No mortality model exists for most DOGE cuts — where one does, the largest projects about 14 million added deaths by 2030

Asked (summary)

Map all cuts made by DOGE against the expected loss of life projected from each cut, ranked in order of number of projected casualties.

Most DOGE contract, grant, lease and staffing cuts have no published numerical mortality model, so “all cuts” cannot honestly be assigned casualty numbers. This report instead covers every distinct published mortality estimate located in a live-web scan through 11 September 2026 for 2025–2026 federal cuts associated with DOGE or that DOGE helped execute: 23 estimate/scenario rows from at least seven source hosts. Figures are modeled additional deaths, not observed deaths; horizons range from one year to 2050, and several rows overlap or are components of larger totals rather than additive.

All 23 published estimates, ranked by upper bound

Each bar spans the low-to-high published range on a logarithmic scale (thousand-scale and million-scale estimates stay legible). A tick marks the point estimate where one is given. Amber rows are declared components of a larger total above them — do not add rows together. Hover or tap a row for scenario, method and source.

Global / international estimate US domestic estimate Declared component of a larger total point estimate

Where the deaths are projected to occur

The question asks for a map, but the geographies in the underlying studies are textual — “55 countries”, “26 high-burden TB countries”, “globally” — with no country-level allocations published. Drawing them onto a world map would invent precision the sources do not contain, so geographic scope is shown as stated. Only one estimate (Medicaid, 28,785 deaths) is US-domestic; nearly all projected mortality falls in low- and middle-income countries served by USAID, PEPFAR and TB programmes.

Findings

The 14.05 million by-2030 projection for current USAID cuts is the single largest figure located, and the 4.5 million maternal/child figure is a declared component of it, not an addition — thelancet.com

Scenario choice moves the TB estimate twenty-fold: 65,200–130,000 deaths in the minimal case versus 1.57–2.81 million in the worst case across the same 26 countries — thelancet.com

The only US-domestic projections are 28,785 excess deaths in the first year of proposed Medicaid cuts and about 200,000 premature deaths through 2050 from rolling back 31 EPA rules — an administration action associated with DOGE-era cuts rather than a DOGE line item — bu.edu, environmentalprotectionnetwork.org

The 400,000–1.71 million PEPFAR budget-crisis range comes from a MedRxiv preprint that has not completed peer review — medrxiv.org

All rows and sources

#	Cut / programme	Scenario & horizon	Geography	Low	High	Method	Source

Method: live-web scan through 2026-09-11 across more than 1,500 unique search results, with extraction from peer-reviewed papers, university summaries, modelling studies and documented impact counters. 23 rows, one per distinct published mortality estimate or scenario; numbers are modeled additional (projected) deaths, not verified deaths. Sources span seven hosts; no single URL backs more than half the rows. Rows overlap across sources, assumptions and horizons (one year to 2050) and must not be summed. Most DOGE cuts have no published mortality model and are therefore absent. Long scenario text is abbreviated in chart labels; full detail is in the table tooltips and sources.

This report was generated automatically by Keenable SELECT at a user's request, from publicly available web sources linked herein. Keenable does not review, verify, or endorse its contents and makes no representation as to accuracy, completeness, or timeliness; AI-based extraction may contain errors. Nothing in this report is investment, legal, financial, or other professional advice. All trademarks and referenced content remain the property of their respective owners; no affiliation or endorsement is implied. To report an error, rights concern, or request removal: legal@keenable.ai.