the Departments of State, Veterans Affairs, Justice, and Homeland Security2 (the Energy Department’s nuclear weapons funding is part of the $1.5 trillion figure). Only the parts of these agencies’ budgets linked to war and policing were counted. For example, the DHS budget includes disaster relief funding, but that funding isn’t included in the aforementioned $282 billion. Conversely, DHS funds ICE, and ICE’s budget is certainly included in that figure.3

Public investment

The remaining 20% of Trump’s budget proposal — $441 billion — is for government functions related to agriculture; community and regional development; education; health; housing and income assistance; international affairs; natural resources and environment; science, space, and technology; training, employment, and social services; transportation and more.

Funding for this array of purposes is labeled as “public investment” in the chart below. In 2025, public investment spending made up 40% of the enacted budget. Trump wants to halve that share, equal to a $300 billion cut.

Public investment is intended to be productive spending, designed to yield a return — better transit, more affordable food, healthier workers, etc. — to enhance economic growth or human security. Spending on war and policing does not share that aspiration. It’s a cost of enforcing order — less about managing disorder than enforcing a certain social order (at home) and geopolitical order (abroad), neither of which are sustainable.

If you’re looking for a reasonable balance between these two types of spending, you won’t find it in Trump’s budget request. You won’t find it in recently enacted budgets, either.

(Article continues after graph.)

^Alt text for screen readers: 80% of Trump’s budget goes to war and policing. Proposal halves funding for public investments, still increases deficit. The red-orange stacked column chart is as follows: Military, $1.5 trillion; Department of State military aid, $11.2 billion; Department of Veterans Affairs, $145.3 billion; Departments of Homeland Security and Justice, $125.4 billion; Other, $441.2 billion. Amounts: requested discretionary funding plus supplemental discretionary via reconciliation. Military: budget function 050. Agency amounts only include funding related to war and policing. Data: President’s Budget for fiscal year 2027, Office of Management and Budget.

The chart above will draw criticism for three main reasons:

1. I clearly tried to make the graph look like a Rothko painting but didn’t quite pull it off.

Fair enough :(

2. I categorized veterans care as war-related spending.

Why isn’t the VA budget considered a public investment? First, it’s a war cost. Second, the care it delivers is exclusive, not public. Third, like other war-related funding, it crowds out other priorities. I’m not suggesting the VA budget be cut like I have for the Pentagon; I’m arguing that, like Pentagon spending, its growth displaces available funding for social programs and other public investments.

Military and non-military spending are the two fundamental categories of the budgets Congress passes each year. Generally speaking, congressional Republicans aim to increase the former and cut the latter; congressional Democrats used to prefer the opposite, but now want to increase both.

Both parties are heavily invested in the fight over the non-military amount — what happens to that number is a major part of how they gauge their political success. Both parties badly want to report “wins” to their base on this issue.

Republicans and Democrats appear to have come to a tacit agreement in recent years: increase non-military spending for military-adjacent programs (like veterans care) and policing functions while cutting public investment funding by nearly as much. Under this arrangement, both parties end up winners (different story for workers). Republicans can claim victory because public investments were cut, having been squeezed out of the non-military budget. Meanwhile, Democrats can tout historic increases in non-military spending.

It’s this budgetary sleight-of-hand that allowed Democratic leaders to boast in December 2022 that their 2023 budget bill — the final spending bill passed under the Democratic trifecta — included “the highest level for non-defense funding ever,” while cutting public investment spending by $55 billion ($99 billion in 2026 dollars).

In 1980, veterans care took up 4% of the non-military spending in the discretionary budget, 7% in 2000, and 10% in 2020. It jumped to 15% under Biden. Including the relevant DHS and DOJ funding, 24% of ostensibly non-military spending in Biden’s last enacted budget (2024) went to military-adjacent, militarized, or policing programs — the highest share on record to that point. It’s 25% this year, and 33% in Trump’s proposed budget for next year.

Trump’s proposal didn’t fall from the sky. The graph below shows that where the president wants to take us we’ve been headed for a while. It’s as if Trump observed a preexisting defect in US politics, then proceeded to inject it with vast quantities of epinephrine and anabolic steroids. (Note: To more clearly present the gap between military and public investment spending discussed above, the graph excludes VA, DHS, DOJ, and DOS funding.)

^Alt text for screen readers: US budgets less for improving lives, more for ending them. Discretionary spending by function, constant 2026 dollars. This chart shows war spending increasing steadily since 2017 and shows public investment spending peaking in 2020 before steadily falling ever since. Trump’s 2027 budget includes $1.5 trillion in war spending and only $432.3 billion in public investment. Amounts reflect discretionary funding plus supplemental discretionary funding via reconciliation. Budget function 050 in war category; others minus 700, 750 in public investment. Data: Office of Management and Budget, Congressional Research Service.

3. The discretionary budget doesn’t capture all government spending.

Indeed, discretionary spending is the money that the president proposes and Congress enacts each year through funding bills.4 Not included in the chart above is standard mandatory spending — primarily payments for benefits programs that are funded automatically through formulas — or interest on the national debt. Including those other two, the total projected budget for 2027 is $8.3 trillion.

Viewed in the context of all federal spending, funding for war and policing appears far less overwhelming. Or does it?

With the additional war- and policing-related spending in mandatory accounts (mostly disability and other payments to veterans), the total climbs from $1.8 trillion to $2.2 trillion.

So while it’s 80% of the proposed discretionary budget but only 26% of the entire projected budget, that 26% is more any other government expenditure. Here are the top five largest sources of US spending across discretionary and mandatory accounts for 2027:

1. War and policing: $2.2 trillion

2. Social Security: $1.8 trillion

3. Medicare: $1.2 trillion

4. Interest on debt: $1.1 trillion

5. Medicaid: $1.0 trillion

Including its share of the interest on the national debt, there’s $2.5 trillion for war and policing in the 2027 budget — 29% of the $8.3 trillion total.

SPECIAL THANKS TO: Abe B., Alan F., Alexander L., Alissa Q., Amin, Andrew R., AT., B. Kelly, Barbara B., Bart B., BeepBoop, Ben, Ben C.,* Bill S., Bob N., Brett S., Byron D., Carol V., Catherine L., Chris, Chris G., Claudia, Cole H., Coleman J., D. Kepler, Daniel M., Dave, David J., David S.,* David V.,* David M., Dharna N., Elizabeth R., Emily H.,* Errol S., Foundart, Fran Q., Francis M., Frank R., Fred R., Gary W., Gladwyn S., Graham P., Griffin R., Hunter S., IBL, Irene B., Isaac, Isaac L., Jacob, James G., James H., James N., Jamie LR., Jcowens, Jeff, Jennifer, Jennifer J., Jessica S., Jerry S., Joe R., John, John, John A., John K., John M., Jonathan S., Joseph B., Joshua R., Julia G., Julian L., Katrina H., Keith B., Kesh L., Kheng L., Lea S., Leah A., Leila CL., Lenore B., Linda B., Linda H., Lindsay, Lindsay S.,* Lora L., Lou B., Mapraputa, Marie R., Mark L., Mark G., Marvin B., Mary Z., Marty, Matthew H.,* Megan., Melanie B., Michael S., Mitchell P., Nick B., Noah K., Norbert H., Omar A., Omar D.,* Peter M., Phil, Philip L., Ron C., Rosemary K., Sari G., Scarlet, Scott H., Silversurfer, Soh, Springseep, Stan C., TBE, Teddie G., Theresa A., Themadking, Tim C., Timbuk T., Tony L., Tony T., Tyler M., Victor S., Viviane A., Wayne H., William H.,* William P.

* = founding member

-Stephen (Follow me on Instagram, Twitter, and Bluesky)

Subscribe now

Share

Subscribe now

*This note is for Polygraph VIPs only. Read on if you are one, or support my work to become one.

Read more

Share

Big thanks to Fran Q., Polygraph’s latest VIP member! Join Fran and the other paid subscribers thanked at the bottom of each note if you’d like to support my work.

Latest VIP newsletter: Iran War: Cost of the first week

Subscribe now

How much has the US spent on the Iran War?

The Trump administration won’t say and prefers the cost not be discussed.

The US spent an estimated $28.7 billion in the first two weeks of the Iran War, or $2.1 billion a day on average. This is based on my analysis of officials’ statements, federal procurement and operations data, and reporting on military deployments and armament use. This estimate refers only to direct war costs — near-term expenses for military operations, munitions, and the like — and not indirect costs, which include broader and longer-term factors like economic impact and veterans’ care.

This might be a higher estimate than you’ve seen elsewhere. Those estimates are too low. This one could be too.

Below:

• Summary table

• Why most estimates are too low

• Why this estimate could be too low

• Methodology and itemized war costs

^Alt text for screen readers: US spends $2.1 billion a day on the Iran War. Estimated US war costs, 28 February–13 March 2026. Table format: Expense; Description; Cost. Rows: Operations; Mobilization, administration, combat; $4.25 billion. Weapons; Missiles, interceptors, bombs, other; $20.77 billion. Losses; Damaged or destroyed military assets; $2.71 billion. Subsidies; Pay for Israel’s bombs, interceptors; $1.01 billion. Total; Cost of the first two weeks of the Iran War; $28.74 billion. Average cost per day: $2.05 billion. Estimated direct costs based on author’s analysis of officials’ statements, procurement and O&S data, open-source intelligence, media reports. Figures subject to revision as new information becomes available.

Why most estimates are too low

In a recent closed-door briefing with select lawmakers, Pentagon officials estimated the war had cost $11.3 billion in the first week. That figure, despite not accounting for much besides munition expenses, is the closest the Trump administration has come to saying how much it’s spending on the Iran War. Mainstream media outlets repeat the figure over and over, as their compulsive need to cite figures from government officials clashes with officials not giving any. It’s also the basis for popular cost estimates from think tanks.

For example, CSIS originally said the war had cost $3.7 billion over the first 100 hours, which became one of the most widely cited estimates, though probably not for the reason CSIS intended. In the media, the chasm between the CSIS and Pentagon estimates became a popular trope for how the cost of the Iran War was smashing expectations.1 CSIS then released another estimate — this one built pretty much entirely around the Pentagon’s — which became popular in its own right, even serving as the basis for estimates by news outlets and other think tanks.

The prevailing war cost estimates have two things in common. First, they’re all better than the dozens of AI-generated war cost trackers, which are empty calories empirically. Worse, they’re unaccountable. Numbers wrong? Blame the model. There’s no one to hear your justifiable criticism; it’s like being cut off in traffic by a driverless car.

Second, because the prevailing cost estimates are ultimately based on the Pentagon’s, they reproduce its inherent flaws. The biggest is munition expenditures (which is pretty much all the Pentagon’s estimate included, given all the other types of war costs it left out). The result is a pervasive understatement of how expensive the Iran War actually is.

The Pentagon tracks munition “burn rate” costs by merging operational logistics (what’s being fired in what quantities) with financial values (what each one costs). My beef is with how it determines unit costs. I don’t think the Pentagon’s method is wrong; I do think it’s irrelevant — so immersed in the internal logic of departmental bookkeeping that its conclusions are too insular to be of much use. The war costs that show up on the Pentagon’s ledger are different (read: lower) than the war costs the US public will be asked (read: forced) to pay.

For example, the Navy logs the number of SM-2 interceptors it’s firing into the Ordnance Tracking System; that quantity gets fed into the Navy Enterprise Resource Planning system, which multiplies the reported quantity by the SM-2’s current unit cost; that figure goes into the Agency Financial Report, which eventually gets rolled up into the Pentagon’s consolidated financial statements. The cost to fire one SM-2 is based on something called Moving Average Cost, which determines unit cost by dividing the total value of the SM-2’s stockpile by the number of SM-2s in it. The value of a munition’s stockpile is a historical average that changes with every new purchase of that munition. If its unit cost is $1.1 million and 50 are withdrawn from the stockpile and fired, the Pentagon’s ledger would show $55 million in expenses from consuming those munitions. The war cost estimate unnamed Pentagon officials gave lawmakers likely reflects this accounting method, as it’s the department’s standard.

How does this standard lead to underestimates of war costs? The aforementioned interceptor illustrates the mechanics perfectly: The US military has reportedly fired lots of SM-2 interceptors during the Iran War, but the Pentagon doesn’t buy SM-2s anymore; they’re being phased out. The SM-6 — the more advanced and expensive successor — is purchased instead. So each time an SM-2 is fired in the ongoing war, the Pentagon’s accounting system registers a cost closer to the SM-2’s $1.1 million unit cost from 2011 than the $5.3 million unit cost in 2026 for the SM-6 that will replace it. For US taxpayers, the cost of firing 50 SM-2s isn’t $55 million; it’s $265 million.2

Some prevailing estimates have used replacement costs, but for estimates they designed to fit within the Pentagon’s $11.3 billion figure, which reflects Moving Average Cost. Using replacement costs doesn’t correct much when you’re working within a vastly understated pot of funds.

Why this estimate could be too low

Three reasons, among several others:

1. US military assets and personnel poured into the Middle East and Europe as part of a historic buildup ahead of the war. I haven’t yet accounted for the prepositioning of assets, and my estimate for personnel is probably too low. The only pre-war personnel cost I included is for activating reservists, which reflects mobilization at the scale of last year’s war with Iran, adjusted for the 3.8% troop pay increase in 2026. For the June 2025 war, the US mobilized 17,193 reservists — 12,396 deployed to CENTCOM and 4,797 to EUCOM — totaling $1.96 billion, based on my analysis of Army, Navy, Marines, and Air Force budget documents. This money came from the $8 billion war fund made available by Sec. 1421 of P.L. 119-4 and executed on 16 June 2025, a day after Pentagon chief Pete Hegseth announced the US was surging forces into the Middle East in a “defensive” maneuver. The US bombed Iran five days later (using funding from Biden’s Israel aid bill — one of the more obscure reasons why concern over Palestine doesn’t make one a single-issue voter). I haven’t yet found data indicating specifically how much higher my 2026 estimate should be.3

2. The single most expensive military asset reportedly damaged or destroyed by Iranian forces is a $1.3 billion AN/FPS-132 early-warning radar at a US base in Qatar. Because the radar was originally funded by a US foreign military sale to Qatar, I did not include any potential restoration costs for it in my estimate. Methodologically, this is a questionable decision, as it assumes Qatar has the appetite to pay the US to replace a billion-dollar radar destroyed in a war the US basically started for fun. (It certainly wasn’t out of necessity.)

3. Pentagon chief Pete Hegseth said the ongoing war “has seven times the intensity of Israel’s previous operations against Iran during the 12-day war.” If that translates to anywhere near seven times the cost, I’ve badly underestimated US war costs.

Israel spent an estimated $6.7 billion on the 12-day war in June 2025, based on my analysis of Israeli budget documents. In September 2025, the Knesset approved a supplemental bill to cover military costs related to its June war with Iran and its planned “conquering” of Gaza in May. A table from the explanatory section of the bill (below) shows 22.7 billion shekels for the military (lit. “Ministry of Security”), 245 million for national insurance allowances, 6.2 billion for intelligence agencies, and 1.7 billion for interest payments. Ignore the rightmost column with budget codes.

Including the funding for intelligence agencies, there were 28.94 billion shekels in total military spending in the bill, including 6.4 billion for IDF personnel — the bill increased personnel expenses (“הוצאות כח אדם”) to 39.02 billion shekels, up from the 32.66 billion in the original 2025 budget. The Israeli government attributed the additional personnel funding to the May–August 2025 Gaza offensive and the remaining 22.5 billion shekels — $6.7 billion — to the war with Iran.4 This is plausible. The original 2025 budget assumed Israel’s March assault (during the purported ceasefire) but likely did not account for the full costs of its May assault (also during the purported ceasefire). Israel called up tens of thousands of reservists for the May operation, which was plagued by setbacks and ultimately failed, hinting at significant overruns in personnel costs. It’s also plausible that the bulk of the supplemental funding was for the June 2025 war, which quickly went over budget due to operations (including the tens of thousands of flight hours for bombing missions and Mossad actions in Iran), munitions (like the thousands of bombs Israel dropped on Iran), and firing an estimated $1.25 billion in interceptors.

If by “seven times the intensity” Hegseth meant seven times the cost of the June 2025 war, my estimate is about $20 billion too low.

Methodology and itemized war costs

Operations

Estimated cost: $4.25 billion

The Congressional Budget Office (CBO) has data showing how much it costs to operate various combat units. I estimated the type and quantity of units participating in the Iran War by referring to open-source intelligence and media reports. For example, using USNI’s fleet tracker, I counted 15 destroyers operating in the Arabian Sea, Persian Gulf, Red Sea, and Eastern Mediterranean at the start of the war. Each costs $4.2 million per week to operate.

However, CBO’s data is calibrated to peacetime and not wartime, and war makes everything more expensive, including war.

To estimate how much more expensive war makes military operations, I analyzed operating (specifically, Operation and Support) spending in the regular or “base” Pentagon budget and in the supplemental war budget from 2003–2014. To isolate the proportional increase in operating costs attributable to war, I did two things. First, I measured wartime operations spending relative to base operations spending rather than in absolute terms to help control for cost growth not directly related to war (base operations spending rose by roughly 13% above inflation from 2003–2014). Second, I measured operations spending on a per–active-duty troop basis to help control for cost growth solely attributable to the scale of the Iraq and Afghanistan wars.

War increased operating costs by 33% on average from 2003–2014, which translates to a 1.33 wartime operating cost multiplier that I used to convert the CBO data from peacetime to wartime costs. Although the multiplier is derived from data for a particular time (2003–2014) and place (Iraq, Afghanistan), it provides a benchmark for how combat has historically increased operating costs. I suspect this multiplier is low, based on my more granular research. To name just one example, the inflation-adjusted increase in operating costs for three helicopters — CH-47D, UH-60L, and AH-64D — from 1998 (peacetime) to 2007 (wartime) was 140%, 262%, and 541%, respectively.

Weapons

Estimated cost: $20.77 billion

The weapons itemized below are grouped into offensive and defensive munitions. I make this distinction because each group requires its own methodology.5

Offensive

The last time the Pentagon provided a figure on the number of munitions it dropped or fired during the war was on March 3, when CENTCOM Commander Bradley Cooper said the US had “struck nearly 2,000 targets with more than 2,000 munitions.” On March 4, Chairman of the Joint Chiefs of Staff Dan Caine told the press, “I know there have been a lot of questions about munitions…I want to tell you, teammates,6 as a matter of practice, I don’t want to be talking about quantities…we consider it an operational security matter.” You know the Trump administration really doesn’t want people talking about war costs when they deny transparency into the vaguest of metrics.

However, the Pentagon discloses the number of targets it has struck, albeit imprecisely — “over 6,000” as General Caine said in a March 13 briefing — which can be used to infer munition expenditures.

How do you estimate munition costs based on the number of targets alone?

First, guess what General Caine meant by “over 6,000” targets. My guess: 6,400.

Second, multiply the estimated targets by the average number of munitions US and/or NATO forces used per target during a recent war. Or three wars, if you’re a show-off with poor time management.

• Operation Iraqi Freedom (first 30 days only), Iraq, 19 Mar 2003–18 Apr 2003

  • 29,199 munitions released / 19,898 targets struck = 1.47 munitions/target*

  • *Adjusted to 1.01 munitions/target to account for non-precision-guided munitions, though this might be an over-correction.7

• Operation Unified Protector (NATO), Libya, 31 Mar 2011–31 Oct 2011

  • 8,112 munitions released / ~6,000 targets struck = 1.35 munitions/target

• Operation Inherent Resolve, Iraq & Syria, 16 Oct 2014–20 Jan 2017

  • 65,461 munitions released / 39,608 targets struck = 1.65 munitions/target

Average of these wars’ average munitions per target = 1.36 munition/target. 6,400 targets times 1.36 munitions per target = 8,704 munitions.8

Of these 8,704 munitions, I assumed a 35-65 split between the very expensive, long-range, “stand-off” munitions used en masse during the early days of the war and the less expensive, short-range “stand-in” munitions the US has used more frequently since. The estimated split is based on officials’ statements.

Stand-off munitions

Quantities were estimated based on media reports, open-source intelligence, historical data, and forensic analysis of arms and munitions use. Munitions (unit cost) include: Tomahawk ($2 million), ATACMS ($1.6 million), PrSM ($2.5 million), AGM-88 HARM/AARGM/AARGM-ER ($1.9 million for the AARGM-ER, which replaces the AARGM, which replaced the HARM) JASSM ($2.9 million), JSOW (shares JASSM’s $2.9 million unit cost, as JASSM replaces the JSOW. However, I halved the associated cost to account for JASSM having twice the payload of JSOW.)

This includes 725 Tomahawk missiles, matching the rate at which they were fired during the first two weeks of Operation Iraqi Freedom — approximately 400 were fired during the first four days of both the 2003 Iraq War and 2026 Iran War. Other estimated quantities: 75 ATACMS, 55 PrSM, 1,016 JASSM, 600 AGM-88, 575 JSOW.

Stand-in munitions

The estimated quantities9 of the following guided bombs are based on the distribution of guided bombs dropped during the first 30 days of the 2003 Iraq War (2,000lb bombs, 36%; 1,000lb, 17%; 500lb, 47%).

• 2,000lb GBU-31 (MK84/BLU-109 warhead + JDAM)

  • Quantity: 1,728

  • Unit cost: $107,266 (warhead $39,250 + JDAM $68,016)

• 1,000lb GBU 32 (MK83/BLU-110 warhead + JDAM)

  • Quantity: 846

  • Unit cost: $78,923

• 500lb GBU 38 (MK82/BLU-111 warhead + JDAM)

  • Quantity: 2,248

  • Unit cost: $74,975

Estimated Hellfire expenditures were based on media reports and open-source intelligence.

• Hellfire missile

  • Quantity: 400

  • Unit cost: $249,506 (JAGM replacement cost, as JAGM is replacing the Hellfire)

These figures exclude the unspecified but reportedly large number of LUCAS one-way drones fired (~$35,000 each) as they were likely used predominately to saturate enemy defenses rather than destroy specific targets (Cooper likely excluded these drones from his munitions count on March 3). Also excluded are many of the “tens of thousands of pieces of ordnance” General Caine said were dropped on March 2.

*See more cost data here.

Defensive

Counter-missile

• Number of missiles Iran fired: 1,224

• Share of Iranian missiles that survived the boost phase: 93%

  • During the June 2025 war, 93% of Iranian missiles (546/588) survived the boost phase.

• Share of surviving missiles considered threats and engaged by American, Israeli, or Gulf air defenses: 81%

  • The share was 62% during the June 2025 war, but with the reduced distance between Iran and US bases in the Gulf — which reduces the time for determining whether a missile is a threat — I assumed that half the missiles assessed as non-threats in 2025 are considered threats in 2026.

• Missiles US and US-friendly forces attempted to shoot down: 922

• Average number of interceptors (missiles used to shoot down incoming missiles) used to shoot down each Iranian missile: 1.7

  • Although 1.1 were used per Iranian missile in the June 2025 war (361 interceptors for 339 missiles), US forces averaged two interceptors for each Houthi missile in 2024–25, and used 30 Patriot interceptors to shoot down 13 Iranian missiles fired at al-Udeid air base that Iran warned were coming.

• Interceptors fired at Iranian missiles: 1,567

• Share of interceptors fired by US forces: 86%

  • The US fired 72% of the interceptors during the June 2025 war (Israel fired 101; the US fired 230 defending Israel and 30 defending al-Udeid). For the 2026 war, I estimated conservatively that the US would assume half of the non-US burden from 2025, given Iran’s focus on attacking US military bases, and the relative decline in air defense capacity of Gulf countries compared to Israel.

• Interceptors fired by US forces: 1,348

I assumed a reverse mix of the interceptors the US fired during in June 2025 (with SM-3 split evenly with SM-6/SM-2s) because of the emphasis on US military bases as targets, where Patriot batteries are prevalent. This yielded an estimated 778 Patriot interceptors fired, which I later adjusted slightly to 800 to match the figure given by officials. (Will anyone be impressed by how close I got using my model? Probably not — most people have stopped reading at this point and the people who still are likely aren’t impressed by that sort of thing, presumably having made much more impressive estimates themselves recently. Thanks for reading, by the way.) I offset the resultant increase by subtracting 22 from the SM-3 total, due to the interceptors being in comparatively fewer stockpiled quantities.

• Patriot

  • Quantity: 800

  • Unit cost: $4.6 million

• THAAD

  • Quantity: 156

  • Unit cost: $15.5 million

• SM-3

  • Quantity: 185

  • Unit cost: $28.7 million

• SM-6/SM-2

  • Quantity: 207

  • Unit cost: $5.3 million (SM-6 unit cost, as it replaces the SM-2)

Counter-drone

• Number of drones Iran fired: 2,700

• Share of drones determined to be a threat and engaged by air defenses: 87%

  • It was 73% during the June 2025 war, but I halved the number of non-threatening drones due to the significantly shorter distance from origin to target.

• Percent of drones shot down by the US: 70%

  • The US shot down 18% during the June 2025 war, but I quadrupled the rate for 2026 due to the targeting of US bases.

• Drones shot down by US forces: 1,648

• Interceptors required to shoot down each drone: 2

  • This might be generous, considering it took US fighter aircraft two Sidewinder missiles to shoot down a balloon over Lake Huron in 2023.

• US munitions used on Iranian drones: 3,297

The approximated mix of the 3,297 munitions is based on the mix (also approximated) that US forces used to shoot down Houthi drones. For example, 40% of Houthi drones shot down during Operation Rough Rider were felled using APKWS, so I assumed the same figure here.

• AMRAAM

  • Quantity: 330

  • Unit cost: $1.4 million

• AIM-9X Sidewinder

  • Quantity: 989

  • Unit cost: $485,000

• APKWS

  • Quantity: 1,319

  • Unit cost: $39,494

• Coyote

  • Quantity: 330

  • Unit cost: $126,538

• Other

  • Total cost (est): $1.6 million*

  • *This includes a range of other munitions reportedly used to down Iranian drones, like 30mm rounds fired from Apache helicopters and 20mm fired from C-RAMs. This is a very rough estimate.

Losses

Estimated cost: $2.71 billion

Most of it is due to damaged infrastructure (largely radars) on US bases in the Gulf. The damage initially reported is usually revealed to be much worse once satellite imagery is examined. Unfortunately, satellite imagery can be on tape delay or worse, per ABC: “Since the conflict began, Planet Labs imposed a 14-day delay in releasing images from the region, while Vantor does not share imagery of U.S. military locations.” This is servile, cowardly, and ultimately undemocratic behavior. These companies should reassess their policies, including those that allowed invertebrates to assume leadership positions.

A notable loss of equipment was the three F-15 fighter aircraft shot down over Kuwait on March 1. How much each one costs depends on the media outlet. Good Morning America, for example, says they’re $31 million each. The Pentagon paid $103 million each in 2025.

Additional losses itemized here.

Subsidies

Estimated cost: $1.01 billion

The US-Israeli war with Iran is for the sole strategic benefit of Israel. It is against the strategic interests of the United States. What’s more, the US is covering a significant chunk of Israel’s war costs. That is apparent in pending arms sales that include tens of thousands of bombs, many of which come directly from US stocks (rather than waiting months or years for a contractor to produce them). These are classified as sales, but they aren’t sales in the conventional sense. US military aid functions like a gift card, where Israel is the purchaser but not the funder. Americans pay for these arms deals.

The $1.01 billion estimate reflects the value of bombs and interceptors Israel has expended that can be covered by military aid.

Bombs

Israel reportedly dropped 11,000 bombs during the first two weeks of the war. To estimate the cost, I multiplied the reported quantity by an even mix of the following precision-guided bombs (unit costs reflect 2026 procurement data and include the cost of the warhead and guidance kit, if applicable): 2,000lb GBU-31, $107,266; 1,000lb GBU-32, $78,923; 500lb GBU-38, $74,975; 250lb Small Diameter Bomb, $67,000.

Interceptors

Israel has fired an estimated 820 interceptors from its David’s Sling, Iron Dome, and Arrow 3 platforms through roughly the first two weeks of the war. In 2025, the US provided $40 million, $110 million, and $50 million in missile aid, respectively, for these systems, enough to fund 40 interceptors for David’s Sling, 2,750 for the Iron Dome, and 17 for the Arrow 3. Expect Trump to include billions of dollars for Israeli interceptors in any war funding request.

I may revisit this methodology. For bombs, why didn’t I just put the value of the pending arms deals? For interceptors, why did I only include the $200 million in US missile aid earmarked for procurement and not the other $300 million for R&D, when the latter frees up Israeli funding for missile procurement?

There’s always time for methodological adjustments during endless war.

SPECIAL THANKS TO: Abe B., Alan F., Alexander L., Alissa Q., Amin, Andrew R., AT., B. Kelly, Barbara B., Bart B., BeepBoop, Ben, Ben C.,* Bill S., Bob N., Brett S., Byron D., Carol V., Catherine L., Chris, Chris G., Claudia, Cole H., Coleman J., D. Kepler, Daniel M., Dave, David J., David S.,* David V.,* David M., Dharna N., Elizabeth R., Emily H.,* Errol S., Foundart, Fran Q., Francis M., Frank R., Fred R., Gary W., Gladwyn S., Graham P., Griffin R., Hunter S., IBL, Irene B., Isaac, Isaac L., Jacob, James G., James H., James N., Jamie LR., Jcowens, Jeff, Jennifer, Jennifer J., Jessica S., Jerry S., Joe R., John, John, John A., John K., John M., Jonathan S., Joseph B., Joshua R., Julia G., Julian L., Katrina H., Keith B., Kesh L., Kheng L., Lea S., Leah A., Leila CL., Lenore B., Linda B., Linda H., Lindsay, Lindsay S.,* Lora L., Mapraputa, Marie R., Mark L., Mark G., Marvin B., Mary Z., Marty, Matthew H.,* Megan., Melanie B., Michael S., Mitchell P., Nick B., Noah K., Norbert H., Omar A., Omar D.,* Peter M., Phil, Philip L., Ron C., Rosemary K., Sari G., Scarlet, Scott H., Silversurfer, Soh, Springseep, Stan C., TBE, Teddie G., Theresa A., Themadking, Tim C., Timbuk T., Tony L., Tony T., Tyler M., Victor S., Wayne H., William H.,* William P.

* = founding member

-Stephen (Follow me on Instagram, Twitter, and Bluesky)

Subscribe now

Share

Read more

Share

*Thank you, Alexander L. and Kesh L., for becoming paid subscribers! Please consider joining Alexander, Kesh, and the other VIPs thanked at the bottom of each note to support my work.

*Latest newsletter for VIPs: Israel doesn’t fund the vast majority of its US weapons purchases — US taxpayers do. Here’s precisely how much Americans pay for arms “sales” to Israel. Read more.

*Read my comments on US military aid to Israel in L’Orient–Le Jour.

Subscribe now

Situation

The House ended the partial government shutdown last week by passing legislation containing five full-year spending bills for 2026 — Pentagon ($839 billion); Labor, Health and Human Services, and Education ($195 billion); Transportation and Housing and Urban Development ($103 billion); Financial Services and General Government ($26 billion); State Department ($50 billion) — and a temporary funding extension for the Department of Homeland Security (prorated based on its $89 billion 2025 budget).

The spending package originally contained a full-year Department of Homeland Security (DHS) funding bill, but the Senate replaced it with a two-week funding measure after Customs and Border Protection (CBP) agents murdered legal observer Alex Pretti, RN, which followed the murder of Renee Good by Immigration and Customs Enforcement (ICE) agents.1 The Senate approved the amended legislation on January 30, sending it back to the House.2 Democratic leaders said they would use those two weeks to negotiate ICE reforms as a condition for approving the full-year DHS funding bill.

Assessing Democratic leadership’s proposed ICE reforms

On Wednesday, House and Senate Minority Leaders Hakeem Jeffries (D-NY) and Chuck Schumer (D-NY) proposed ten reforms for ICE, which you can read here. None of them involve reducing ICE funding.

The question isn’t whether these proposed reforms are any good. The question is whether they’re worth $11 billion. Why? Because there’s $11 billion for ICE in the pending 2026 DHS bill, and agreeing to the reforms unlocks the Democratic votes needed to pass it.

If you’re ICE, it’s an easy choice: take the money. First let Jeffries and Schumer concede some of their ten demands, then ignore the rest once the DHS bill passes. You’ll eventually get a strongly worded letter from Jeffries and Schumer for disregarding the agreed-upon reforms, but it’s not like your livelihood (budget) will be in danger. If Democratic leaders won’t try to cut ICE’s funding after it murders someone, they’re not going to block funding over masks and military apparel.

Trump administration goals — including deporting 1 million people per year and producing media content that shows the use of overwhelming force against immigrants — require ICE to function as a paramilitary force and occupy major metropolitan areas. Based on their proposal, Jeffries and Schumer are saying they can stop ICE’s warrantless arrests, arbitrary detention, and summary executions without ending its occupation of US cities, even though those illegal actions are part and parcel of military occupations. Money is policy: you can’t give ICE a budget larger than all but 15 militaries worldwide and expect the agency to demilitarize its behavior. Jeffries and Schumer know this.

The limits of Democratic opposition to state violence

The Jeffries–Schumer proposal speaks to a broader theme: Even when Democrats object to state violence, they’re still willing to fund it. This applies to both paramilitary (ICE) and military spending.

On January 29, Sen. Tim Kaine (D-VA) introduced a war powers resolution prohibiting US forces from attacking Iran without congressional approval. A smart move, considering recent reports warning of an imminent US attack and Trump himself setting the precedent for the US bombing Iran — something of a red line for past administrations — in June. This time around, the goal appears to be regime change.

The next day, Kaine voted to hand Trump a record military budget. The $839 billion Pentagon funding bill, when combined with the military spending in the reconciliation bill and other 2026 funding bills,3 totals $1.02 trillion.

Within 48 hours, Kaine objected to a unilateral war with Iran and approved the funding needed to fight one. He’s willing to take action reasserting Congress’ warmaking powers, but checking executive overreach using Congress’ power of the purse is apparently too much to ask.

Kaine’s position represents the extent to which the Democratic leaders want the party to be anti-war. For party leadership, it’s OK to be pro-war and it’s OK to introduce non-binding resolutions against war, but funding war is non-negotiable. Increasing military spending is acceptable if not encouraged; cutting it is anathema.

To illustrate this point, when the GOP reconciliation bill was being considered in the House last summer, Democratic leadership discouraged its members from proposing amendments striking the bill’s more than $150 billion in military spending. (I heard this directly from Hill staff.) Keep in mind that most amendments Democrats submitted to the Big Beautiful Bill weren’t given a vote; they were symbolic. But House Democratic leaders didn’t even want to give the faintest suggestion that the party opposed a massive increase in military spending.4

Democratic leadership has apparently decided that the party will not constrain war abroad or at home by reducing military or paramilitary spending, regardless of who’s president.

Reining in the US paramilitary budget

An alternative to the Jeffries–Schumer ICE proposal: Refuse to pass a DHS bill with ICE funding.

Some argue that such a demand would be pointless because ICE already has funding from the GOP reconciliation bill ($19 billion annually over the next four years, on average). Republicans, they say, would scoff at Democrats’ demand, let temporary DHS funding lapse, and allow other DHS agencies to shut down indefinitely. Democrats’ hands are tied.

I don’t find this argument convincing. First, I don’t think the GOP would be fine leaving the rest of DHS unfunded. Take FEMA, for example. How long would congressional Republicans be willing to deny a Republican administration the ability to respond to hurricanes, floods, and the like over extra ICE funding? Or take TSA — how much patience does the Trump administration have for an extended airport security staffing shortage? Republicans’ hands are tied.

Second, the fact that ICE already has funding is not an excuse to give them more. The $11 billion in the pending DHS bill would be on top of the projected $19 billion from the reconciliation bill for 2026.5 Giving ICE that extra $11 billion would nearly triple its 2025 budget, which was already a record high.

The table below ranks the pending $30 billion ICE budget among the highest funded militaries. It comes in at 16th, exceeding the military spending of countries like Canada, Spain, Brazil, Mexico, Sweden, Iran, and Switzerland. (Combined funding for ICE and the adjacent CBP stands at $66 billion, which would rank 8th.)

^Alt text for screen readers: ICE funding exceeds the military budgets of all but 15 countries. This table ranks the proposed $30 billion ICE budget for 2026 among the largest military budgets worldwide. It has 3 columns — ranking, military or agency, and funding — and 40 rows, 39 of which are country military budgets and 1 is the ICE budget. ICE is ranked 16th, larger than the military budgets of countries including Canada, Spain, Brazil, Mexico, Sweden, Iran, and Switzerland. The ICE budget refers to the proposed 2026 discretionary spending in H.R.7147 plus average annual reconciliation funding in P.L.119-21. Military budgets are sourced from the Stockholm International Peace Research Institute.

SPECIAL THANKS TO: Abe B., Alan F., Alexander L., Alissa Q., Amin, Andrew R., AT., B. Kelly, Bart B., BeepBoop, Ben, Ben C.,* Bill S., Bob N., Brett S., Byron D., Carol V., Catherine L., Chris, Chris G., Claudia, Cole H., Coleman J., D. Kepler, Daniel M., Dave, David J., David S.,* David V.,* David M., Dharna N., Elizabeth R., Emily H.,* Errol S., Foundart, Francis M., Frank R., Fred R., Gary W., Gladwyn S., Graham P., Griffin R., Hunter S., IBL, Irene B., Isaac, Isaac L., Jacob, James G., James H., James N., Jamie LR., Jcowens, Jeff, Jennifer, Jennifer J., Jessica S., Jerry S., Joe R., John, John, John A., John K., John M., Jonathan S., Joseph B., Joshua R., Julia G., Julian L., Katrina H., Keith B., Kesh L., Kheng L., Lea S., Leah A., Leila CL., Lenore B., Linda B., Linda H., Lindsay, Lindsay S.,* Lora L., Mapraputa, Marie R., Mark L., Mark G., Marvin B., Mary Z., Marty, Matthew H.,* Megan., Melanie B., Michael S., Mitchell P., Nick B., Noah K., Norbert H., Omar A., Omar D.,* Peter M., Phil, Philip L., Ron C., Rosemary K., Sari G., Scarlet, Scott H., Silversurfer, Soh, Springseep, Stan C., TBE, Teddie G., Theresa A., Themadking, Tim C., Timbuk T., Tony L., Tony T., Tyler M., Victor S., Wayne H., William H.,* William P.

* = founding member

-Stephen (Follow me on Instagram, Twitter, and Bluesky)

Subscribe now

Share

Read more

Share

*ICYMI: How much aid has the US given Israel? After uncovering over $7 billion in previously unreported funds, this is the most accurate tally of US aid to Israel to date.

*Big thanks to Claudia for becoming a paid subscriber! Please consider joining Claudia and the other Polygraph VIPs thanked at the bottom of each newsletter.

Subscribe now

House vote

Last night, the House passed a bill that pushes this year’s military budget past the $1 trillion mark. The legislation garnered the support of 149 Democrats. Just 3% of Democratic voters want to increase military spending, but 70% of House Democrats just voted for an increase.

The question is whether Democrats should be let off the hook for that vote because the Pentagon bill was bundled with two other spending bills. One reason they shouldn’t is that three-quarters of the money in the three-bill “minibus” is military spending. There’s $839 billion for the Pentagon, but only $195 billion for the Departments of Labor, Health and Human Services, and Education, and $103 billion for the Departments of Transportation and Housing and Urban Development.

Another reason is that military spending enables the performative shows of force that increasingly define Trump’s foreign policy. The scale of these violent and dangerous performances will grow to fit the expanded military budget the House just gave him (pending Senate approval). The aforementioned $839 billion is on top of the $156 billion in military spending included in the Big Beautiful Bill1 and is in addition to the military spending in other 2026 budget bills, including $34 billion from the Energy Department bill (mostly for nuclear weapons); $20 billion from Military Construction and Veterans Affairs; $7 billion from Commerce, Justice, and Science; $3 billion from Homeland Security; $400 million from Transportation; and $50 million from Financial Services. Of the 12 annual spending bills, 7 include military spending.

Taken together, the pending 2026 military budget totals $1.02 trillion. Note that I only counted the money that fits within the US government’s definition of military spending,2 which excludes certain costs of war (e.g., Veterans Affairs budget) and paramilitary forces (e.g., ICE budget).3

A third reason is that this money is controlled by an executive department that can’t pass an audit. The Pentagon doesn’t properly track its spending or resources. This means that House members overwhelmingly approved a historic transfer of wealth without knowing how taxpayer dollars will actually be used.

Naturally, 89% of House Republicans — who last year slashed social welfare in the name of reducing government waste, fraud, and abuse — voted to increase funding for the single largest driver of that waste, fraud, and abuse.

30 years of failed audits

It got lost in the news cycle, but over the holidays, the Pentagon failed its eighth consecutive audit. Sounds bad. But even this framing shortchanges the department’s breathtaking negligence. The Pentagon has had 30 years to pass an audit, and it still hasn’t passed one.

The Chief Financial Officers Act (CFO) of 1990 required 24 major executive branch departments and agencies to prepare annual financial statements and have them audited. However, the CFO Act required only limited financial statements for certain agency activities. The Government Management Reform Act (GMRA) of 1994 filled this gap, directing those 24 CFO Act agencies to prepare agency-wide audited financial statements for fiscal year 1996 by March 1, 1997.4

The Pentagon submitted agency-wide financial statements for audit from 1996–2001, and failed all of them. Technically, they resulted in “disclaimers of opinion” — when an organization’s books are such a mess that there’s not even enough evidence to form an audit opinion. That’s tantamount to failure, but it’s somehow worse than the lowest grade auditors can give (“adverse opinions”). At least in those cases, there are financial records for auditors to evaluate. The Pentagon couldn’t even produce the materials auditors grade. In 2001, Congress decided to spare the Pentagon further humiliation by passing a law allowing the department to skip the annual agency-wide audit beginning in 2002. This was undone by a 2014 law that led to the resumption of full audits in 2018. The Pentagon has failed all eight audits since then, all receiving disclaimers of opinion. If auditors can’t even gather enough evidence to say what specific problems are going on, it’s likely that actual fraud is far greater than reported fraud. So the recently disclosed $10.8 billion in fraud isn’t just the tip of the iceberg, it’s the tip of that tip.

The underlying theory of the CFO Act and GMRA was that adopting private sector-style financial accounting and reporting standards would make federal agencies more efficient and accountable. Financial audits — third-party investigations of an organization’s books — support government accountability by revealing how agencies spend public money and manage the things they purchase.

Is there another system (a complex, if you will) more in need of greater accountability? The Pentagon owns 82% of the federal government’s total physical assets, eats about 50% of the budget Congress passes each year, and sends 54% of its own budget to for-profit companies. It doesn’t know the number, value, or location of all its matériel scattered across the globe on land, sea, and sky. In 2025, the Pentagon couldn’t account for at least 43% of its assets and 64% of its budgetary resources.5 How can Congress or the New York Times claim the military is under-resourced when the Pentagon itself doesn’t know what it already has?

Congress made the problem worse last night by approving a massive increase in military spending. How? Consider the following. The Pentagon couldn’t account for 47% of its assets in 2022 and 43% in 2025 — a modest improvement. However, the Pentagon is missing more stuff now than it was in 2022. Because it acquired more and more assets during the intervening years, the dollar value of the items it couldn’t account for increased by $342 billion, totaling $2 trillion in 2025. So while the share of the assets it can’t account for fell by 9% (or four percentage points), the dollar value of those missing assets increased by 21%. Congress rewards the Pentagon’s incompetence faster than it makes progress on audits.

The graph below shows how long it’s taken major federal agencies to pass an audit. (If you’re unsure of any of the acronyms, see here.) After the agency-wide audit requirement kicked in for 1996 (2005 for the Department of Homeland Security), some agencies passed the same year, while others took longer. All told, 23 of the 24 CFO Act agencies have received a clean audit opinion.

On average, it takes a federal agency just under two and a half years to pass an audit. The Pentagon hasn’t passed one in thirty.

^Alt text for screen readers: The Pentagon has had 30 years to pass an audit. It still hasn’t passed one. All other major federal agencies have received a clean audit opinion. This chart shows how many years each agency took to pass an audit. All CFO Act agencies were required by law to undergo agency-wide audits beginning in fiscal year 1996, except for the Department of Homeland Security, which was in 2005. Here is each agency and the first year each one passed an audit: Agriculture, 2002; Commerce, 1999; Pentagon, never passed an audit; Education, 1997; Energy, 1996; HHS, 1999; Homeland Security, 2013; HUD, 1998; Interior, 1997; Justice, 2001; Labor, 1997; State, 1997; Transportation, 1999; Treasury, 2000; Veterans Affairs, 1999; EPA, 1997; GSA, 1996; NASA, 1996; NSF, 1998; NRC, 1996; OPM, 2000; SBA, 1996; SSA, 1996; USAID, 2003.

SPECIAL THANKS TO: Abe B., Alan F., Alissa Q., Amin, Andrew R., AT., B. Kelly, Bart B., BeepBoop, Ben, Ben C.,* Bill S., Bob N., Brett S., Byron D., Carol V., Catherine L., Chris, Chris G., Claudia, Cole H., Coleman J., D. Kepler, Daniel M., Dave, David J., David S.,* David V.,* David M., Dharna N., Elizabeth R., Emily H.,* Errol S., Foundart, Francis M., Frank R., Fred R., Gary W., Gladwyn S., Graham P., Griffin R., Hunter S., IBL, Irene B., Isaac, Isaac L., Jacob, James G., James H., James N., Jamie LR., Jcowens, Jeff, Jennifer, Jennifer J., Jessica S., Jerry S., Joe R., John, John, John A., John K., John M., Jonathan S., Joseph B., Joshua R., Julia G., Julian L., Katrina H., Keith B., Kheng L., Lea S., Leah A., Leila CL., Lenore B., Linda B., Linda H., Lindsay, Lindsay S.,* Lora L., Mapraputa, Marie R., Mark L., Mark G., Marvin B., Mary Z., Marty, Matthew H.,* Megan., Melanie B., Michael S., Mitchell P., Nick B., Noah K., Norbert H., Omar A., Omar D.,* Peter M., Phil, Philip L., Ron C., Rosemary K., Sari G., Scarlet, Scott H., Silversurfer, Soh, Springseep, Stan C., TBE, Teddie G., Theresa A., Themadking, Tim C., Timbuk T., Tony L., Tony T., Tyler M., Victor S., Wayne H., William H.,* William P.

* = founding member

-Stephen (Follow me on Instagram, Twitter, and Bluesky)

Subscribe now

Share

Share

*ICYMI: War is Trump’s answer to the affordability crisis

*Thank you, Catherine L., for supporting Polygraph! If you’re enjoying this newsletter, please consider joining Catherine and the other paid subscribers thanked at the bottom of each note.

Subscribe now

Situation

Last week, the House passed legislation funding the State Department for the rest of the year by a 341–79 margin, as part of a two-bill “minibus.” This morning, the Senate released the text of a four-bill spending minibus, including the Pentagon.

Between the pending 2026 budget bills for the State Department and Pentagon, there’s more than $4 billion in aid for Israel. Here’s a breakdown of this $4,017,117,000:

State Department bill

• $3.3 billion in military aid, of which “not less than” $250.3 million is for Israel to spend on arms from Israeli companies. The remaining amount functions as a gift card for Israel to buy US weapons.1

• This amount is in line with the terms of the ten-year Memorandum of Understanding between the US and Israel that has been in effect since 2019. However, contrary to popular belief, the MOU isn’t legally binding. The US gives billions of dollars in military aid to Israel each year because each year, Congress decides to do so.

• $14.6 million in economic aid, including $117,000 for an Israeli scholarship program, $3 million for Israeli development programs, $5 million for “historical, archaeological, and cultural initiatives, including in Jerusalem, that strengthen and deepen the United States-Israel special relationship,” and $6.5 million for resettling refugees in Israel.

• Meanwhile, the bill prohibits funding for UNRWA, originally established by the UN General Assembly to provide relief to the Palestinians expelled by Israel during the Nakba.2 It also prohibits funding for UN investigations into Israeli human rights violations. The bill requires certification that the Palestinian Authority is “acting to counter incitement of violence against Israelis and is supporting activities aimed at promoting peace, coexistence, and security cooperation with Israel” in order to be eligible for humanitarian or development assistance.

Pentagon bill

• $702.5 million in military aid, including $500 million for missile programs, $80 million for anti-tunneling, $75 million for counter-drone, and $47.5 million for “emerging technology.”

• The Pentagon’s Israel aid programs are labeled as “cooperative,” implying an equally shared burden and benefit. In reality, it’s just more military aid. For example, $273 million of the $500 million in missile aid in 2026 is for the Arrow 3 system, even though the US has never purchased the Arrow system and has never intended to.3

How much aid has the US given Israel?

By the time the House passed the 2026 bill with $3.3 billion for Israel, the US had already given Israel $352 billion in aid since its founding. This includes $263 billion in military aid and $89 billion in economic assistance. These inflation-adjusted amounts4 are shown on a cumulative basis in the graph below.

These figures are the products of my own research — research that I wouldn’t have made the effort to do if I thought existing figures fully captured US aid to Israel. I found over $7 billion in aid that was previously unaccounted for, mostly through a FOIA request on an obscure military assistance program. This amounts to a 2% improvement over the existing data — framed this way, my contribution sounds considerably less impressive.

The methodology section (just south of the chart) compares my findings to the three most widely cited sources on the question posed above: USAFacts, Council on Foreign Relations, and the Congressional Research Service. No two sources offer the same findings, largely because foreign aid data has to be pieced together. While the US government maintains a foreign aid database (ForeignAssistance.gov), it doesn’t cover all assistance programs, funding levels for those programs only go back to 2001, and there’s a lag time in reporting. Moreover, due to transparency issues in US foreign assistance — particularly regarding Israel — everyone’s findings are estimates. Some estimates are better than others, though.

*

The chart below starts in 1951, when the US began providing economic aid to Israel. Military assistance started in 1959. Cumulatively, the latter overtook the former in 1974, and the gap has progressively widened, particularly since economic aid fell off in the late 1990s–early 2000s. Israel still receives economic assistance — $24 million per year over the last decade, on average — but the cumulative effect is nearly imperceptible when juxtaposed with the billions of dollars in annual military aid. (For examples of how Israel uses the weapons it acquires with US military aid, see here and here.)

The military aid illustrated in the graph adhere to a bureaucratic definition of the term, meaning that only funding provided to Israel through official military assistance and “cooperative” programs is counted. For example, the $6 billion the US spent directly defending Israeli territory against counterattacks by Iran in April 2024, October 2024, and June 2025 doesn’t count as military assistance, even though the US was literally assisting Israel militarily. The point is that yes, the aid figures are enormous, and no, they don’t fully capture the financial costs of the US’s Israel policy.

^Alt text for screen readers: The U.S has given Israel $352 billion in aid. Cumulative U.S. aid to Israel, by type. Total aid, 1951 to 2025, $352 billion; including $263 billion in military aid and $89 billion in economic aid. Figures in constant 2025 U.S. dollars. Data: ForeignAssistance.gov, Freedom of Information Act request (Data Liberation Project), CRS, DSCA, GAO, author’s analysis of legislation and reporting language.

Methodology

It’s just methodological talk from here on, so feel free to attend to the rest of your emails. Thanks for opening this one.

Foreign aid data isn’t stored in one place. However, the bulk of it can be found at ForeignAssistance.gov, which is jointly managed by the State Department and USAID (or what’s left of it). First, use the live dashboard for funding data going back to 2001. Second, consult the static U.S. Overseas Loans and Grants “Greenbook” for the years before 2001. Third (optional), cross-reference the overlap period (2001–20), and reconcile any discrepancies.

This is more or less the starting point for each of the following four sources on all-time US aid to Israel: USAFacts, Council on Foreign Relations, Congressional Research Service, and Polygraph. What separates them is the level of effort in accounting for the foreign aid not included in State-USAID data.

USAFacts

A hefty portion of military aid to Israel is funded by the Pentagon. Most Pentagon-funded aid is for missile programs. This missile aid is not included in the State-USAID data. Not counting missile aid in a typical year means omitting $500 million in military aid to Israel. In 2024, it meant omitting $5.7 billion.

USAFacts based its graph only on State-USAID data and is therefore missing $18.5 billion5 from the missile aid omission alone.

^USAFacts (accessed 19 Jan 2026).

Council on Foreign Relations

CFR reports $330 billion in total US aid to Israel as of 2024, including $244 billion in military aid and $86 billion in economic aid (in constant 2024 dollars).

This is a considerably better estimate than the one above because CFR bothered to include missile aid. However, in the notes section of their graph, CFR says it started counting missile funding in 2006. Why? Pentagon funding for Israel’s Arrow program dates back to the $52 million included in the 1990 military spending bill. Excluding missile aid from 1990–2005 means CFR’s estimate is $2.3 billion6 lower than it otherwise would be.

CFR has a $100 million annual budget.

^CFR (accessed 19 Jan 2026)

Congressional Research Service

The best estimate so far. Not only does CRS include missile aid data in full, it collected all that data for us in the first place. The report containing its estimate is the single best resource on US aid to Israel; it even cites an article written by your boy.

The table below summarizes its tally. It’s slightly inconvenient that there’s no year-by-year breakdown or an inflation-adjusted total, making it all the more annoying that USAFacts and CFR set out to do those things and failed. Not a big deal.

I have questions, though. For example, why does its military aid count for 2021–25 omit funding for the Pentagon’s tunnel, drone, and emerging technology aid programs? What about the Excess Defense Articles program (discussed below)? More confounding is that CRS mentions all those programs in its report, and even itemizes annual funding levels for some of them.

^CRS (accessed 19 Jan 2026)

Polygraph

My estimate is 2% higher than CRS’s: $5.2 billion more in nominal terms or about $7 billion more in constant 2025 dollars.7 This 2% mostly comes from resolving a discrepancy between how much aid the State Department said went to Israel through a certain military aid program versus how much aid for that program appears in the State Department’s database (ForeignAssistance.gov).

The program in question is the Excess Defense Articles (EDA) program. It’s essentially the international equivalent of the 1033 program, through which the Pentagon sends matériel it deems excess to police for free. EDA does the same thing for foreign countries. CRS notes that Israel has received $6.6 billion worth of matériel through the EDA program. However, less than a sixth of that amount appears in State-USAID data, which means less than a sixth of the funding is represented in the military aid totals of USAFacts, CFR, and CRS.

The problem is that the government says one thing, but its data shows another. I solved this problem using a FOIA request published by the Data Liberation Project, which shows complete EDA data for 1993–2023. It reveals unreported data that account for the discrepancy between the State Department’s official number on the value EDA transfers to Israel and the sum of the figures in its database.8

The discrepancy comes down to how the value of arms transferred through the EDA program are counted. The Pentagon’s own EDA dataset (which only goes back to 2013) lists two values for each item transferred through the EDA program: the original cost and an adjusted cost. The original cost is the item’s acquisition value, i.e., how much the US government originally paid for the thing. The adjusted cost is between 5% and 50% of the original cost. The State Department’s official number refers to original cost, but for whatever reason, only the adjusted costs are tallied on ForeignAssistance.gov.

The adjusted value is only there for the rare instances of an EDA sale, and is irrelevant when the item is given away for free, which for Israel is 99.7% of the time. Only 9 of 3,426 EDA transfers to Israel from 1993–2023 were sales. (I threw out those 9 cases for this study.)

The value of transfers through the 1033 program — the domestic equivalent of the EDA program — are recorded by original acquisition cost. For several reasons, I believe that this is the most accurate way to track EDA transfers too. The administration agrees: a State Department fact sheet from April 2025 states, “Since 1992, the United States has provided Israel with $6.6 billion worth of equipment under the Excess Defense Articles program, including weapons, spare parts, weapons, and simulators.” If the State Department understood the value of EDA transfers in terms of their adjusted costs, it would’ve said $1 billion, not $6.6 billion. Using the FOIA data, I counted $6,562,489,557 for the aggregate original value of EDA transfers to Israel from 1992–2024 (constant 2025 dollars), which rounds to the State Department’s official figure.

The FOIA request provides EDA data for 1993–2023. My graph includes estimated EDA data for 1990–92 and 2024–25. A 1994 GAO report provided EDA data for 1990–92 as a lump sum ($454,583,000). I divided that amount evenly across those three years ($151,527,667), before adjusting for inflation. For 2024, I estimated $83,694,574 based on the $14,555,578 in adjusted value reported for 2024 on ForeignAssistance.gov multiplied by 5.75, the average ratio of original to adjusted value from 1990–2023. For 2025, I estimated $137,343,797, based on the average acquisition value of transfers from 1990–2023.

*NB: A previous version misstated Israel’s GDP per capita compared to US states. That sentence has been removed.

SPECIAL THANKS TO: Abe B., Alan F., Alissa Q., Amin, Andrew R., AT., B. Kelly, Bart B., BeepBoop, Ben, Ben C.,* Bill S., Bob N., Brett S., Byron D., Carol V., Catherine L., Chris, Chris G., Cole H., Coleman J., D. Kepler, Daniel M., Dave, David J., David S.,* David V.,* David M., Dharna N., Elizabeth R., Emily H.,* Errol S., Foundart, Francis M., Frank R., Fred R., Gary W., Gladwyn S., Graham P., Griffin R., Hunter S., IBL, Irene B., Isaac, Isaac L., Jacob, James G., James H., James N., Jamie LR., Jcowens, Jeff, Jennifer, Jennifer J., Jessica S., Jerry S., Joe R., John, John, John A., John K., John M., Jonathan S., Joseph B., Joshua R., Julia G., Julian L., Katrina H., Keith B., Kheng L., Lea S., Leah A., Leila CL., Lenore B., Linda B., Linda H., Lindsay, Lindsay S.,* Lora L., Mapraputa, Marie R., Mark L., Mark G., Marvin B., Mary Z., Marty, Matthew H.,* Megan., Melanie B., Michael S., Mitchell P., Nick B., Noah K., Norbert H., Omar A., Omar D.,* Peter M., Phil, Philip L., Ron C., Rosemary K., Sari G., Scarlet, Scott H., Silversurfer, Soh, Springseep, Stan C., TBE, Teddie G., Theresa A., Themadking, Tim C., Timbuk T., Tony L., Tony T., Tyler M., Victor S., Wayne H., William H.,* William P.

* = founding member

-Stephen (Follow me on Instagram, Twitter, and Bluesky)

Subscribe now

Share

8

If anyone from DLP is reading this: Thank you for your terrific work. Also, there are several instances of the Pentagon recording the original value in the adjusted value column, and vice-versa, evidenced by the instances in which the adjusted value is higher than the original (which should never be the case for EDA transfers). These mistakes are present in the source PDF, i.e., they’re not your fault. Happy to talk more.

*Sen. Elizabeth Warren (D-MA) and Rep. Chris Deluzio (D-PA) cited Polygraph in a recent letter to the Treasury Secretary on military contractor stock buybacks. You can help Polygraph appear in more footnotes by becoming a paid subscriber, like Dharna N. just did (thank you, Dharna!)

Subscribe now

*Latest newsletter for VIPs: Trump’s war for oil has bipartisan backing: Last month’s bipartisan passage of the $901 billion military bill gave Trump tacit approval to attack Venezuela.

Share

Situation

Last weekend, Trump bombed Venezuela, kidnapped its president, and threatened Cuba, Colombia, Greenland, Mexico, Venezuela (really the whole Western Hemisphere), and Iran with military force. This week, Trump said he’s going to propose a $1.5 trillion military budget next year, as Congress prepares spending bills that would give him $1 trillion this year.

The warning signs have always been there with Trump. We’re now seeing the beginning of what those signs were warning against. War is no longer just part of Trump’s political agenda; war is his political agenda. It’s a one-size-fits-all solution to all of Trump’s problems. Or so he hopes.

War is Trump’s answer to the affordability crisis

Foreign policy starts at home. For Trump, home is where his biggest political problem lies. The Epstein investigation looms large, but even that isn’t as massive a political liability as the affordability crisis, a major factor in his historically bad approval ratings. He’s desperate for a distraction.

I’m not about to argue that there’s one issue solely responsible for Trump going all-in on neoconservative aggression. There are many relevant factors, evidenced by Trump transparently fighting a war for oil companies in Venezuela, the performative aspect of his broader foreign policy, and the longstanding status of the United States as an imperial power. (I use “imperial” descriptively here, not pejoratively.1) Rather, my argument is that wars abroad can distract from problems at home, that this function is Trump’s intent, and that it’s a key part of the administration’s strategy. We’re witnessing a violent, cynical act of political self-preservation.

Trump is using war to divert the public’s attention away from his inability to govern. Here are three reasons why.

1. Trump has failed to address voters’ top concern

Affordability — the trendy term for economic security, referring to one’s ability to reliably make ends meet — is deteriorating in the United States, and has been since roughly early 2022.2 It’s been voters’ top concern for about as long. In July 2022, YouGov added “inflation/prices” as an option in its near-weekly survey on what issue Americans say is the most important. This analog for affordability immediately became the most popular selection. Analyzing the polling data, I reported last month that US voters had ranked affordability as their top concern for 41 months straight. It’s now been 42 consecutive months.3

If voters had no longer had a problem affording increasingly high costs, they wouldn’t continue ranking it as their top concern. But since they do, they do. The polling data’s message is that Trump has failed to fix the issue most responsible for his 2024 election victory, and one he campaigned heavily on. Pervasive insecurity at the human level — including economic insecurity and food insecurity, particularly among the working class — was once Joe Biden and Kamala Harris’s greatest political liability. Now Trump owns it.

2. Trump has no plan to fix the affordability crisis

Trump has no credible solution to the affordability crisis, a fact the administration tacitly admits from time to time. For example, the latest annual report on food insecurity from the US Department of Agriculture (USDA) was quietly released last week, covering 2024. (It was supposed to be published in October, but delayed due to the government shutdown and presumably the 18,000 layoffs at the USDA.) The study found that 48 million Americans live in food insecure households, or 14.4% of the population — a 32.1% increase over 2019. Food insecurity is the highest it’s been in over a decade.4

The Trump administration is so confident they’ll be able to fix the rising hunger problem that they not only canceled the next annual food security report, but every one thereafter, claiming that they (the reports) “do nothing more than fear monger.” Unless Congress intervenes, the USDA will not provide a study later this year revealing the effects of further fraying the social safety net during an affordability crisis. Convenient, considering the Big Beautiful Bill’s sizable social welfare cuts, particularly to SNAP (formerly, Food Stamps).

^Alt text for screen readers: Food insecurity has grown by nearly a third since 2019. Americans living in food insecure households, %: 2019, 10.9%; 2020, 11.8%; 2021, 10.4%; 2022, 13.5%; 2023, 14.3%; 2024, 14.4%. Data: United States Department of Agriculture, December 2025.

3. Trump is making the affordability crisis worse

On January 1, enhanced Affordable Care Act (ACA) Marketplace subsidies expired, thereby doubling monthly health insurance premiums for roughly 20 million people. Two days later, Trump launched his attack on Venezuela.

The type, timing, and sequence of these events so perfectly exemplify my argument (that Trump is using war to divert the public’s attention away from his inability to govern) that I feel like they almost undermine it. Last week’s lapse in welfare followed by a dramatic escalation in warfare is a parody-level illustration of my point. After all, I’m not implying direct causality, but in that particular case, the Trump administration seems to be.

However, that exaggeration is closer to the truth than naïvely arguing that the affordability crisis has no bearing whatsoever on Trump’s escalation in Venezuela and seemingly everywhere else. That would be like saying it’s purely a coincidence that the one pandemic-era policy Democrats managed to keep alive all this time (at least until a few days ago) — after promising to make many of them permanent — is the one that sends tens of billions of dollars each year to health insurance companies.5

The expiration of the enhanced ACA Marketplace subsidies marks the official end of the pandemic welfare state, illustrated below. Notice how relief policies overlap in 2021? Now look at 2021 in the food security graph above — welfare expansion in 2021 produced a record drop in hunger, despite the economic recession. (The US also shrunk its military budget and ended a forever war in 2021.6)

^Alt text for screen readers: The rise and fall of the U.S. pandemic welfare state. A sample of pandemic-era relief, 2020 to 2025. This range chart plots the start of pandemic relief policies in green and their end in gray. Foreclosure ban, March 1, 2020 to July 31, 2021; eviction ban, September 1, 2020 to August 26, 2021; additional unemployment coverage, March 27, 2020 to September 4, 2021; extended unemployment benefits, March 27, 2020 to September 4, 2021; child tax credit expansion, March 11, 2021 to December 31, 2021; earned income tax credit expansion, March 11, 2021 to December 31, 2021; universal free school meals, March 18, 2020 to June 30, 2022; extra SNAP benefits, March 18, 2020 to February 28, 2023; uninterrupted Medicaid coverage, March 18, 2020 to March 31, 2023; Women, Infants and Children Nutrition Program increase, June 1, 2021 to September 30, 2023; child care stabilization grants, April 15, 2021 to September 30, 2023; student loan pause, March 13, 2020 to September 30, 2023; emergency rental assistance, December 27, 2020 to September 30, 2025; Affordable Care Act Marketplace enhanced subsidies, March 11, 2021 to December 31, 2025. Source: Author analysis (Stephen Semler).

SPECIAL THANKS TO: Abe B., Alan F., Alissa Q., Amin, Andrew R., AT., B. Kelly, Bart B., BeepBoop, Ben, Ben C.,* Bill S., Bob N., Brett S., Byron D., Carol V., Chris, Chris G., Cole H., Coleman J., D. Kepler, Daniel M., Dave, David J., David S.,* David V.,* David M., Dharna N., Elizabeth R., Emily H.,* Errol S., Foundart, Francis M., Frank R., Fred R., Gary W., Gladwyn S., Graham P., Griffin R., Hunter S., IBL, Irene B., Isaac, Isaac L., Jacob, James G., James H., James N., Jamie LR., Jcowens, Jeff, Jennifer, Jennifer J., Jessica S., Jerry S., Joe R., John, John, John A., John K., John M., Jonathan S., Joseph B., Joshua R., Julia G., Julian L., Katrina H., Keith B., Kheng L., Lea S., Leah A., Leila CL., Lenore B., Linda B., Linda H., Lindsay, Lindsay S.,* Lora L., Mapraputa, Marie R., Mark L., Mark G., Marvin B., Mary Z., Marty, Matthew H.,* Megan., Melanie B., Michael S., Mitchell P., Nick B., Noah K., Norbert H., Omar A., Omar D.,* Peter M., Phil, Philip L., Ron C., Rosemary K., Sari G., Scarlet, Scott H., Silversurfer, Soh, Springseep, Stan C., TBE, Teddie G., Theresa A., Themadking, Tim C., Timbuk T., Tony L., Tony T., Tyler M., Victor S., Wayne H., William H.,* William P.

* = founding member

-Stephen (Follow me on Instagram, Twitter, and Bluesky)

Subscribe now

Share

Subscribe now

*This note is for Polygraph VIPs only. Read on if you are one, or support my work to become one.

Read more

*See how your senators voted on the $901 billion military bill and how much they accepted from military contractors beforehand in this newsletter for Polygraph VIPs. (House version here.)

*Why I promote paid newsletters here instead of sending them to everybody directly: Paywalled emails bother me and I can’t bring myself to send them. (NB: This is admittedly a moronic business decision on my part.)

Share

Subscribe now

Situation

On Tuesday, Trump announced that he had ordered a naval blockade on Venezuela, an act of war, that will continue until the country gives the US “all of the oil, land, and other assets that they previously stole from us.”1 Trump is presumably alluding to the nationalization of Venezuela’s oil industry in 1976 or its renegotiation of foreign oil contracts in 2007. If so, he’s framing a loss of oil company revenue as theft and using oil companies and “us” as synonyms.

The US orchestrated regime change in Iran in 1953 in response to its oil nationalization, and toppled the government of Iraq as part of an oil-related war in 2003.2 There are no good outcomes to US regime change in Venezuela.

Venezuela has the world’s largest proven oil reserves, and Trump’s announcement made clear that the war is squarely about that. Drugs, formerly the pretext for war, were relegated to a secondary issue, a mere byproduct of the country’s oil. What had been the glaring subtext for Trump’s escalation (oil) is now the stated justification. I’m grateful for the transparency.

As of yesterday, net support for an invasion of Venezuela among US voters stands at –36% (22% support to 58% oppose, including “strongly support” at 10% and “strongly oppose” at 44%). But Trump doesn’t need public opinion on his side to wage a major war. What he really needs is reassurance that Congress is willing to fund one.

Senate approves $901 billion military bill

On Wednesday, the Senate passed the 2026 National Defense Authorization Act (NDAA), authorizing $901 billion in military spending, after the House approved it a week earlier. Because the Trump administration plans to use $119 billion of the Big Beautiful Bill’s extra $156 billion military spending next year, the NDAA’s passage signals congressional support for a $1.020 trillion military budget in 2026.3 Adjusted for inflation, the only years with larger military budgets were 2010 — when nearly 200,000 US troops, excluding contractors, were in Iraq and Afghanistan — and 1944 and 1945, during World War II.

The bill passed by a 77–20 margin. Only 10% of voters want a larger military budget, but nearly 80% of the Senate just voted for one. That’s even more out of step with public opinion than House votes on the same bill. If there were a contest over which chamber can distance itself further from voter attitudes on military spending, the Senate would be winning, albeit not by much.

As I did with the 424 House votes, I looked at the money behind the 97 Senate votes, comparing how senators voted with how much each accepted from arms industry donors last election cycle. Paid subscribers already saw the full data in an interactive table (senators are ranked by political contributions received by default, but if you sort that column from lowest to highest, you’ll suddenly see a lot more no/nay votes at the top of the table.)

The chart below compares yes/yea votes and no/nay votes on the NDAA, showing the median amount each group received from military contractors before voting on the $901 billion military bill. I’m now about to state the obvious, but in a very specific way. Money in politics matters: senators who voted to increase military spending received 3.5 times as much arms industry cash as senators opposed.

This money doesn’t predict votes, but it makes aligning congressional votes with public opinion much more difficult.

In many cases, the massive cash incentive effectively rules out the possibility of simply talking a lawmaker into supporting a lower military budget using facts and reason alone. It quite literally makes them unreasonable. Public opinion isn’t enough either — public pressure is needed, and at an amount lawmakers could reasonably describe as fearful. Enough to make the money from military contractors seem trivial.4

^Senators who voted to increase military spending received 3.5 times the arms industry cash of those opposed. Median amount received (2024 election cycle), by vote. Voted YES on $901 billion military bill: $224,028. Voted NO on $901 billion military bill: $63,956. Votes: Fiscal Year 2026 National Defense Authorization Act (NDAA). Political contributions (2019 to 2024): OpenSecrets.

SPECIAL THANKS TO: Abe B., Alan F., Alissa Q., Amin, Andrew R., AT., B. Kelly, Bart B., BeepBoop, Ben, Ben C.,* Bill S., Bob N., Brett S., Byron D., Carol V., Chris, Chris G., Cole H., Coleman J., D. Kepler, Daniel M., Dave, David J., David S.,* David V.,* David M., Elizabeth R., Emily H.,* Errol S., Foundart, Francis M., Frank R., Fred R., Gary W., Gladwyn S., Graham P., Griffin R., Hunter S., IBL, Irene B., Isaac, Isaac L., Jacob, James G., James H., James N., Jamie LR., Jcowens, Jeff, Jennifer, Jennifer J., Jessica S., Jerry S., Joe R., John, John, John A., John K., John M., Jonathan S., Joseph B., Joshua R., Julia G., Julian L., Katrina H., Keith B., Kheng L., Lea S., Leah A., Leila CL., Lenore B., Linda B., Linda H., Lindsay, Lindsay S.,* Lora L., Mapraputa, Marie R., Mark L., Mark G., Marvin B., Mary Z., Marty, Matthew H.,* Megan., Melanie B., Michael S., Mitchell P., Nick B., Noah K., Norbert H., Omar A., Omar D.,* Peter M., Phil, Philip L., Ron C., Rosemary K., Sari G., Scarlet, Scott H., Silversurfer, Soh, Springseep, Stan C., TBE, Teddie G., Theresa A., Themadking, Tim C., Timbuk T., Tony L., Tony T., Tyler M., Victor S., Wayne H., William H.,* William P.

* = founding member

-Stephen (Follow me on Instagram, Twitter, and Bluesky)

Subscribe now

Share

To see the money behind the House votes on the same bill, see this newsletter for Polygraph VIPs.

*This note is for Polygraph VIPs only. Read on if you are one, or support my work to become one.

Read more

Share

*See how your representative voted on the $901 billion military policy bill and how much they received beforehand from military contractors in the latest newsletter for Polygraph VIPs.

*Thank you, Fred R., for becoming a Polygraph VIP! Please consider joining Fred and the other paid subscribers thanked at the bottom of each newsletter to support my work.

Subscribe now

Situation

The House passed the 2026 National Defense Authorization Act (NDAA) on Wednesday, which authorizes $901 billion in military spending — $8 billion more than Trump asked for. An additional $156 billion was included in the Big Beautiful Bill, so the NDAA’s passage signals congressional support not only for its many ugly provisions, but also for a $1 trillion-plus military budget next year.

Historic increases in military spending like the one authorized by the NDAA enable Trump’s foreign policy and practically define his domestic agenda. Money is policy. If you oppose war with Venezuela, occupations of US cities, or a historic transfer of public wealth to private companies partly paid for by cuts to social programs, you vote against the NDAA. If you’re fine with or get excited about those things, you vote for the bill.

I recommended that members of Congress oppose the bill. Only a quarter of them did: the bill passed by a 312–112 margin.1

1. Increasing military spending: Public opinion vs. House votes

If congressional votes reflected public opinion, the bill wouldn’t have passed. Only 1 in 10 voters want a bigger military budget, but more than 7 in 10 House members voted for one on Wednesday.

The discrepancy is illustrated in the following graph. Each pair of bars compares public opinion on military spending and House votes on the same issue. The percentages are the share of voters who support increasing military spending and the share of House members who voted to authorize one through the NDAA.

Feel free to comment below on which party you find more annoying in this context. On one hand, Republicans overwhelmingly voted for the NDAA, and are further out of step with the views of their base than their House counterparts. On the other hand, the bill wouldn’t have passed without the support of Democrats, who just handed a trillion-dollar military budget to a guy they’ve called authoritarian for the past decade.

^Alt text for screen readers: Only 1 in 10 voters want a bigger military budget, but 7 in 10 House members just voted for one. House votes on the $901 billion NDAA clash with public opinion. Support for increasing military spending among Democratic voters: 3%. House Democrats: 55%. Republican voters: 16%. House Republicans: 92%. Voters (all): 10%. House (all): 74%. Polling on public support for increasing military spending: Institute for Global Affairs at Eurasia Group (Nov 2025). Votes: 2026 National Defense Authorization Act (Dec 2025).

2. Cash4Votes

I looked at the money behind the 424 House votes on the NDAA. Specifically, I compared how each House member voted with the amount they accepted from arms industry donors last election cycle (2023–24).2 I sent the data to paid subscribers in an interactive table, where they can sort by name, party, vote, and arms industry cash received. (Recommendation: in the political donations column, go back and forth between sorting amounts from lowest to highest and highest to lowest while keeping your eye on the vote column. It’s wild.)

On average, members who voted to authorize $901 billion in military spending received four times as much money from military contractors as those opposed. (This four-to-one ratio between yes and no votes holds whether you take the mean or median for each group. The chart below refers to the mean.)

I can’t imagine you’re surprised by the results. I’m not surprised, either: I’ve run this analysis since 2018 and found the same correlation — politicians voting to increase military spending taking more money from military contractors — every single year.

Considering the share of military contractors’ revenue from the US government, much of the money they give members of Congress originally came from taxpayers. Arms companies are recycling funds back to the people who signed off on that funding in the first place, helping ensure that Congress doesn’t represent public opinion on how federal dollars should be allocated. Military spending doesn’t just buy weapons, it buys a way of life.

^Alt text for screen readers: Lawmakers who voted to increase military spending received 4 times more money from military contractors. Average amount received (2024 election cycle), by vote. Voted YES on $901 billion military bill: $57,000. Voted NO on $901 billion military bill: $14,000. Votes: 2026 National Defense Authorization Act. Political donations (2023–24): OpenSecrets.

SPECIAL THANKS TO: Abe B., Alan F., Alissa Q., Amin, Andrew R., AT., B. Kelly, BartB., BeepBoop, Ben, Ben C.,* Bill S., Bob N., Brett S., Byron D., Carol V., Chris, Chris G., Cole H., Coleman J., D. Kepler, Daniel M., Dave, David J., David S.,* David V.,* David M., Elizabeth R., Emily H.,* Errol S., Foundart, Francis M., Frank R., Fred R., Gary W., Gladwyn S., Graham P., Griffin R., Hunter S., IBL, Irene B., Isaac, Isaac L., Jacob, James G., James H., James N., Jamie LR., Jcowens, Jeff, Jennifer, Jennifer J., Jessica S., Jerry S., Joe R., John, John, John A., John K., John M., Jonathan S., Joseph B., Joshua R., Julia G., Julian L., Katrina H., Keith B., Kheng L., Lea S., Leah A., Leila CL., Lenore B., Linda B., Linda H., Lindsay, Lindsay S.,* Lora L., Mapraputa, Marie R., Mark L., Mark G., Marvin B., Mary Z., Marty, Matthew H.,* Megan., Melanie B., Michael S., Mitchell P., Nick B., Noah K., Norbert H., Omar A., Omar D.,* Peter M., Phil, Philip L., Ron C., Rosemary K., Sari G., Scarlet, Scott H., Silversurfer, Soh, Springseep, Stan C., TBE, Teddie G., Theresa A., Themadking, Tim C., Timbuk T., Tony L., Tony T., Tyler M., Victor S., Wayne H., William H.,* William P.

* = founding member

-Stephen (Follow me on Instagram, Twitter, and Bluesky)

Subscribe now

Share

Share

*This note is for Polygraph VIPs only. Read on if you are one, or support my work to become one.

Read more

Share

*Thank you, William H. and Coleman J., for becoming paid subscribers! Please consider joining them to support my work.

Subscribe now

UPDATE (10 Dec 17:46): As Trump starts a war with Venezuela, 312 members of the House — including 115 Democrats — just voted to authorize over $900 billion in military spending.

UPDATE (10 DEC 18:09): Roll call, here.

Summary of the annual military policy bill

The House is scheduled to vote today on a bill that authorizes over $900 billion in military spending. (I’ll update this page accordingly, so be sure to check back here.)

The text of the 2026 National Defense Authorization Act (NDAA) — the annual military policy bill — was released Sunday night, and there’s a lot to complain about in its more than 3,000 pages.

For example, Sec. 1706 requires an assessment of the countries restricting arms transfers to Israel over its human rights abuses, which weapons are being withheld, and how the US can “mitigate such gaps.” In other words, the bill says that the US government should function as Israel’s tool for bypassing international arms embargoes. The House Armed Services Committee said this provision “combats antisemitism.” I wish I were kidding.

You won’t find many policies that genuinely align with working-class interests. Sec. 8369, which repeals the devastating Caesar sanctions that indiscriminately punished ordinary Syrians for the last several years, is one of the few bright spots.

Members of Congress will tout the bill’s 3.8% troop pay increase, but not mention the ways the legislation sets up troops to be used as bait or cannon fodder. This includes the high-risk, little-to-no-reward provisions under Sec. 1254 geared toward antagonizing China, including one that instructs the US to “expand the scope and scale of multilateral military exercises,” in the Indo-Pacific, “including more frequent combined maritime operations through the Taiwan Strait and in the South China Sea.”

Corporate interests are represented well. For example, Sec. 803 establishes a pilot program that expands what expenses military contractors can legally bill the government for, including interest payments. Basically, the Pentagon is experimenting with covering contractors’ debt (and other previously unallowable costs).

Corporate interests are also favored based on what’s not in the bill text. Reportedly due to a last-minute arms industry lobbying blitz, “right-to-repair” provisions were stripped from the final version of the NDAA. Right now, military personnel often aren’t legally allowed to fix their own equipment — they’ve got to go back to the contractor for repairs and upkeep. This is a big reason why the operations and maintenance portion of the military budget is larger than all the others, including the accounts for personnel and procurement.1

One reason the lobbying effort to strip the right-to-repair provisions was successful was that the members of Congress who negotiate the final version of the NDAA are often the top recipients of arms industry campaign contributions. They also receive a lot of money from the arms industry lobbyists themselves.2 Among other privileges, this cash makes it a hell of a lot easier to get meetings with members’ offices. I can personally attest to this fact.

Context: a $1 trillion military budget

Today’s vote is effectively a yea-or-nay on a $1 trillion military budget: the $901 billion in the pending National Defense Authorization Act (NDAA) is on top of the extra $156 billion in military spending included in the One Big Beautiful Bill Act, passed in July.3 The White House said it will use $119 billion of that in 2026, saving the rest for future years. This means that the members of Congress who vote yea/yes on today’s NDAA are signaling their support for a $1.020 trillion military budget, $8 billion more than Trump asked for.

NB: $1.020 trillion is the amount that’s able to be endorsed or rejected today. This amount does not include military-related spending outside the discretionary military budget, including pensions for retired military personnel, military aid funded by the State Department, veterans’ health care and other benefits, DHS, ICE, and other “security spending.”

Should Congress pass the 2026 military policy bill?

No.

The Trump administration is leading us into an Iraq-style war in Venezuela, is using US military forces to occupy US cities, and has pledged to resume nuclear weapons testing. What message would voting for the $901 billion NDAA send to the administration? What policies would a $1.020 trillion military budget enable? Money is policy.

The Pentagon still can’t pass an audit. A trillion-dollar military budget would trigger a historic redistribution of wealth that would primarily benefit the 1%. Social programs are being slashed to fund historic increases in Pentagon spending — the One Big Beautiful Bill Act managed to do both at the same time — per the recommendation of the Commission on the National Defense Strategy: “The ballooning U.S. deficit also poses national security risks. Therefore, increased security spending should be accompanied by additional taxes and reforms to entitlement spending.”

Mitch McConnell (R-KY) recently argued in the Wall Street Journal that “spending more on defense is simply necessary.” (WSJ didn’t disclose the fact that McConnell has accepted $1.2 million in political donations from the arms industry, so I thought I’d mention it here.) McConnell’s wrong — it’s not necessary. People who treat more military spending as the answer to all questions think in terms of theatrics, not outcomes, which leaves no room to reflect on whether their preferred policy is part of the problem.

How much is enough, anyway? Seven out of every ten dollars the world spends on militaries are spent by the United States and its main allies. Military spending worldwide reached $2.7 trillion last year, according to the Stockholm International Peace Research Institute. Of that total:

• 37% was by the United States.4

• 55% was by the US and the 31 other NATO countries.

• 70% was by the US, the 31 other NATO countries, and the US’s 21 major non-NATO allies — an official designation under US law that does not include security commitments, but indicates a close relationship with the US and entails special privileges, particularly with respect to arms transfers.5

^Alt text for screen readers: US, major US allies spend 70% of global military budget. This stacked column chart displays 2.7 trillion in military spending worldwide, including $997 billion by the United States, $509 billion by US NATO allies, $383 billion by US major non-NATO allies, and $829 billion by the rest of the world. Figures are for 2024. US total includes related funds outside the discretionary military budget, such as pensions and State Department’s military aid. Data: Stockholm International Peace Research Institute.

SPECIAL THANKS TO: Abe B., Alan F., Alissa Q., Amin, Andrew R., AT., B. Kelly, BartB., BeepBoop, Ben, Ben C.,* Bill S., Bob N., Brett S., Byron D., Carol V., Chris, Chris G., Cole H., Coleman J., D. Kepler, Daniel M., Dave, David J., David S.,* David V.,* David M., Elizabeth R., Emily H.,* Errol S., Foundart, Francis M., Frank R., Gary W., Gladwyn S., Graham P., Griffin R., Hunter S., IBL, Irene B., Isaac, Isaac L., Jacob, James G., James H., James N., Jamie LR., Jcowens, Jeff, Jennifer, Jennifer J., Jessica S., Jerry S., Joe R., John, John, John A., John K., John M., Jonathan S., Joseph B., Joshua R., Julia G., Julian L., Katrina H., Keith B., Kheng L., Lea S., Leah A., Leila CL., Lenore B., Linda B., Linda H., Lindsay, Lindsay S.,* Lora L., Mapraputa, Marie R., Mark L., Mark G., Marvin B., Mary Z., Marty, Matthew H.,* Megan., Melanie B., Michael S., Mitchell P., Nick B., Noah K., Norbert H., Omar A., Omar D.,* Peter M., Phil, Philip L., Ron C., Rosemary K., Sari G., Scarlet, Scott H., Silversurfer, Soh, Springseep, Stan C., TBE, Teddie G., Theresa A., Themadking, Tim C., Timbuk T., Tony L., Tony T., Tyler M., Victor S., Wayne H., William H.,* William P.

* = founding member

-Stephen (Follow me on Instagram, Twitter, and Bluesky)

Subscribe now

Share

Share

*ICYMI: How Trump’s trillion-dollar war machine enriches the 1%

Please read and share with your representative and senators.

*Latest post for Polygraph VIPs: Trump: No economic hardship, just bad vibes

*Many thanks to Emily H. for becoming a paid subscriber — Polygraph continues to attract the support of the very finest folks. Grateful for you all.

Subscribe now

Situation

In November, voters ranked economic insecurity as their top concern — just as they did in each of the 40 months before that.

YouGov runs a recurring survey asking Americans what they believe is the most important issue facing the country.1 “Inflation/prices” was first listed as an option in July 2022, and it’s been the most popular selection in 168 of the 171 near-weekly polls conducted since. (It was the second-most popular response the three times it wasn’t the most popular.2) Average the near-weekly data out into months and affordability has been voters’ number one issue for 41 consecutive months.3

Economic security refers to one’s ability to reliably make ends meet. Being secure in this way means having the resources to make ends meet now and feeling comfortable that you’ll have the resources to do so in the future. This forward-looking dimension is part of any definition of security because the concept itself is future-oriented: it implies continuous protection against a given threat, a preservation of well-being, both now and extending into the future.4 “Inflation/prices” expresses economic security as a poll question by capturing the relevant concerns related to both the present (high prices) and future (rising prices).

“Affordability” has emerged as the trendy synonym for economic security, though the latter’s still quite common. It’s used in describing the problem (“lack of affordability”), the scale of the problem (“affordability crisis”), and policy platforms centered on fixing the problem (“affordability agenda”). The term’s everywhere now.

And it’s about time. The chart below lists the six most popular responses to the YouGov poll from January 2021 to November 2025, and shows the percentage of Americans who ranked each as the most important issue. Five are very salient issues. Then there’s economic insecurity — affordability — which is the issue. Since mid-2022, more than one in five Americans have said it’s their top concern.5

To give you an idea of how salient the next five most popular concerns are, notice how prominent issues like abortion, crime, gun control, foreign policy, and taxes aren’t there instead. This makes it all the more remarkable that affordability has topped all of them for 41 straight months, and that it’s taken the media and Washington almost as long to decide that the affordability crisis is headline-worthy.

What led to affordability’s political breakthrough?

^Alt text for screen readers: Affordability has topped voter concerns since mid-2022. Issues voters ranked as most important, 2021 to 2025. “Inflation/prices” has remained the most popular response every month since July 2022. Figures are monthly averages. Data: YouGov. “Inflation/prices” became an option in July 2022. Issues shown are the six most popular selections on average from January 2021 to November 2025. Other issues listed: “jobs, economy” “civil rights/liberties” “health care” “immigration” “climate, environment.”

*NB: For a more nuanced look at the data, I added an interactive version of the graph at the bottom of this newsletter. Or you can access YouGov’s survey data directly to see the data on all polled issues.

What led to affordability’s political breakthrough

Longtime Polygraph readers know that I haven’t shut up about the economic (and food) security crisis since mid-2023, so it’s surreal seeing “affordability crisis” suddenly all over the headlines in late 2025.

I attribute affordability’s breakthrough in Washington to two things: 1) voters punishing out-of-touch parties and politicians; and 2) voters supporting candidates who challenge the political establishment.

1. Voters punishing out-of-touch parties, politicians

If voters had no problem affording increasingly high costs, they wouldn’t rank it as their top concern. But since they do, they do. High prices harm people’s current well-being, and rising prices threaten their future well-being. There is plenty of data on economic insecurity to back this up.

For example, only about a third of Americans live comfortably financially, according to the Federal Reserve, and more than four in ten have difficulty paying their bills, according to the Consumer Financial Protection Bureau. Data from the most recent five-year period from each of these surveys are shown in the graph below.6

^Alt text for screen readers: More Americans struggle to pay bills than live comfortably. The last time more people lived comfortably financially was 2021. This chart shows the percentage of Americans having difficulty paying the bills from 2020 to 2024 (35, 34, 36, 38, 43) and living comfortably (35, 39, 34, 33, 34).

Despite the abundance of economic insecurity data signaling a crisis, there is a popular — though not for much longer — narrative that insists there’s no substance behind widespread economic negativity, it’s just bad “vibes.” And because it’s bad vibes and nothing more, no policy changes are needed, just better marketing.

Empirically, there’s so much wrong with the “vibecession” narrative that even articles written by its top preachers can be swiftly rebuked in a footnote. It’s also as condescending as you’d expect. Here’s self-described vibes guy Paul Krugman writing in the New York Times in late 2023:

Biden administration officials are trying hard to sell their economic accomplishments, as they should — if they don’t, who will? But will public opinion turn around? Nobody knows. We’re living in a world in which what people believe may have little to do with facts, including the facts of their own lives.

Biden, and later Harris, chose the vibes-guy approach. Here’s something I wrote about Democrats in August 2023:

They’ve certainly got the right to brag about high GDP, low unemployment, and generally-improving inflation numbers. But these numbers don’t reflect the conditions of day-to-day economic life. And those conditions are demonstrably poor. While Republicans don’t have a credible plan to fix things like food or financial insecurity, they’ve still managed to outdo Democrats, who seem to have unanimously agreed to not mention this problem at all. By insisting everything’s fine when it isn’t, Democrats risk coming off as the most annoying people on the planet to millions of voters who wouldn’t otherwise think so.

In November 2024, the electorate delivered a middle finger to an irritatingly out-of-touch incumbent party: While the share of US voters unhappy with economic conditions increased by 18 percentage points from 2020 to 2024, the percentage of those who backed the Democratic candidate fell by 52 points.7

Being a vibes guy ultimately amounts to telling voters they’re fine when they’re not. People hate that, and are generally eager to punish you for it. There might not be a more politically poisonous narrative than the “vibecession”: Democrats embraced it in 2024 and lost. Republicans embraced it in 2025 and lost, and will likely lose the 2026 midterms for the same reason.

Donald Trump is a vibes guy. In a recent Fox News interview, Trump rejected widespread reports of economic insecurity. Here are a couple excerpts:

Ingraham: “Is this a voter perception issue of the economy, or is there more that needs to be done by Republicans on Capitol Hill in terms of policy?”

Trump: “More than anything else, it’s a con job by the Democrats. They just…put out something, say today, costs are up. They feed it to the anchors of ABC, CBS, and NBC…Costs are way down.”

[...]

Ingraham: “So are you saying that voters are misperceiving how they feel? Because you said Biden did that…[Biden] was saying things were great…Why are people saying they’re anxious about the economy?”

Trump: “I don’t know that they are saying that. I think polls are fake. We have the greatest economy we’ve ever had.”

This was a really bad look, particularly because Trump won a lot of votes by campaigning on lowering costs and constantly slamming the Biden-Harris administration for price hikes. It was little surprise that CNN reported the following two days after the Fox News interview:

Trump’s advisers acknowledge that they have an affordability problem…Americans’ outlook on the economy is dour, and the administration’s efforts to ease their financial anxieties aren’t resonating.

“You can’t convince people that their experience, what they’re feeling at home, isn’t reality,” one of the officials said.

Trump reportedly called the affordability crisis a “fake narrative” and a “con job” by Democrats again this week. Too late — Republicans have already acknowledged the affordability crisis and publicly committed to addressing it. Last week, for example, a GOP senator published “A GOP Playbook for an Affordability Offensive” in the Wall Street Journal.

After taking turns denying its existence, Democrats and Republicans now recognize the existence of the affordability crisis at the same time. I think this marks the beginning of the end of the indiscriminately toxic “vibecession” narrative. The smart vibes guys will hedge their past takes (Krugman already has) or move on to something else. The dumb vibes guys won’t see the writing on the wall — they’ve got that special kind of antisocial arrogance, the type that leads one to conclude, All these women at this Whole Foods must be too shy to ask for my phone number, so I guess I’ll just start yelling it.

You won’t see me doing that.

2. Voters supporting challengers to the political establishment

Up until a few months ago, it looked like the Republican and Democratic parties were fine playing political table tennis with voters over affordability — capitalizing off each other’s negligence each election but not proposing to fix the problem once elected — and doing so indefinitely. Harris ignored voters’ number one issue in the 2024 election, Trump ignored it after he won, and establishment Democrats didn’t appear interested in addressing it in 2025 until Zohran Mamdani felled one of their own in the New York City mayoral elections (first in the primary then general election) by running on an affordability agenda.

Affordability is not something establishment politicians want to address because it risks shifting economic politics primarily from the national to human level. Such a shift would risk changing who politicians primarily make policy for: after all, legislating mainly for the sake of national economic growth isn’t the same as doing so mainly for human economic well-being. That would, in turn, risk changing who politicians are primarily accountable to. Instead of answering first to the small group of political and economic elites who define what’s in “the national interest,” more mind would have to be paid to the workers who actually make the economy go.

All this would require actually governing, which probably doesn’t sound so appealing for politicians who’ve made a career off ruling. This helps explain why an immense amount of campaign cash was invested by establishment Republicans and Democrats alike to stop Mamdani, and why the highest-ranking Democrat in office, Chuck Schumer (D-NY), refused to endorse him.

Unfortunately for the political establishment, people support ideas like a public option for groceries and voted accordingly. An affordability agenda speaks to their sense and lived reality of economic insecurity, which Republican and Democratic leadership have taken turns ignoring for the past 41 months.

SPECIAL THANKS TO: Abe B., Alan F., Alissa Q., Amin, Andrew R., AT., B. Kelly, BartB., BeepBoop, Ben, Ben C.,* Bill S., Bob N., Brett S., Byron D., Carol V., Chris, Chris G., Cole H., Coleman J., D. Kepler, Daniel M., Dave, David J., David S.,* David V.,* David M., Elizabeth R., Emily H.,* Errol S., Foundart, Francis M., Frank R., Gary W., Gladwyn S., Graham P., Griffin R., Hunter S., IBL, Irene B., Isaac, Isaac L., Jacob, James G., James H., James N., Jamie LR., Jcowens, Jeff, Jennifer, Jennifer J., Jessica S., Jerry S., Joe R., John, John, John A., John K., John M., Jonathan S., Joseph B., Joshua R., Julia G., Julian L., Katrina H., Keith B., Kheng L., Lea S., Leah A., Leila CL., Lenore B., Linda B., Linda H., Lindsay, Lindsay S.,* Lora L., Mapraputa, Marie R., Mark L., Mark G., Marvin B., Mary Z., Marty, Matthew H.,* Megan., Melanie B., Michael S., Mitchell P., Nick B., Noah K., Norbert H., Omar A., Omar D.,* Peter M., Phil, Philip L., Ron C., Rosemary K., Sari G., Scarlet, Scott H., Silversurfer, Soh, Springseep, Stan C., TBE, Teddie G., Theresa A., Themadking, Tim C., Timbuk T., Tony L., Tony T., Tyler M., Victor S., Wayne H., William P.

* = founding member

-Stephen (Follow me on Instagram, Twitter, and Bluesky)

Subscribe now

Share

Share

*My latest in Popular Information: How Trump’s trillion-dollar war machine enriches the 1% Please read and share with your representative and senators.

*Thank you, Alissa Q., for becoming Polygraph’s latest VIP member! Please consider joining Alissa and the others thanked at the bottom of each note to support my work:

Subscribe now

Situation

The issue at the heart of the recent government shutdown was health insurance subsidies, specifically, enhanced subsidies for the Affordable Care Act (ACA) Marketplace. They originated in the 2021 American Rescue Plan, were extended by the 2022 Inflation Reduction Act, and will expire at the end of this year if they aren’t extended again.

Democrats had refused to vote on any budget deal to reopen the government without funding for enhanced subsidies until eight Democratic senators (Cortez Mastro, Durbin, Fetterman, Hassan, Kaine, King, Rosen, Shaheen) joined Republicans to advance a funding bill that excluded them, clearing the way for the bill’s enactment. The eight Democrats say they secured from Republicans a future vote on Marketplace subsidies — that is, they made a pinky promise with the party that nearly repealed the entire ACA a few years ago. They’re also touting “wins” such as backpay for federal employees and resumed SNAP funding, but those are issues related to the shutdown itself, neither of which was a partisan sticking point.

Democrats could have tied the shutdown to any number of demands, but decided on Marketplace subsidies, launching a whole ad campaign around it. That Democrats came away from the highly publicized and self-promoted standoff with nothing has prompted some lawmakers and several major progressive organizations to call on Chuck Schumer (D-NY), the historically unpopular Senate Majority Leader and currently the highest-ranking Democrat in office, to resign.1

What happens if the Marketplace subsidies expire? Enrollment in the Marketplace is 24.3 million. Most, if not all, will face higher monthly health insurance premiums in January, for two reasons.

First, Marketplace insurance companies have decided to increase their premiums by 26% next year.2 This is in keeping with the trend of private health insurance costs growing considerably faster per enrollee than Medicare or Medicaid.

Second, nine in ten people with Marketplace coverage have health insurance plans subsidized by the government. The enhanced subsidies expire at the end of December.

An expensive basic necessity is getting even more expensive, and a policy that helps people afford it is being taken away. This type of double whammy has become something of a theme, hasn’t it?

Background on the ACA Marketplace, subsidies

Feel free to skip this part.

Because the US health insurance system is so fragmented, there are major coverage gaps. One is the bloc of 40–50 million people who don’t get health insurance from their jobs, are too young to qualify for Medicare, and earn too much to qualify for Medicaid. There are several obvious solutions to this problem — lower the age threshold for Medicare, raise the income threshold for Medicaid, eliminate thresholds for either/both, etc. What we got instead was the ACA Marketplace. It’s expensive for both the government and health insurance “consumers.”

In addition to the operating and administrative costs of the Marketplace, the US government also spends quite a lot subsidizing private health insurance companies’ premiums. Federal support for the Marketplace totaled $125 billion in 2024. Extending the enhanced subsidies alone is projected to be $35 billion annually, on average — though that estimate is probably too low, because it was made before Marketplace insurers decided to raise their premiums by 26% next year.3 Per enrollee, Marketplace coverage is far more expensive than Medicaid.

Immense federal subsidies notwithstanding, Marketplace coverage is still quite expensive for health insurance “consumers” ($500 per month/$6,000 per year for one of the cheaper individual plans). There are over 27 million uninsured Americans, and about two-thirds say their lack of coverage is due to the high cost of health insurance.

Democrats caving on health care is worse than you think

The looming expiration of Marketplace subsidies doesn’t help Republicans, but it’s likely more politically damaging to Democrats.

The bulk of the media critique has focused on the pattern of Democrats failing to stand up to Trump, thereby alienating their base. As a result, you’ve likely recently heard critics describe the Democratic Party as “feckless,” “cowardly,” “weak,” “if irritable bowel syndrome were a political party,” and so on.

A related but more important pattern is the Democratic Party’s alienation of working-class voters. Democrats reneging on their no enhanced subsidies, no budget deal ultimatum isn’t just bad optics — it betrays the lower-income people the previous Democratic administration funneled from Medicaid to the Marketplace, promising affordable coverage.

1. Historic Marketplace enrollment, historic Medicaid disenrollment

Biden campaigned on record-level growth in Marketplace enrollment during his abbreviated 2024 reelection bid and listed it as one of his top healthcare achievements near the end of his term. In January 2025, the Biden administration announced that it set “another all-time record” for Marketplace sign-ups. Biden himself boasted that “nearly 24 million Americans have signed up for Affordable Care Act coverage. That means that enrollment has nearly doubled since I took office. That’s no coincidence.”

Biden actually sold himself short there — Marketplace enrollment for 2025 ended up at 24.3 million, more than double the number he inherited in 2021:4

• 2018: 11.8 million

• 2019: 11.4 million

• 2020: 11.4 million

• 2021: 12.0 million

• 2022: 14.5 million (+2.5M)

• 2023: 16.4 million (+1.9M)

• 2024: 21.4 million (+5.0M)

• 2025: 24.3 million (+2.9M)

The historic growth in Marketplace enrollment under Biden was made possible by two main factors.

First, the enhanced Marketplace subsidies that made private coverage less expensive, enacted as part of the March 2021 American Rescue Plan. After Marketplace enrollment had basically flatlined after 2015, the introduction of the enhanced subsidies prompted the 2.5 million-enrollee jump from 2021 to 2022 and the smaller but still sizable 1.9 million increase from 2022 to 2023.

Second, the Medicaid “unwinding.” That the largest enrollment jump in the Marketplace occurred from 2023 to 2024 (and second-largest from 2024 to 2025) isn’t a coincidence. As the chart below shows, the record increase in Marketplace signups in 2024 is mostly attributable to the Medicaid purge that began in 2023. From April 2023 to September 2024, 26.8 million people lost Medicaid coverage.5

The Medicaid unwinding was the largest nationwide healthcare coverage transition since the Affordable Care Act itself. Still, some people write off over 25 million losing health coverage because Medicaid enrollment shot up from March 2020 to March 2023.

In early 2020, Congress wisely sought to avoid lower-income Americans being kicked off health insurance during a public health crisis. In March 2020, Congress passed the Families First Coronavirus Response Act, which required states to maintain enrollment of nearly all those on Medicaid in exchange for an increase in federal funding to accommodate the larger enrollment (see Sec. 6008). This became known as the “continuous enrollment” or “continuous coverage” condition.6 This condition would be active so long as the COVID-19 Public Health Emergency (first declared on January 31, 2020, and renewed by the acting administration many times thereafter) remained in effect.7

Because Medicaid was still taking on new participants, enrollment grew from 72 million in March 2020 to 95 million three years later. By March 2023, more than one in four Americans were covered by Medicaid, including half of all children.8

Continuous coverage was excellent from a public health perspective. For example, one study found that

“Medicaid beneficiaries experienced a decline in coverage interruptions in 2021 and 2022 relative to privately insured individuals…with less reporting of unaffordable healthcare needs and delayed medical care due to cost…The continuous Medicaid coverage provision…was associated with enhanced coverage stability and improved access to care.”

The Medicaid unwinding process — which was an administrative disaster — had the opposite effect on public health. For example, one study in the Journal of Addiction Medicine found that unwinding was linked to

“A rise in the number of people ending medication treatment for opioid use disorder, as well as a decrease in the number of people beginning such treatment…changes were greatest in states that have had the largest disenrollments.”

Florida disenrolled more than 400,000 people from Medicaid in the first three months after unwinding began. Texas disenrolled over 500,000 people in a single month, the vast majority for not submitting the proper paperwork. Two-thirds of the 26.8 million people kicked off Medicaid nationwide as of September 2024 were disenrolled for procedural or administrative reasons, meaning they lost coverage because of red tape or the state dropped them in error, not because they were officially determined to be ineligible.

People are still losing their health insurance as part of the Medicaid unwinding — states have until the end of December 2025 to complete eligibility redeterminations.

^Alt text for screen readers: Kicking millions off Medicaid in 2023 drove record ACA Marketplace growth in 2024. This line chart shows the growth in Medicaid and CHIP enrollees under continuous coverage — from 72 million to 95 million — and the decline under unwinding, from 95 million to 78 million. It also shows that Marketplace enrollment grew from 12 million to 16 million after enhanced subsidies started in March 2021, and from 16 million to 24 million as millions were kicked off Medicaid. Enrollment data: KFF, Medicaid.gov. Figures as of January 2025 for Marketplace, June 2025 for Medicaid. Unwinding to be completed by the end of December 2025.

2. A politically engineered Medicaid mass displacement

Nearly ten months after the Medicaid purge started, a Biden administration press release titled “Record Marketplace Coverage in 2024: A Banner Year for Coverage” announced that

“We are now seeing something closer to the full effect of the Biden Administration’s policies to strengthen Marketplace coverage options, which previously had been partially masked by unusually high Medicaid enrollment due to the COVID-era continuous coverage provisions.”

As that quote suggests, the Biden administration’s intent was to push people from Medicaid into the Marketplace. It made sure that happened by: 1) extending Marketplace subsidies through the Inflation Reduction Act in August 2022, 2) signing off on Medicaid unwinding in December 2022, and 3) channeling those disenrolled from Medicaid into the Marketplace after unwinding began in April 2023.

I understand that attributing blame to Biden for mass Medicaid disenrollments is considered controversial, but it really shouldn’t be. These are facts:

• Medicaid continuous coverage was linked to the COVID-19 Public Health Emergency — coverage would remain in effect so long as the emergency declaration did, which the administration can (and did) repeatedly renew.

• Medicaid continuous coverage did not require funding from Congress — it was funded by formula-driven mandatory spending (not congressionally-approved discretionary spending).

• The President could end Medicaid continuous coverage by a) signing a bill delinking continuous coverage from the Public Health Emergency, or b) declaring an end to the Public Health Emergency. Biden did both.

• Biden’s budget for 2023 — enacted through the Consolidated Appropriations Act, 2023 — delinked Medicaid continuous coverage from the Public Health Emergency, and set April 1, 2023, as the start date for gradually defunding continuous coverage and allowing states to resume disenrollments (see Div. FF, Title V, Sec. 5131). After urging its “swift passage,” Biden signed the legislation into law on December 29, 2022, as the last spending bill enacted by the Democratic trifecta. All but one Democrat voted for it.9

• Had the bill not included that provision, Biden would have triggered Medicaid unwinding anyway after ending the COVID-19 Public Health Emergency on May 11, 2023.

A common defense of Biden’s behavior from Democratic partisans is that Medicaid continuous coverage originated as a temporary provision, so it should be temporary. That’s precisely the same argument Republicans are making now in opposing the extension of the enhanced ACA subsidies:

• Democrats (2023–24): Medicaid continuous coverage originated among many temporary emergency measures authorized by legislation passed between March 2020 and March 2021 in response to COVID-19. The Medicaid unwinding was just a return to the status quo ante. In fact, more people are enrolled in Medicaid now than before the pandemic!

• Republicans (2025): The ACA Marketplace enhanced subsidies were one of many temporary provisions enacted through the 2021 American Rescue Plan and other emergency legislation. The expiration of the subsidies is just a return to the status quo ante. In fact, despite the projected drop in Marketplace coverage, enrollment will still be higher than it was before the pandemic!

3. Migration to the Marketplace

In January 2023, the Biden administration created a Special Enrollment Period for people losing Medicaid to sign up for Marketplace coverage without waiting for the typical open enrollment period. It was scheduled to begin March 31, 2023 — the day before Medicaid unwinding began. Predictably, there was massive and highly unusual mid-year growth in Marketplace signups. Per KFF:

“Unlike most previous years, the individual market grew mid-year in 2023, outside the open enrollment window and at a time when attrition normally occurs. From early April to the end of September 2023, enrollment in the individual market grew by 5.7%, or just over 1 million enrollees.”

In March 2024, Biden extended the Special Enrollment Period to accommodate the spiraling number of Medicaid disenrollments. Per Politico:

“The move aims to minimize the number of people losing health insurance coverage in the run-up to the November election as a result of a nationwide purge of state Medicaid rolls…a process that’s resulted in the biggest reshuffling of the health insurance landscape since the passage of Obamacare itself.”

Biden administration officials worked to transition those who lost Medicaid coverage to the Marketplace. However, Marketplace enrollments didn’t appear to offset the effect of Medicaid unwinding, at least on children. Over 5.5 million kids lost public health insurance coverage from April to September 2023 alone. The uninsured rate among children in 2024 was higher than any year since at least 2017:

• 2017, 5.0%

• 2018, 5.5%

• 2019, 5.2%

• 2020, 5.6%

• 2021, 5.0%

• 2022, 5.4%

• 2023, 5.8%

• 2024, 6.1%

By June 2024, at least 3.5 million predominantly lower-income people kicked off Medicaid had enrolled in the Marketplace. As the Biden administration announced in 2024:

“Nearly 4.2 million more individuals with household incomes under 250% of the federal poverty level enrolled in [Marketplace] coverage compared to last year. This further indicates that lower-income individuals and families have enrolled…as states continue their post-COVID redeterminations of Medicaid.”

Lower income people made up the majority of the growth in the Marketplace enrollment under Biden: For example, 83% of the enrollment growth in the ACA Marketplaces from 2020 to 2024 was attributable to those with incomes below 2.5 times the federal poverty level.

Millions of them will lose coverage due to cost if there isn’t an extension of the enhanced Marketplace subsidies — the subsidies Democrats just caved on in the shutdown standoff. Many of these enrollees have Marketplace coverage in the first place because they were kicked off Medicaid by a bill enacted under a Democratic President, House, and Senate. What message are Democrats sending to the working class?

SPECIAL THANKS TO: Abe B., Alan F., Alissa Q., Amin, Andrew R., AT., B. Kelly, BartB., BeepBoop, Ben, Ben C.,* Bill S., Bob N., Brett S., Byron D., Carol V., Chris, Chris G., Cole H., D. Kepler, Daniel M., Dave, David J., David S.,* David V.,* David M., Elizabeth R., Errol S., Foundart, Francis M., Frank R., Gary W., Gladwyn S., Graham P., Griffin R., Hunter S., IBL, Irene B., Isaac, Isaac L., Jacob, James G., James H., James N., Jamie LR., Jcowens, Jeff, Jennifer, Jennifer J., Jessica S., Jerry S., Joe R., John, John, John A., John K., John M., Jonathan S., Joseph B., Joshua R., Julia G., Julian L., Katrina H., Keith B., Kheng L., Lea S., Leah A., Leila CL., Lenore B., Linda B., Linda H., Lindsay, Lindsay S.,* Lora L., Mapraputa, Marie R., Mark L., Mark G., Marvin B., Mary Z., Marty, Matthew H.,* Megan., Melanie B., Michael S., Mitchell P., Nick B., Noah K., Norbert H., Omar A., Omar D.,* Peter M., Phil, Philip L., Ron C., Rosemary K., Sari G., Scarlet, Scott H., Silversurfer, Soh, Springseep, Stan C., TBE, Teddie G., Theresa A., Themadking, Tim C., Timbuk T., Tony L., Tony T., Tyler M., Victor S., Wayne H., William P.

* = founding member

-Stephen (Follow me on Instagram, Twitter, and Bluesky)

Subscribe now

Share

*This note is for Polygraph VIPs only. Read on if you are one, or support my work to become one.

Read more

Share

*Please consider joining the esteemed list of paid subscribers thanked at the bottom of each note. They make research like this possible.

Subscribe now

Four contractors, four years, $89 billion in buybacks and dividends

In Thursday’s newsletter, I reported that four top military contractors — Lockheed Martin, Raytheon (now RTX), General Dynamics, and Northrop Grumman — increased spending on stock buybacks and dividends from $39 billion under Trump to $89 billion under Biden, based on my review of each company’s 10-K filings from 2017–2024.

In this newsletter, I ask the following: How much of the $89 billion spent on shareholders over the last four years came from US taxpayers? (Answer: about two-thirds. Read on for more.)

Subsidizing shareholder payouts

While corporations in general are increasingly focused on rewarding shareholders — particularly through buybacks — military contractors are not normal corporations. To paraphrase economist John Kenneth Galbraith, the big arms firms are really public firms. Here are the top four US arms firms1 with the share of their 2021–2024 revenue from US government contracts:

• Lockheed Martin: 73%

• Raytheon (RTX): 45%

• General Dynamics: 67%

• Northrop Grumman: 86%

Based on the percentage of each company’s revenue that came from the US government, an estimated $58 billion of the $89 billion that Lockheed Martin, Raytheon (RTX), General Dynamics, and Northrop Grumman spent on buybacks and dividends from 2021–2024 was financed with public funds.2

I arrived at this estimate by recording each company’s sales to the US government, dividing those sales by total revenue, then multiplying that share by the amount spent on buybacks and dividends. As an example, here’s what the Lockheed Martin calculation looks like:

• 2024 revenue: $71,043,000,000

• 2024 sales to US government: $52,044,000,000

• 2024 buybacks + dividends: $6,759,000,000

• US government sales/revenue = 0.7325704151

• (Buybacks + dividends) × (US government sales/revenue) = $4,951,443,436

• Est. taxpayer-funded buybacks + dividends = $4,951,443,436

• Repeat process for 2023, 2022, and 2021.

• Add estimated annual totals together.

• Insert four-year sum ($24,647,931,510) into pumpkin-spice-themed chart, as seen below. Pumpkin spice optional.

I repeated this process 12 more times (three remaining firms × four years).3 These estimates assume that the share of buybacks and dividends funded by US taxpayers reflects the share of company revenue from US government sales, and are conservative estimates for that reason.4 Revenue from foreign arms sales brokered by the US government, which can also be partially funded by the US government, is accounted for separately in company 10-K reports and is therefore excluded from this analysis.

*

To review, four de facto public firms in four years spent $89 billion rewarding shareholders, an estimated two-thirds of which came from public funds that could have otherwise gone to helping people afford groceries. This $58 billion in taxpayer-funded buybacks and dividends would be enough to deliver SNAP benefits to its 41.7 million recipients for the next seven months.

SNAP recipients will reportedly only receive partial benefits this month. It’s a problem, apparently, to subsidize groceries, but it’s somehow fine to subsidize arms company shareholders. Consider spicing up your next holiday gathering by bringing this up at the dinner table.

^Alt text for screen readers: Four military contractors spent $89 billion on buybacks and dividends in four years. Taxpayers funded $58 billion of it, based on each company’s government revenue. Stock buybacks and dividends, by contractor, in billions of dollars: Lockheed Martin, 34; Raytheon (RTX), 31; Northrop Grumman, 14; General Dynamics, 11; total, 89. Estimated stock buybacks and dividends funded by taxpayers: Lockheed Martin, 25; Raytheon (RTX), 14; Northrop Grumman, 12; General Dynamics, 7; total, 58. Data: 2021 to 2024. Taxpayer-funded buybacks and dividends estimated by multiplying reported buybacks and dividends in company 10-K filings by each firm’s share of revenue from U.S. government sales. Company totals may not match column totals due to rounding.

SPECIAL THANKS TO: Abe B., Alan F., Amin, Andrew R., AT., B. Kelly, BartB., BeepBoop, Ben, Ben C.,* Bill S., Bob N., Brett S., Byron D., Carol V., Chris, Chris G., Cole H., D. Kepler, Daniel M., Dave, David J., David S.,* David V.,* David M., Elizabeth R., Errol S., Foundart, Francis M., Frank R., Gary W., Gladwyn S., Graham P., Griffin R., Hunter S., IBL, Irene B., Isaac, Isaac L., Jacob, James G., James H., James N., Jamie LR., Jcowens, Jeff, Jennifer, Jennifer J., Jessica S., Jerry S., Joe R., John, John, John A., John K., John M., Jonathan S., Joseph B., Joshua R., Julia G., Julian L., Katrina H., Keith B., Kheng L., Lea S., Leah A., Leila CL., Lenore B., Linda B., Linda H., Lindsay, Lindsay S.,* Lora L., Mapraputa, Marie R., Mark L., Mark G., Marvin B., Mary Z., Marty, Matthew H.,* Megan., Melanie B., Michael S., Mitchell P., Nick B., Noah K., Norbert H., Omar A., Omar D.,* Peter M., Phil, Philip L., Ron C., Rosemary K., Sari G., Scarlet, Scott H., Silversurfer, Soh, Springseep, Stan C., TBE, Teddie G., Theresa A., Themadking, Tim C., Timbuk T., Tony L., Tony T., Tyler M., Victor S., Wayne H., William P.

* = founding member

-Stephen (Follow me on Instagram, Twitter, and Bluesky)

Subscribe now

Share