Changes in Priorities

This post has a reminder, a new call for comments, a piece of good news, and a bonus.

The reminder: as I requested last month, please comment on the Heavy Duty Diesel rollback, due August 29th (there are only 32 comments posted so far, please add to those!).

The new call for comments: a proposed NSF guidance document would prioritize politics over quality research, so please comment by the August 24th deadline (see below for my comment).

The good news: The recent Senate spending bill included language that would block the White House from implementing the proposed changes that I criticized in June. This bill passed 90–6 (backed by Susan Collins, who moved past “concern” for once), but the key paragraph is not in the House version. If you actually have a voting representative in the House, please contact them and tell them to make sure this paragraph stays in when the House and Senate work out the differences in September!1

The bonus: I include some recommendations for politics blogs at the end of the post.

To Whom It May Concern,

Re: Comment on Draft NSF Guidance on Financial Assistance, Docket No. NSF-2026-OTR-0001 (Information Collection 3145-0058), Guide 25, Suspensions, Terminations, and Research Misconduct

I write to oppose the termination provisions of proposed Guide 25, and specifically the inclusion of “changes in priorities” among the listed reasons NSF may end an award (p. 197). I ask that NSF strike that phrase and state the complete grounds for termination in the text of the Guide.

[G25.A.Policy 1] Guide 25 does not state the grounds on which NSF may terminate an award, as 2 CFR 200.340(b) requires.

Guide 25’s description of grounds for termination is buried in a cross-reference: “NSF may suspend or terminate an award for cause in whole or in part, pursuant to a) 2 CFR 200.339 and 2 CFR 200.340 or b) NSF’s Regulation on Research Misconduct (45 CFR 689)” (p. 197). But 2 CFR 200.340(b) states that an agency “must clearly and unambiguously specify all termination provisions in the terms and conditions of the Federal award.” A recipient reading the guidance cannot determine on what grounds its award may be ended, because those grounds appear nowhere in it. “Changes in priorities” is the one ground the Guide does name, and it is defined neither in Guide 1’s key terms nor anywhere else in the 203-page draft.

[G25 / pp. 6, 9] Guide 25 defers to a proposed rule that is not final and that the Senate has voted to freeze by name.

NSF describes the revision as one that “[a]ligns termination provisions with the proposed revisions to 2 CFR 200” (p. 9), identifying those revisions as OMB’s proposed rule published May 29, 2026 (p. 6; 91 Fed. Reg. 32,198). Under that proposal, § 200.340(a)(2) would authorize termination where an award “does not effectuate program goals, Federal agency priorities, or the national interest as they exist at the time of the termination.”

A court has already held that the current rule does not permit this. In State of New Jersey v. U.S. Office of Management and Budget, No. 1:25-cv-11816 (D. Mass. July 17, 2026), Judge Talwani granted summary judgment and declared that 2 C.F.R. § 200.340(a)(4) (2024) and § 200.340(a)(2) (2021) “do not allow terminations of awards based on new program goals or agency priorities that an agency identifies after granting the award.” NSF is a named defendant in that action. The phrase “as they exist at the time of the termination” in the 2026 proposal appears designed to authorize by regulation what the court held the existing text does not permit, and Guide 25 would carry that language into NSF’s own guidance.

Separately, on August 8 the Senate passed the Continuing Appropriations Act, 2027 by a vote of 90 to 6. Section 157 provides that “[t]hrough December 11, 2026,” a rule arising out of the proposed rulemaking “Regulation for Federal Financial Assistance,” published May 29, 2026, or any substantially similar rule, “shall not be issued or finalized.” That is the rulemaking NSF cites on page 6.

Because Guide 25 states no grounds of its own, its meaning will shift with whatever OMB ultimately adopts. I cannot evaluate a provision defined entirely by reference to a rule that may yet be changed, enjoined, or never take effect at all.

[G25.Procedures 1] The immediate-termination exception is undefined and unbounded.

Guide 25 states that NSF “may immediately suspend or terminate an award when NSF believes such action is necessary to protect the interests of the government” (p. 198, ¶ 1(a)), and ¶ 1(e) confirms that in that event “a notice of non-compliance and corrective action plan will not be issued.” “Interests of the government” is undefined. The trigger is what NSF subjectively believes, and the Guide requires no written statement of reasons and names no approving official. An award could be ended with no notice and no chance to cure.

[G17.B.Policy 1 / G26.B.Policy 2] Incremental funding offers a route to the same result.

Most NSF research is supported through continuing grants funded in annual increments. Guide 17 states that “NSF may reduce continuing increments to respond to broad funding changes or new priorities” (p. 161), and Guide 26 puts those decisions outside its review process, which “does not apply to … decisions not to award a continuing grant or supplement” (p. 202). Revising Guide 25 alone would therefore leave a priority-based path to the same outcome, carrying less process than the one it replaces.

Requested actions

1. Strike “changes in priorities” from the list of common reasons for suspension or termination at page 197.

2. State the complete grounds for termination in the text of Guide 25, as 2 CFR 200.340(b) requires, rather than by cross-reference to a pending rulemaking.

3. Confirm in text that NSF will not terminate an award on the basis of program goals or agency priorities adopted after the award was made, consistent with New Jersey v. OMB.

4. Define “interests of the government”; replace “NSF believes” with an objective standard; require a written statement of reasons and a named approving official for any immediate termination; and limit the exception to fraud, imminent safety risk, or national security.

5. State expressly whether terminations premised on changed priorities are reviewable under Guide 26, and make priority-based reductions of continuing-grant increments reviewable.

6. Correct the citation at Guide 25 § A, Terms and Definitions ¶ 2, from 2 CFR 200.201 to 2 CFR 200.1.

Thank you for the opportunity to comment.

Respectfully submitted,

Marcus C. Sarofim, PhD

I have three go-to blogs for politics. Like Kate Marvel and Hannah Ritchie, these authors are logic- and data-driven, reliable in the areas where I have expertise, and very readable. They are some of the inspirations for my own Substack writing.

The Weekly Sift: Most Mondays, Doug Muder (an ex-mathematician and active Unitarian Universalist who started blogging in 2003) writes a post about news and politics. He usually does one deep dive (this week was a focus on El-Sayed, the Democratic Senate nominee in Michigan) and then a summary of the weekly news (this week he highlighted bad economic news and the Republican Fauci-hunt).

Electoral-Vote: This blog posts almost every day. It is primarily written by Andrew S. Tanenbaum (aka the Votemaster or “V”, a computer scientist) and Christopher Bates (aka John Peter Zenger or “Z”, a historian), who have been joined recently by Alison Regan (aka Lockwood or “L”, a lawyer) and Sari Kaufman (aka Ann Smith Franklin or “A”). Today they highlighted Iran War incompetence and the recent Kennedy Center news as another example of Trump’s infantile ego trips.

Letters from an American: This is another almost daily series from the historian Heather Cox Richardson. Today she highlighted some of the Trump administration’s ongoing horrible cruelty to immigrants as well as corrupt spending on private jets and unscientific statements on vaccines.

Thanks for reading. Now go file those comments — August 24th for NSF, August 29th for heavy-duty diesel.