The $700M Plan to Put Life Back Into American Farmland

Washington pledges $700m to fix American farmland and the agency meant to deliver it is already short-staffed

26 Mar 2026

A quarter of American farmland is losing soil to water erosion. Another sixth is blowing away. Against that backdrop, the federal government has chosen this moment to consolidate its bets on regenerative agriculture.

The USDA has opened applications for a $700m Regenerative Pilot Program, splitting the funds between two established conservation schemes. The more significant change is not the money but the mechanism: for the first time, farmers can apply for soil health, water management, and biodiversity support through a single submission. Outcomes will be tracked at the individual farm level, nudging the program toward performance rather than paperwork.

The program spans an unusual political coalition. Agriculture Secretary Brooke Rollins and Health Secretary Robert F. Kennedy Jr. announced it together, framing healthy soil as a precondition for nutritious food. A new advisory council will pull together producers, food companies, and consumer advocates for quarterly reviews with agency leadership. Private co-investment is also permitted under existing federal rules, potentially stretching the headline figure further.

The structural challenge is familiar. The Natural Resources Conservation Service, which administers the program, has seen significant staff reductions in recent months. Policy experts have raised concerns about whether the agency can process the volume of applications the funding is likely to attract. The USDA says it will expand its network of certified technical service providers to compensate.

That caveat matters. Consolidating the application process reduces friction for farmers, but someone still has to review the submissions, approve the contracts, and track the outcomes. A performance-based model is only as good as the capacity to measure performance.

For an agency stretched thin, a program designed around accountability may prove the hardest kind to run.