Most geographically concentrated agencies

≈ computed · share in one state

FY2026 ended Sep 30 · data still posting · data as of 2026-11-01†

† DoD contract data posts to USAspending about 90 days after award.

Federal agencies ranked by how much of their award money lands in ONE state. Some agencies spread work across the country; others concentrate almost all of it in a single place, usually around one laboratory, shipyard or installation. Both facts are visible in the source records: every award names its awarding agency and its own place of performance, and this ranking simply pairs the two. Computed

  1. 1
    Department of Veterans Affairs$33.2B of $58.3B in VA
    57%
  2. 2
    Department of Homeland Security$26.1B of $46.5B in TX
    56%
  3. 352%
  4. 4
    Department of Energy$1.8B of $4.1B in MD
    43%
  5. 529%
  6. 627%
  7. 7
    Department of Defense$16.6B of $79.6B in TX
    21%
  8. 8
    Department of Transportation$5.4B of $32.2B in CA
    17%
  9. 9
    Department of the Interior$312.9M of $1.9B in NM
    17%
  10. 1016%
  11. 11
    Department of Agriculture$6.9B of $47.7B in CA
    14%
  12. 1214%
  13. 13
    Department of Education$2.7B of $22.7B in CA
    12%
← All leaderboards

Computed from the individual award records loaded here, which are each year's largest contracts and grants rather than every award. Both inputs are from source (an award names its awarding agency and its own place of performance), but the SLICE is ours: this describes where an agency's big awards land, not all of its spending. There is no agency-by-place totals table in the source, and splitting a recipient's geography across its funders would be an inference dressed as a fact.

Why the labels? The graph is only worth anything if the links are trusted. Facts taken straight from a federal filing are Source; figures we calculate are Computed; links we infer (corporate parents, districts, similar contractors, connections between records) are Inferred, worded carefully, never asserted as fact and never a claim that one thing caused another. Spot an error? Tell us.