For the Republic
← Command Center / 📄 Article / 2026-02-14

The Arsonist's Invoice: DOGE at One Year

Draft Complete — Pending Author Review

Thesis

2/10
thesis.md

Article Thesis

Working Title

The Arsonist's Invoice: DOGE at One Year

Subtitle

They promised to cut the waste. They cut the government instead -- and sent us the bill.

Thesis

DOGE did not fail because it was poorly executed. DOGE failed because it was built on a theory of government that is empirically, structurally false: that the federal workforce is where the money is, and that cutting workers equals cutting waste. Federal employee compensation is roughly 8% of total spending. DOGE took a chainsaw to the 8% and left the 92% untouched -- producing what might be the most expensive "efficiency" initiative in American history. The result is negative returns on destruction: a government that is simultaneously more expensive and less capable, verified by the Cato Institute, Yale Budget Lab, and the administration's own dissolving math. But the deeper story is not incompetence -- it is intent. Russell Vought's quiet institutionalization of DOGE at OMB, with a $45 million budget and 150 permanent staff, reveals the endgame: not a government that works better, but a government that cannot work at all.

The Framework

Negative returns on destruction. This is the conceptual engine of the article -- the reusable phrase the reader should walk away with and apply to future policy debates.

The framework works on three levels. First, financial: DOGE's own ledger is underwater. Claimed savings of $214 billion collapse under scrutiny to somewhere between $1.4 billion and $11.7 billion in verified reductions. Meanwhile, the hidden costs -- paid leave for fired-then-reinstated workers, litigation defense, lost IRS enforcement revenue, rehiring and retraining -- conservatively total $135 billion in the first year alone, with Yale Budget Lab projecting $198-323 billion in lost tax revenue over a decade. The initiative designed to save money cost money. Second, institutional: 106,636 years of federal work experience walked out the door. 10,109 STEM PhDs -- 14% of the government's doctoral workforce -- are gone. At research agencies, departures outnumber new hires 11 to 1. The enforcement mechanisms (IRS), the service delivery systems (SSA, VA), and the scientific capacity (NIH, NOAA, EPA) that make government actually function have been degraded in ways that will compound for a decade. Third, democratic: a government that cannot collect taxes, process benefits, inspect food, or respond to disasters is not a government that has been made "efficient." It is a government that has been incapacitated -- and incapacitation is not an accident. It is the product.

The framework draws explanatory power from the gap between the stated purpose (efficiency) and the actual mechanism (capacity destruction). It is the same dynamic as a private equity firm that buys a company, strips its assets, loads it with debt, declares it "restructured," and walks away with a fee -- leaving behind something that technically still exists but can no longer do its job. DOGE is the leveraged buyout of the federal government. The "savings" were the marketing. The product was incapacitation.

Why This Matters Now

February 2026 marks one year since DOGE began operating. The spectacle phase is over -- Musk departed in May 2025, the entity was formally dissolved in November 2025, and the media has largely moved on. But the damage is not over. It is compounding. IRS enforcement degradation is a slow-moving phenomenon: taxpayers are rational actors who adjust their compliance behavior as they observe weakened oversight. The UK's experience with HMRC staff cuts -- 42 billion pounds in uncollected tax, with a deficit worse after austerity than before -- shows exactly where this trajectory leads, and it took years to fully manifest. Meanwhile, Vought's OMB is quietly embedding DOGE operatives as permanent "in-house consultants" across federal agencies, converting them to political positions, and requesting budget increases for the oversight apparatus even as every other agency shrinks. The fire is out. The arsonist left. But the building is still collapsing -- and someone is moving into the rubble with blueprints for something new. The one-year mark is the moment to take stock before the damage becomes the new normal.

The Hook

Open with the invoice. Not a metaphor -- a literal itemized bill. The way a contractor would present it: line items, quantities, costs. "Here is what the Department of Government Efficiency cost the American taxpayer in its first year of operation." Claimed savings: $214 billion. Verified savings: $1.4-$11.7 billion. Hidden costs (paid leave, rehiring, lost productivity): $135 billion. Projected lost IRS revenue: $198-323 billion over a decade. Litigation costs: uncounted. Institutional expertise destroyed: 106,636 years. Services degraded: 6 million SSA cases backlogged. Net balance: deep in the red. The invoice format does the rhetorical work -- it uses DOGE's own language of fiscal accountability against it. They wanted receipts. Here are the receipts.

Key Evidence & Sources

  • Cato Institute (Dec 2025): Federal spending rose $248 billion in 2025 despite the largest peacetime workforce cut on record (271,000 workers, 9%). "No noticeable effect on the trajectory of spending." This is the single most devastating data point because it comes from a libertarian think tank ideologically sympathetic to government reduction.
  • Yale Budget Lab: IRS cuts projected to lose $198-323 billion in tax revenue over a decade. Every dollar spent on IRS enforcement yields $5-12 in revenue. Cutting IRS staff is mathematically the most fiscally destructive thing a government can do.
  • WBUR/On Point (Jan 2026): All 13 of the largest claimed savings items were incorrect. ICE contract claimed at $8 billion; actual value $8 million -- a 1,000x error. NYT Pulitzer winner David Fahrenthold: "DOGE didn't cut a dollar of federal spending."
  • Partnership for Public Service/CBS News (Apr 2025): Hidden costs estimated at $135 billion in FY2025 alone -- excluding litigation, lost revenue, and long-term knowledge loss. Max Stier: "We were seeing the opposite" of addressing waste.
  • Science/AAAS (Jan 2026): 10,109 STEM PhDs departed federal service -- three times the 2024 rate. 106,636 years of experience lost. At 14 research agencies, departures outnumber hires 11:1. NIH lost 1,100+ doctoral staff (vs. 421 in 2024).
  • SSA backlogs (multiple sources, 2025): 6 million pending cases, 600,000 retirement claims backlogged (71% increase), 2.5-hour average call wait times. Whistleblower alleged DOGE uploaded Americans' sensitive data to an unsecured server.
  • UK HMRC parallel (Bloomberg, UK Parliament): Britain's tax authority lost 42 billion pounds in uncollected tax after austerity-driven staff cuts. ROI on enforcement: 18 pounds per 1 pound spent. UK deficit was worse after austerity than before.
  • American Prospect (Feb 2026): Vought institutionalized DOGE at OMB with $45M budget, 150 permanent staff, 4% OMB growth -- while every other agency shrinks. All remaining DOGE staff converted to political positions.

Argument Arc (Brief)

Movement 1: The Invoice. Open with the itemized bill. Establish the central absurdity: the efficiency initiative cost more than it saved. Use the Cato Institute as the credibility anchor -- when the libertarians say it did not work, the debate is over. Walk through the numbers: promised vs. claimed vs. verified vs. actual costs. The gap is not rounding error. It is fabrication at industrial scale.

Movement 2: Where the Money Actually Is. Zoom out to the structural argument. Federal employee compensation is 8% of spending. The other 92% is transfer payments that workforce cuts cannot touch. DOGE attacked the wrong target by design -- because the right targets (entitlements, defense) require congressional action and political courage. The chainsaw was always aimed at the part of government that was small enough to cut unilaterally.

Movement 3: What Was Actually Destroyed. Shift from dollars to capacity. The STEM PhDs, the SSA backlogs, the IRS enforcement collapse, the 6 million people waiting for benefits they already earned. This is where "negative returns on destruction" becomes human. Use the UK HMRC parallel surgically: one paragraph, devastating, historically confirmed.

Movement 4: Steel-Man, Then Shatter. Acknowledge what DOGE did right -- doge.gov, the Treasury payment tracking fix, fraud detection. These are real. Then make the argument that matters: you did not need to fire 271,000 people to build a website. Technology modernization and mass destruction are separate initiatives that were deliberately conflated. The good things DOGE accomplished could have been done without the bad. The bad things were the point.

Movement 5: The Arsonist Moves In. The forward-looking close. Vought's institutionalization of DOGE at OMB -- the $45 million budget, the 150 permanent staff, OMB growing while everyone else shrinks. The spectacle is over. The permanent incapacitation machinery is just beginning. End with the reusable insight: negative returns on destruction is not a bug. It is the business model of a movement that has always believed government should not work -- and is now ensuring that it cannot.

The "So What?"

The reader should walk away understanding three things they likely did not grasp before. First, that DOGE was not merely wasteful or chaotic -- it was structurally incapable of achieving its stated goal, because it targeted the 8% of federal spending that goes to salaries while leaving the 92% untouched. The theory of government was wrong before the first employee was fired. Second, that the costs of capacity destruction compound over time -- the IRS revenue losses, the expertise drain, the service degradation are not one-time events but accelerating curves, and the UK parallel proves it. Third, and most importantly, that DOGE's "failure" to cut spending was not an accident but a feature: the institutionalization at OMB reveals that the goal was never efficiency but permanent incapacitation of the federal government's ability to regulate, enforce, and serve. The reader should have a new phrase -- "negative returns on destruction" -- and a new lens for evaluating any future initiative that promises to fix government by breaking it.

Potential Pitfalls

  • The Penn Wharton data. Tax collections through April 2025 were broadly in line with projections. If we overstate the IRS revenue collapse, critics will point to this. Handle it carefully: acknowledge the short-term data, then explain that audit compliance is a slow-moving phenomenon. Taxpayers adjust over years, not months. The UK data confirms the lag.
  • Technology modernization is real. If the article dismisses DOGE's genuine achievements -- doge.gov, payment tracking, fraud detection -- it loses credibility with fair-minded readers. Steel-man honestly, then distinguish between reform and destruction.
  • "Government needed reform" is a popular position. 72% of Americans support the concept of government efficiency. The article must not come across as defending the status quo. Acknowledge that legacy systems were outdated, that some processes were genuinely inefficient, that modernization was overdue. Then make the case that reform and demolition are different things.
  • The $135 billion cost figure is from one analysis. The Partnership for Public Service is credible and nonpartisan, but the estimate has not been independently replicated. Use it as an estimate, not a fact carved in stone.
  • Tone risk. The temptation is rage. But the most devastating version of this article is clinical, not furious -- the invoice format itself implies cold, factual accounting. The anger should come from the precision, not from the rhetoric. Let the numbers do the emotional work.

Research Assessment

The source material is exceptionally strong -- 15 substantive sources spanning financial analysis (CBS/PSP, Cato, Yale Budget Lab, NPR, Newsweek), human impact (Federal News Network, SSA backlogs, Science/AAAS), institutional analysis (Kettl/GovExec, American Prospect), historical parallel (UK/HMRC), counterarguments (steelman compilation, Penn Wharton), public opinion (polling compilation), and legal costs (lawsuit compilation). The research summary is thorough and includes a well-developed evidence map.

Minimal supplemental research is needed. The main gaps are: (1) precise litigation costs, which no source has totaled -- this is a known unknown that should be flagged rather than fabricated; (2) post-2025 IRS revenue data to confirm the Yale projections are materializing, which simply does not exist yet; and (3) contractor spending increases that may have offset workforce savings, which is a strong hypothesis without definitive data. These gaps are manageable -- the article can acknowledge them honestly without weakening the thesis. The existing evidence base is more than sufficient for a 1,500-word article.