Historically (since ~mid-20th century) the idea was to create federal agencies that were given lump sums by Congress and which then used a kind of internal peer

Historically (since ~mid-20th century) the idea was to create federal agencies that were given lump sums by Congress and which then used a kind of internal peer review process to determine how to distribute these funds. There's a history of some political intrigue there (universities try to get their people into the federal agency so they can direct funds back to their home university, etc.). This has avoided some of the usual sort of line-item Congressional pork activity, for example (left) Bernie Sanders got the F-35 engine factory for Vermont, and (right) Richard Shelby got the ULA rocket program for Alabama, etc. There have been some coordinated political attacks on federal agencies though, if you look into the history of the USGS, Congress basically said "stop doing environmental pollution research or we'll cut your funding" in the 1990s. https://www.paloaltoonline.com/weekly/morgue/cover/1995_Feb_1.USGS0001.html https://www.paloaltoonline.com/weekly/morgue/cover/1995_Feb_... > "USGS data is considered so reliable and objective that it is often used for testimony during court hearings", Conomos said... "We do scientific studies and basic data collection, in some cases for 100 years or more. We know the history of streams and the environment nationwide. Private consultants don't do that." The NSF is probably among the most independent agencies at present. NIH seems pretty tied up with the pharmaceutical sector, and FDA is even worse. DOE is just entirely captured by the nuclear waste and fossil fuel sector at this point. EPA and USGS, they're basically hamstrung.