Research Summary: The Arsonist's Invoice — DOGE at One Year

Date: 2026-02-14 Format: Article (~1,500 words) Sources gathered: 15

Topic

The Arsonist's Invoice: DOGE at One Year. After one year of the largest peacetime workforce reduction in American history, DOGE's actual ledger shows it cost taxpayers more than it saved, and the damage to federal capacity will compound for a decade.

Thesis Direction

Refined thesis (based on deep research):

DOGE didn't fail because it was poorly executed (although it was). DOGE failed because its theory of government was wrong. The premise — that the federal workforce is where the money is, and that cutting workers equals cutting waste — is empirically, structurally, demonstrably false. Federal employee compensation is roughly 8% of total spending. The vast majority of federal expenditure is transfer payments that workforce cuts cannot touch. DOGE took a chainsaw to the 8% and left the 92% untouched — and in the process, destroyed the enforcement mechanisms (IRS), the service delivery systems (SSA, VA), and the institutional expertise (10,000+ STEM PhDs) that make government actually function.

The result is negative returns on destruction: a government that is simultaneously more expensive (paid leave, rehiring, lost tax revenue, litigation) and less capable (degraded services, lost expertise, demoralized workforce). The enshittification of governance.

The deeper frame: this was always about capacity destruction, not efficiency. Russell Vought's institutionalization of DOGE at OMB reveals the endgame — a permanently weakened federal government that cannot regulate, cannot enforce, cannot serve. The "savings" narrative was the marketing. The product was incapacitation.

Note on thesis evolution: The initial scan framed this primarily as a financial story (costs vs. savings). Deep research revealed it's equally a capacity story and an institutional story. The financial numbers are damning, but the expertise drain (106,636 years of federal experience lost) and the service degradation (6 million SSA cases backlogged) may matter more for the long-term argument. The article should lead with the financial absurdity (it cost more than it saved) but build toward the deeper structural claim (it was designed to break government, and it worked).

Evidence Map

The 15 sources organize into five clusters that build the article's argument:

Cluster 1: The Financial Ledger (Sources 01, 02, 03, 05, 10)

These sources establish the core financial argument — DOGE cost more than it saved.

Source 01 (CBS/PSP analysis): The $135 billion cost estimate vs. $160 billion in claimed savings
Source 02 (Cato Institute): Spending rose $248 billion even as workforce was cut 9%; the libertarian confirmation
Source 03 (Yale Budget Lab): IRS cuts projected to lose $198-323 billion in tax revenue over a decade
Source 05 (NPR): Only $2 billion in verifiable savings vs. $55 billion claimed
Source 10 (Newsweek): Politico verified only $1.4B of $52.8B in contract savings; CBS found 97% overstatement

Cluster 2: The Savings Audit — Checking the Math (Sources 04, 05, 10)

These sources document the systematic inflation of DOGE's claimed achievements.

Source 04 (WBUR/On Point): All 13 largest claimed savings were incorrect; $8B ICE claim was $8M
Source 05 (NPR): Detailed methodology showing DOGE counted ceiling values, not actual spending
Source 10 (Newsweek): MIT economist: "numbers have proved completely unreliable"

Cluster 3: The Human Cost — Services and Expertise (Sources 07, 08, 09, 11)

These sources document what was actually destroyed.

Source 07 (Federal News Network): 317,000 workers lost; 46% of Americans report personal impact
Source 08 (Kettl/Gov Executive): Cuts fell hardest on GS-7/GS-9 workers (youngest, least powerful); SES down 9.4%
Source 09 (Science/AAAS): 10,109 STEM PhDs lost; 106,636 years of experience; 11:1 departure-to-hire ratio
Source 11 (SSA backlogs): 6 million pending cases; 600,000 retirement claims; 2.5-hour wait times

Cluster 4: The Institutional Story (Sources 06, 15)

These sources explain what happened after the spectacle ended.

Source 06 (American Prospect): Vought institutionalized DOGE at OMB; $45M budget request; 150 permanent staff
Source 15 (Lawsuits): 24,000 workers fired and ordered reinstated; multiple court findings of illegality

Cluster 5: Context and Counterargument (Sources 12, 13, 14)

These sources provide historical perspective, the strongest defense, and public opinion.

Source 12 (UK austerity/HMRC): Britain's tax authority lost 42 billion pounds after staff cuts; 18:1 ROI on enforcement
Source 13 (Steelman): Technology modernization, doge.gov transparency, financial tracking improvements, Penn Wharton data
Source 14 (Polling): 72% support "government efficiency" concept; 37% support DOGE specifically; -8 net favorability

Strongest Evidence For the Thesis

Cato Institute confirming spending rose while workforce was cut — this is the single most powerful data point because it comes from an ideological ally of the administration. A libertarian think tank saying "it didn't work" is devastating.
The IRS math — Yale Budget Lab's projection that IRS cuts will lose $198-323 billion in tax revenue over a decade, when the entire annual savings from a 10% workforce cut is approximately $40 billion. The enforcement ROI is $5-12 per dollar spent. Cutting IRS staff is literally throwing away money.
All 13 largest claimed savings were incorrect (WBUR/On Point) — not some, not most, ALL of the top claims. Politico verified only $1.4 billion of $52.8 billion in contract savings. This isn't rounding error; it's fabrication.
106,636 years of federal work experience lost — the STEM PhD departure data from Science/AAAS. This is irreplaceable in any meaningful timeframe. The 11:1 departure-to-hire ratio means the bleeding hasn't stopped.
6 million SSA cases backlogged — real people waiting for retirement and disability benefits they've already earned. This is where the abstraction becomes personal.

Strongest Evidence Against the Thesis

Penn Wharton Budget Model (April 2025): Tax collections through April were "broadly in line" with CBO projections, undermining the $500 billion short-term revenue loss prediction. The feared immediate IRS revenue collapse didn't materialize. Counter-counter: Audit compliance effects are slow-moving; the real damage shows up over years, not months. The Yale Budget Lab projections are about the decade, not the quarter.
Technology modernization was genuine. Doge.gov, the Treasury payment tracking fix, and AI-driven fraud detection represent real improvements. Even critics acknowledge these. Counter-counter: These improvements didn't require firing 271,000 people. Technology modernization and mass layoffs are separate initiatives that were deliberately conflated.
OPM cut contract spend by 50%. A real, measurable reduction from $484M to $242M. Counter-counter: This is $242 million in savings — significant but a rounding error against $7.6 trillion in spending.
The federal workforce arguably was due for modernization. Some legacy systems were genuinely outdated (70-year-old magnetic tapes at GSA). The cultural resistance to performance metrics in parts of government is real. Counter-counter: Reform and destruction are different things. You can modernize without gutting institutional capacity.
92% of departures were "voluntary." Supporters argue this means workers chose to leave. Counter-counter: The "fork in the road" emails were widely viewed as coercive, and many departed workers were probationary employees with the least job protection.

Research Gaps

Precise litigation costs. No source has attempted to total the government's legal defense expenses across dozens of simultaneous DOGE-related lawsuits. The PSP's $135 billion estimate explicitly excludes these costs.
Long-term IRS revenue data. The Penn Wharton data only goes through April 2025. We don't have post-2025 filing season data to confirm whether the predicted revenue losses are materializing. The Yale projections are models, not observed data (yet).
Agency-level cost-benefit analysis. No source provides a complete cost-benefit for individual agencies (what DOGE cost Agency X vs. what it saved). The analysis is all at the macro level.
Contractor spending. Several sources suggest that as federal employees were cut, contractor spending may have increased to fill the gaps. This would further erode the "savings" narrative but I could not find definitive data.
Vought's OMB in detail. The American Prospect piece from February 5, 2026, is the most recent source on institutionalization, but the details of what the embedded DOGE staff are actually doing day-to-day at agencies remain opaque.

Recommended Approach

Lead with the financial absurdity. The headline number — it cost more than it saved — is the hook. Use the Cato Institute as the credibility anchor: even libertarians confirm spending rose.

Build the ledger. Walk through the actual numbers: claimed savings ($214B), verified savings ($1.4B-$11.7B), hidden costs ($135B), projected revenue losses ($198-323B over a decade). The gap between what was promised and what was delivered is staggering at every level of scrutiny.

Go human. The SSA backlogs and the STEM PhD drain are where the abstraction becomes tangible. 6 million people waiting for benefits. 106,636 years of experience walking out the door. This is what "negative returns on destruction" looks like in real life.

The UK parallel is powerful but use it surgically. Don't make this a comparative politics essay. One paragraph: "Britain tried this. HMRC cut staff and lost 42 billion pounds in uncollected tax. For every pound spent on enforcement, they recovered 18. We are running the same experiment and getting the same results."

Steel-man honestly. The technology modernization and transparency improvements are real. Acknowledge them. Then make the argument: you didn't need to fire 271,000 people to build a website. The good things DOGE did could have been accomplished without the destruction. The destruction was the point.

Close with Vought. The institutionalization story is the forward-looking hook. DOGE as spectacle is over. DOGE as permanent incapacitation machinery is just beginning. The $45 million budget request for a permanent DOGE staff embedded at OMB is the tell: they're not trying to fix government. They're trying to prevent it from functioning.

Handle the Penn Wharton data carefully. The fact that short-term tax receipts held up is the strongest counterargument. Acknowledge it directly, then explain why it doesn't disprove the thesis: audit compliance is a slow-moving phenomenon. Taxpayers are rational actors. As they observe that enforcement has weakened, compliance will decline over years, not months. The UK data confirms this pattern.

The framework to name: "Negative returns on destruction." This is the reusable concept the audience should walk away with. It's not just that DOGE didn't save money — it's that destruction as a management philosophy produces negative returns. Worse service at higher cost. The enshittification of governance.

Source Inventory

source-01-cbs-doge-costs-135-billion.md — CBS News/Partnership for Public Service analysis estimating DOGE cost $135 billion in hidden costs against $160B in claimed savings
source-02-cato-workforce-cuts-spending-rose.md — Cato Institute finding: largest peacetime workforce cut on record, but spending rose $248 billion; "no noticeable effect on trajectory of spending"
source-03-yale-budget-lab-irs-revenue.md — Yale Budget Lab projections: IRS cuts will lose $198-323 billion in tax revenue over a decade; $769 billion projected tax gap in 2026
source-04-wbur-on-point-doge-one-year.md — WBUR On Point one-year retrospective: all 13 largest claimed savings incorrect; NYT Pulitzer winner says "didn't cut a dollar of federal spending"
source-05-npr-doge-overstates-savings.md — NPR analysis cross-referencing DOGE claims against federal databases: actual verifiable savings approximately $2 billion vs. $55 billion claimed
source-06-american-prospect-vought-doge-institutionalized.md — American Prospect on Vought institutionalizing DOGE at OMB: $45M budget, 150 permanent staff, OMB growing while everyone else shrinks
source-07-fed-news-network-workforce-transformed.md — Federal News Network year-end retrospective: 317,000 workers lost; 46% of Americans report personal impact; service disruptions across agencies
source-08-govexec-kettl-tail-wagging-dog.md — Donald Kettl analysis for Government Executive: cuts fell on lowest-grade workers, SES down 9.4%, union membership gutted, performance ratings collapsed
source-09-science-stem-phd-brain-drain.md — Science/AAAS: 10,109 STEM PhDs lost, 106,636 years of federal experience, 11:1 departure-to-hire ratio at research agencies
source-10-newsweek-doge-dead-actual-savings.md — Newsweek DOGE obituary: Politico verified only $1.4B of $52.8B in claimed contract savings; CBS found 97% overstatement; MIT economist calls numbers "completely unreliable"
source-11-ssa-backlogs-delays.md — SSA service degradation: 6 million pending cases, 600,000 retirement claims backlogged (71% increase), 2.5-hour wait times, data breach whistleblower complaint
source-12-uk-austerity-hmrc-comparison.md — UK austerity historical parallel: HMRC lost 42 billion pounds in uncollected tax after staff cuts; 18:1 ROI on enforcement; UK deficit WORSE after austerity
source-13-doge-achievements-steelman.md — Strongest counterarguments: technology modernization, doge.gov transparency, financial tracking fix, Penn Wharton tax data, OPM contract reduction
source-14-polling-public-opinion-doge.md — Public opinion: 72% support efficiency concept but only 37% support DOGE; -8 net favorability; 59% concerned about Musk conflicts of interest
source-15-lawsuits-legal-costs.md — Legal challenges: 24,000 workers fired and reinstated by courts; Privacy Act violations confirmed; multiple preliminary injunctions; uncounted litigation costs

The Arsonist's Invoice: DOGE at One Year

Source Material