Promises are inputs. These are the outputs.
This series has documented a 48-year pattern: knowledge without consequence, then declarations without deeds. This installment converts the pattern into six falsifiable tests — each with a defined pass condition, a place to check, and a status as of September 2026. In the September 2026 snapshot, every test reads pending. This is an archived status assessment, not a live tracker refreshed on each visit. Update this tracker as each one is passed or failed.
The benchmarks are country-neutral by design: any country's farmland purchases deserve the same scrutiny these tests apply.
USDA's 2026 proposed rule would raise late-filing penalties to 1.5% per week (2.5% for adversaries), remove downward-adjustment discretion, and streamline appeals. A proposal is not a rule.
Pass condition: a final AFIDA rule published in the Federal Register, with an effective date and an enforceable penalty schedule — not an advance notice, not a proposal.
Where to check: the Federal Register — search "Agricultural Foreign Investment Disclosure Act" final rules. USDA press releases — compare the announcement against the Register entry; announcements without Register entries are pending.
Status as of September 2026: PENDING — proposed rule stage. No final rule published. No effective date set.
Why it matters: the 1978 schedule already allowed 25% penalties and produced eight assessments in eleven years. A new schedule proves nothing until it is final — and, per Test 2, used. Watch for the Register entry, not the press release.
Eight penalties in eleven years; zero from 2015–2018; typical assessments around 1% against a 25% ceiling. USDA announced a "renewed commitment" to penalties in 2023. The numbers have not been published.
Pass condition: USDA's annual report to Congress includes, for the reporting year: penalties assessed (count and total), penalties collected (count and total), and the valuation basis for each — published, on the record, every year.
Where to check: USDA's annual AFIDA report to Congress — the penalty section, if one appears. Compare against the Arkansas benchmark: maximum rate, collected, publicized.
Status as of September 2026: PENDING — no assessed-and-collected penalty ledger has been published. The 2023 "renewed commitment" remains a statement, not a record.
The test within the test: assessed is not collected. A published list of assessments without collection figures is half a ledger. Demand both — Arkansas collected its $280,000 in 23 days; the federal program should be able to say what it collected, too.
GAO found USDA shares AFIDA data with CFIUS once a year, from a paper-derived process. The reviewers deciding on farmland transactions work from a stale, annual spreadsheet. The 2025 plan promised a USDA–Treasury coordination agreement.
Pass condition: demonstrable transaction-time AFIDA data flow to CFIUS and other national-security agencies — documented in a GAO follow-up, an agency report, or the coordination agreement's published terms.
Where to check: GAO follow-up reporting on GAO-24-106337 recommendations. The published USDA–Treasury/CFIUS agreement — read the data-sharing terms, not the press release. CFIUS annual report references to agricultural-land data sources.
Status as of September 2026: PENDING — agreement promised; no transaction-time feed demonstrated on the public record.
Why it matters: a national-security review process fed by once-annual paper data is security theater with extra steps. The six federal bodies that declared the concern all depend, directly or indirectly, on this feed working. It is the single most consequential plumbing fix in the plan.
The 2025 plan promised an online AFIDA filing system with geospatial data and purchase-purpose information — replacing paper forms filed at county offices — plus a public portal for reporting false or failed compliance.
Pass condition: a working online filing system accepting AFIDA reports with parcel-level geospatial identifiers — verifiable by attempting a filing or by USDA's published system documentation — and a functioning public reporting portal.
Where to check: USDA announcements — then verify the portal exists and accepts filings. Whether new filings in the annual data carry geospatial fields.
Status as of September 2026: PENDING — promised; no operational system demonstrated on the public record.
The credibility condition: a portal that exists but is optional, while paper filing continues unchanged, does not modernize the system — it adds a website to it. The test is whether the geospatial data actually appear in the filings the public and CFIUS receive.
In 2024, 59.9% of AFIDA records carried a zero or blank purchase price — up from 33.6% in 2010. Purchase Price is separate from Current Value and Estimated Value. A zero or blank purchase price does not establish that a market valuation is unavailable or that a penalty cannot be assessed.
Pass condition: USDA publishes the zero/blank purchase-price share as an official data-quality metric in each annual report — and the share declines year over year.
Where to check: each year's AFIDA holdings file — recompute the share. The annual report to Congress — look for the metric, not just the totals.
Status as of September 2026: PENDING — the share is rising, not falling; no official data-quality metric published.
Under FIRRMA (2018), CFIUS can review certain foreign real-estate transactions for national-security risk. Farmland transactions are in scope. Public outcomes specific to agricultural land are scarce.
Pass condition: CFIUS's annual report to Congress — or Treasury publications — disclose agricultural-land transaction outcomes: reviews opened, mitigation agreements, or blocks involving farmland, in identifiable form.
Where to check: CFIUS annual report to Congress (Treasury). Treasury press releases on real-estate transaction actions.
Status as of September 2026: PENDING — no standing public series of farmland-specific CFIUS outcomes identified.
Fairness note: CFIUS operates under confidentiality constraints by statute; full transaction-level disclosure is not expected. The test is for aggregated, agricultural-land-identifiable outcomes — enough for oversight, consistent with the committee's existing public reporting.
How to update this tracker: each test flips to passed only on public, verifiable evidence meeting its pass condition — a Register entry, a published ledger, a documented feed, a working portal, a published metric, a reported outcome. Announcements, proposals, and MOUs do not flip a test. When a test flips, note the date and the evidence. If a test fails definitively (e.g., a final rule that weakens enforcement), mark it failed with the reason. This archived tracker requires a dated source review before its statuses are treated as current. Pending is not failed. The tracker's discipline is to record only what the public record shows — and to let the accumulation of pendings speak for itself.
This series held up the government's own declarations: Congress built the disclosure law in 1978 and received the reports every year. USDA held 25% penalty authority for 48 years and used it eight times in eleven. The 2018 Farm Bill ran 529 pages without a word on foreign land while the files reported 32.53 million foreign-held acres. Arkansas showed enforcement was a choice. Twenty-nine states legislated. GAO audited. Six federal bodies declared farm security national security.
At no point was the government wrong about the concern — only slow to act on it. The concern was interdepartmental before it was ever this publisher's: Agriculture, Defense, Homeland Security, Justice, Treasury, and the government's own auditor, each on their own letterhead.
"You said this yourselves — the threat, the numbers, the penalties, the plan. You were right all along. Six tests stand pending. The next move belongs to the institutions that made the declarations."
DISCLAIMER: This article is a policy research product, not legal advice and not a finding of wrongdoing. Its strongest use is as a standing oversight checklist: bring it to every hearing, every USDA briefing, and every farm-bill markup, and ask which test flipped since the last one.