Overview

Project 2025 (formally the 2025 Presidential Transition Project) is a $22 million initiative organized by the Heritage Foundation to provide a governing blueprint, personnel infrastructure, and implementation strategy for a conservative administration. Launched April 21, 2023, over 100 conservative organizations participated as coalition partners, and more than 400 contributors—the majority with ties to the first Trump administration—authored its proposals.

The document is unusual in American political history for two reasons. First, its scale: a 920-page operational blueprint published nearly two years before the election it targeted. Second, its self-consciousness: the plan does not merely propose policies but describes how to overcome anticipated internal government resistance to those policies, screens personnel for willingness to “test the limits” of law, and drafts Day-One executive orders in advance. That combination — advocacy plus pre-staged operational execution — is what places some portions of the document in a legally distinct category from ordinary think-tank output. This distinction is developed in the Truth and Reconciliation Framework section below.

Heritage Foundation and the Mandate for Leadership Series

The Heritage Foundation (founded 1973) has published the Mandate for Leadership series since 1981. The first edition guided the Reagan administration, which adopted two-thirds of its 2,000 recommendations. Subsequent editions (1985, 1989, 1996, 2001, 2005, 2015, 2018) had progressively narrower influence as Heritage’s function shifted from providing policy inputs to any Republican White House toward becoming an ideologically coordinating institution.

The ninth edition — Mandate for Leadership: The Conservative Promise — published April 2023 as Project 2025’s centerpiece. At 920 pages with 30 chapters spanning every federal department, it was the earliest pre-election Mandate edition, released months before the first Republican primary debate. The 2023 edition also differs from its predecessors in scope: previous Mandates concentrated on policy content, while the 2023 volume devotes substantial text to how the executive branch is to be reshaped, staffed, and defended against internal dissent.

Heritage president Kevin Roberts described the underlying doctrine as “institutionalizing Trumpism” — a formulation notable because it accepts that the movement’s continuity is a matter of institutional design rather than of any individual’s continued office-holding.

Key Personnel

Kevin Roberts — Heritage Foundation president since 2021. Declared the initiative a “second American Revolution, which will remain bloodless if the left allows it to be.” Roberts remained in place through the 2024 backlash and has continued to serve as Heritage’s public face for the initiative into 2026.

Russell Vought — Heritage senior fellow and author of the Mandate‘s executive-power chapter. Founder of the Center for Renewing America. Confirmed February 6, 2025 as Director of the Office of Management and Budget, a position from which he has coordinated impoundment doctrine, agency restructuring, and civil-service reclassification. In leaked comments in 2024, Vought described being “blessed” by Trump to develop a “second phase” of the project — a statement that materially complicated Trump’s campaign-season denials.

Paul Dans — Project director and Mandate co-editor (2023–2024). Former Chief of Staff at the Office of Personnel Management under Trump. Left the project in August 2024 amid political backlash, though he had previously stated on the record that Trump was “very bought in.”

Spencer Chretien — Associate Director of Presidential Personnel for the project. Former Special Assistant to the President in Trump’s first term. Directs the personnel database.

Steven Groves — Mandate co-editor. Margaret Thatcher Fellow at Heritage; former deputy White House press secretary and special assistant to the President.

John McEntee — Developer of the Presidential Personnel Office loyalty-screening questionnaire in Trump’s first term. Returned to a senior personnel role in the second term. His questionnaire architecture is the operational backbone of Pillar II.

The Four Pillars

Pillar I — Policy Guide: The 920-page Mandate with 30 chapters of policy recommendations, produced by 400+ scholars. The New York Times identified 307 authors/contributors; 182 had ties to Trump. CNN counted 140+ first-term Trump officials among participants.

Pillar II — Personnel Database: A system to recruit, vet, and catalog 20,000+ potential appointees emphasizing ideological alignment over credentials. Applicants screened on political philosophy; those reluctant to “test the limits” of presidential power were flagged and rejected. This is the pillar most legally consequential because it deliberately selects for personnel who have pre-committed to a posture of statutory and constitutional maximalism.

Pillar III — Presidential Administration Academy: Online and in-person training teaching recruits how government functions and how to function in government. Curriculum emphasizes bureaucratic combat: how to override career officials, how to structure orders for reduced judicial reviewability, and how to manage inspectors general and internal watchdogs.

Pillar IV — 180-Day Playbook: Unpublished agency-specific implementation plans — Day-One executive orders, personnel targets, and restructuring mechanisms. The withholding of Pillar IV from public disclosure is itself analytically significant: the visible 920-page Mandate is a policy argument, but Pillar IV is closer in character to an operational plan.

Schedule F / Schedule Policy/Career Civil Service Reclassification

The plan’s most structurally consequential proposal converts career civil servants in “policy-related” positions from competitive service (with due process protections) to at-will employment. Because the U.S. Constitution assigns Congress the power to structure the civil service and Congress in 1978 legislated Merit Systems Protection Board due process, the reclassification’s legal footing is contested.

Timeline:

• October 21, 2020: Trump signs Executive Order 13957 creating Schedule F. Never implemented before Biden rescinds it in January 2021.
• January 20, 2025: Trump signs an executive order titled “Restoring Accountability to Policy-Influencing Positions Within the Federal Workforce” directing reinstatement.
• February 2026: OPM issues a final rule implementing “Schedule Policy/Career,” renaming Schedule F and adding procedural adjustments intended to survive Administrative Procedure Act challenges.
• March 17, 2026: A federal court accepts a public-sector union’s proposal to file an amended complaint against the rule. Public Employees for Environmental Responsibility and four other unions file a separate suit in the U.S. District Court for the District of Maryland.
• June 3, 2026: Trump signs a further executive order operationally converting approximately 8,000 career employees (narrowed from an initial 50,000 estimate) — a scale-down that appears designed to make the initial conversion legally defensible while establishing the reclassification mechanism.

Consequences for converted employees. Individuals moved into Schedule Policy/Career lose notice-of-removal rights, appeal rights to the Merit Systems Protection Board, and — of particular relevance to accountability — independent whistleblower investigation channels. Litigation to date has focused on three theories: violation of the Civil Service Reform Act of 1978, violation of the Administrative Procedure Act’s arbitrary-and-capricious standard, and constitutional separation-of-powers claims. As of mid-2026 no ruling has definitively resolved the reclassification’s legality.

Why this matters for the framework in this document. The Schedule Policy/Career conversion is the mechanism by which the personnel screening in Pillar II becomes durable. Once employees are at-will, refusing an unlawful order becomes a fireable act. Statements in the Mandate that appear to anticipate this — for example, that policy-influencing personnel should share the president’s “willingness to test the limits” of legal constraints — take on evidentiary character in any future proceeding examining unlawful orders acted upon by such personnel.

Key Policy Proposals by Department

DOJ: Bring DOJ under direct presidential control, eliminate prosecutorial independence, target political opponents.

EPA: Roll back climate regulations, withdraw from Paris Accords, expand fossil fuel extraction on federal lands.

Education: Eliminate the department, end Title IX transgender protections, redirect funding to school choice.

HHS: Restrict abortion administratively, defund Planned Parenthood from Medicaid, direct NIH against gender-affirming care.

DHS: Expand expedited deportation, militarize border, penalize sanctuary cities, increase workplace raids.

DOD: Purge DEI programs, restructure civilian oversight, restore COVID-discharged service members.

Independent agencies. The Mandate proposes that “independent” regulatory agencies (FTC, FCC, SEC, NLRB) be treated as fully within presidential control, contrary to Humphrey’s Executor v. United States (1935) — a doctrine the Supreme Court has narrowed but not overturned. The unitary-executive theory undergirding this proposal is a matter of contested constitutional interpretation and, standing alone, is protected legal advocacy.

Trump’s Distancing vs. Adoption

During the 2024 campaign, Trump denied involvement: “I know nothing about Project 2025. I have no idea who is behind it.” His campaign issued memos stating the project “had nothing to do with the campaign.” These denials came after months of Democratic-campaign pressure that had made the project a defining issue.

After winning, Trump openly embraced the blueprint. Russell Vought (who authored the executive-power chapter) became OMB Director. John McEntee (who developed the loyalty-screening questionnaire) returned to a senior role. Paul Dans had previously stated Trump was “very bought in.” Vought’s own leaked “blessed by Trump” comments in 2024 established that the campaign-season denials were themselves calculated messaging rather than accurate representations of the plan’s status.

The gap between campaign denial and post-inauguration implementation is itself relevant to the framework below: statements of disavowal made for electoral purposes do not immunize subsequently implemented content from either legal analysis or historical accountability.

Implementation (2025–2026)

By February 2026, independent tracking by the Center for Progressive Reform and Governing for Impact documented 53 percent of Project 2025’s domestic administrative agenda initiated or completed — 283 of 532 recommended actions across 20 federal agencies. Coverage in PBS NewsHour and the Alabama Reflector placed the year-one implementation figure similar in magnitude and identified acceleration in Q4 2025 and Q1 2026 as the Schedule Policy/Career mechanism began to take effect.

POLITICO identified 37 direct correlations between executive orders and Project 2025, including “nearly verbatim language.” The Center for Reproductive Rights found 85 percent of reproductive health actions stemmed from Project 2025 recommendations. Fullerton Observer and other outlets published side-by-side comparisons of executive-order text and Mandate chapters.

Discrete high-profile implementations:

• Withdrawal from the Paris Climate Accords (Day One, January 2025)
• Government-wide DEI elimination and dismissal of DEI-designated career staff
• OPM final rule implementing Schedule Policy/Career (February 2026)
• Executive order operationally converting ~8,000 employees (June 2026)
• Expedited-removal expansion under DHS
• Impoundment-doctrine assertions by OMB over congressionally appropriated funds
• Consolidation of authority over previously independent regulatory agencies
• Elimination or restructuring of inspector-general offices at multiple departments

Categories where implementation has been slower or has been blocked in litigation: abolition of the Department of Education (requires legislation, not achieved), full Title IX regulatory rollback (partial, subject to injunctions), certain immigration-enforcement expansions (blocked by district courts).

The Second Phase

In 2024 comments not initially intended for public release, Russell Vought described a “second phase” of the initiative — continued architectural work extending beyond the Mandate‘s original scope. This second phase has become visible in 2026 through several developments:

• Additional Heritage-adjacent policy volumes published in early 2026 concentrating on state-level implementation and on federal-state coordination for policies unable to survive court challenges at the federal level
• Expansion of the Presidential Administration Academy curriculum to include instruction on defending reclassification-era personnel decisions
• Coordination with the American Legislative Exchange Council (ALEC) on state-level statutory implementations of federal deregulatory positions

The second-phase framing matters for accountability analysis because it establishes that the project is a continuing coordinated effort rather than a one-time transition planning exercise. Continuing conduct has different characteristics under statutes of limitation and under conspiracy law than a discrete completed act.

Coalition and Funding

110+ organizations sat on the advisory board, including: America First Legal (Stephen Miller), Center for Immigration Studies, Moms for Liberty, Turning Point USA, Claremont Institute, and ALEC. Nearly half received funding from dark-money networks connected to Leonard Leo. Total documented funding for Project 2025 itself exceeds $22 million; the surrounding coalition ecosystem represents multiples of that figure across all coalition-partner organizations.

Funding structure is relevant to accountability analysis because 501(c)(3) tax status prohibits certain political activities. The line between educational activity (permissible under 501(c)(3)) and political campaign activity (prohibited) is one of the enforcement levers that follows specific factual findings about how the project’s outputs were used.

Public Backlash

Project 2025 became a defining 2024 campaign issue. Democrats used it to argue that Trump’s second term would be authoritarian. Constitutional scholars compared its civil-service proposals to authoritarian consolidation in Hungary and Brazil. Public polling in mid-2024 showed the project’s brand name became a net-negative political asset — the reason Trump’s campaign publicly distanced itself even as senior campaign figures continued participating.

Truth and Reconciliation Framework: Protected Speech vs. Pre-Admission of Conduct

The remainder of this document proposes a framework for evaluating Project 2025’s contents. The framework is intended to be neutral, legally grounded, and applicable to any comparable policy blueprint regardless of ideological direction. The same framework applied to a hypothetical left-wing operational document proposing analogous administrative-state restructuring would produce analogous outputs.

The core question the framework answers: When does a written policy blueprint cross the line from constitutionally protected political speech into evidence of pre-admission of conduct that may violate law?

1. The First Amendment Baseline

American constitutional doctrine strongly protects political advocacy, including advocacy for radical restructuring of government. Three doctrines set the floor:

Brandenburg v. Ohio (1969). Advocacy of law-breaking is protected unless it is (a) directed to producing imminent lawless action and (b) likely to produce such action. A book-length policy document proposing future administrative actions almost never satisfies both prongs.

NAACP v. Claiborne Hardware Co. (1982). Vigorous political advocacy, including advocacy that offends or that contemplates aggressive tactics, is protected. Advocacy is not liable for the subsequent unlawful acts of persons who acted after hearing the advocacy, absent specific direction of unlawful acts.

Snyder v. Phelps (2011). Political speech on public issues receives the highest level of protection even when its content is offensive to a majority of listeners.

**Rule 1: Publishing a policy blueprint — even a radical one, even one proposing dismantling of federal agencies or expansion of presidential power — is quintessential protected speech. Heritage’s constitutional right to publish the Mandate is not seriously in question.**

2. Where Speech Becomes Evidence, Not the Offense

The First Amendment does not bar the use of speech as evidence in prosecutions or accountability proceedings for conduct. It bars punishing speech as such. Two doctrines matter here:

Wisconsin v. Mitchell (1993). A defendant’s speech is admissible to prove intent or motive. A racist statement can be evidence of a bias-motivated crime; a written plan can be evidence of the intent behind a subsequent act.

Federal conspiracy law (18 U.S.C. § 371). Requires an agreement plus at least one overt act. Written documents can constitute or evidence the agreement. Conspiracy prosecutions routinely admit into evidence policy memoranda, planning documents, and internal writings that would, standing alone, be protected speech.

**Rule 2: The First Amendment protects the speaker from being prosecuted for the speech. It does not immunize the speech from being used as evidence in prosecutions for conduct that follows.**

The distinction is subtle but decisive. The Mandate itself cannot be the crime. But specific statements within the Mandate can be admissible evidence of intent, plan, or agreement when combined with subsequent unlawful conduct by named implementers.

3. The Five-Tier Triage Matrix

The framework classifies statements within a policy blueprint into five tiers based on the character of the statement and its relationship to subsequent conduct.

Tier 1 — Protected Advocacy Only. Character: Argument that a policy should be adopted, an agency abolished, a treaty exited, a statute repealed. Directed at persuading the electorate or Congress. Examples in Project 2025: Advocacy of abolishing the Department of Education; advocacy of withdrawing from the Paris Accords. Legal status: Fully protected. Not admissible as evidence of anything but the speaker’s viewpoint. No accountability lever exists or should exist for content in this tier, regardless of how much a reader disagrees with it.

Tier 2 — Contested Legal Theory. Character: Advocacy that a lawful action is more expansive than currently practiced, based on a contested reading of statute or constitution. Examples: Unitary-executive advocacy; impoundment doctrine; the theory that “independent” agencies are constitutionally accountable to the president. Legal status: Protected. Can constitute persuasive authority. Courts, not political opponents, are the appropriate mechanism for resolving whether the underlying legal theory prevails. If a court subsequently rules the theory incorrect, the statement of the theory remains protected; the action taken in reliance on it becomes subject to challenge.

Tier 3 — Loyalty and Selection Evidence. Character: Statements about the criteria to be used for selecting personnel, particularly criteria that select for willingness to override normal constraints. Examples: Screening applicants for willingness to “test the limits” of presidential power; rejecting candidates on the basis of past adherence to institutional norms rather than on qualifications. Legal status: In the abstract, protected as recruitment criteria. Admissible as evidence of shared intent if plans, selection criteria, and subsequent conduct converge on unlawful acts. The relevant chain of custody is: (i) a criterion published in a blueprint, (ii) a named appointee who was selected against that criterion, (iii) subsequent conduct by that appointee matching the plan.

Tier 4 — Anticipatory Admissions. Character: Language that presupposes knowledge that the planned conduct will require overriding existing legal or constitutional constraints. The speaker acknowledges that what is being planned will require exceeding current legal authority. Examples: Personnel screening for willingness to defy anticipated court orders; language framing the reform effort as a “revolution” whose success depends on overcoming institutional resistance likely to be judicially validated. Legal status: Speech itself remains protected. However, the language constitutes strong evidence of mens rea — knowledge — when the constrained conduct is later performed. In a subsequent prosecution or civil proceeding, such statements are admissible to defeat claims that the actor believed the conduct was lawful.

Tier 5 — Operational Playbook Awaiting Illegal Execution. Character: Specific operational instructions, particularly Pillar-IV-style Day-One executive orders and 180-day agency plans, that are subsequently executed by named implementers in a manner that a court finds violates statute, treaty, or constitutional constraint. Legal status: The blueprint text itself remains protected as pure speech. However, once execution occurs and unlawfulness is established, the blueprint becomes evidence of a coordinated plan. This is the tier at which the combination of the blueprint plus the coordinated execution can constitute a conspiracy under 18 U.S.C. § 371 — the blueprint being written evidence of the agreement, and each executed action being an overt act in furtherance.

Triage rule. A statement can only be placed in Tier 4 or Tier 5 based on facts external to the statement itself. Reading a blueprint in isolation cannot place any statement above Tier 2. Placement in Tier 3, 4, or 5 requires evidence about subsequent conduct, chain of custody to named implementers, and a court or independent fact-finder’s determination that the executed conduct violated law.

4. Actor Categories and Calibrated Accountability

Not everyone associated with a large-scale policy project bears the same relationship to any subsequent unlawful conduct. The framework distinguishes:

Category A — Blueprint Authors. Individuals who wrote or edited Mandate chapters. Bear the highest culpability for the content of the writing but the lowest culpability for its execution, absent evidence they also participated in executing it. Russell Vought is a rare individual in both roles.

Category B — Coalition Advisory Board Members. Institutions that lent their names and resources but did not write. Culpability depends on the specific factual role each played and on whether their organizational participation involved 501(c)(3)-prohibited political activity.

Category C — Confirmed Appointees Selected Under Project 2025 Criteria. Individuals selected through the Pillar-II database. Bear culpability calibrated to (i) their awareness of the criteria under which they were selected and (ii) the specific conduct they subsequently performed. Being selected under Project 2025 is not itself grounds for accountability; being selected under it and then performing conduct that a court finds unlawful is.

Category D — Rank-and-File Executors. Career staff or new hires who executed instructions without policy-making authority. Culpability generally limited to circumstances where the unlawfulness of an order was clear on its face (the Nuremberg standard as adapted in U.S. law: an order is not a defense if a reasonable person would recognize its unlawfulness).

Category E — Congressional Acquiescents. Members of Congress who declined oversight, or who voted to confirm nominees selected under criteria that were publicly known. Political accountability at the ballot box; generally not legal accountability.

5. Standards of Proof

Different accountability tracks require different standards:

• Political accountability (elections, censure): No standard beyond persuading the electorate or a legislative body.
• Civil accountability (removal from office by impeachment, private civil suits, injunctive relief): Preponderance of the evidence or clear-and-convincing evidence depending on the forum.
• Administrative accountability (bar from federal contracts, revocation of 501(c)(3) status, professional discipline): Substantial evidence on the administrative record.
• Criminal accountability (conspiracy, deprivation of rights under color of law, obstruction): Beyond a reasonable doubt.

Most Project 2025-related accountability, if it occurs, will be political, civil, and administrative — not criminal. Criminal accountability requires meeting the very high bar of Tier 5, with the additional showing of individual intent by a specific defendant.

6. Institutional vs. Individual Accountability

Post-authoritarian transitional-justice practice from South Africa, Chile, Argentina, Spain, Peru, and Germany distinguishes two accountability tracks:

Institutional consequences. Organizations that participated in unlawful conduct may lose tax-exempt status, be barred from federal contracting, or face restructuring. These are administrative determinations that do not require criminal findings against individuals. Applied to Project 2025, the relevant questions are whether Heritage’s 501(c)(3) activities included prohibited political-campaign activity, and whether coalition partners similarly stepped over prohibited lines.

Individual consequences. Prosecution for the most serious crimes; lustration (administrative bars on holding certain offices) for lesser conduct; truth-telling requirements as a precondition for continued participation in public life. Individual criminal prosecution is generally reserved by transitional-justice practice for the gravest offenses, with lustration handling the broader accountability need.

The distinction matters because it separates accountability from political vengeance. Lustration is applied to categories of conduct (having selected personnel under a defined loyalty criterion, for example) rather than to persons per se. It is time-limited, forward-looking, and does not carry criminal stigma.

7. The Reconciliation Offer

Truth commissions offer a structured reduction of sanction in exchange for full, verifiable disclosure. Applied prospectively to a Project 2025-style initiative, the reconciliation offer has three elements:

1. Full disclosure of authorship, methods, and intent. Individuals who voluntarily provide sworn testimony on their roles, the operational details of Pillars II and IV, and the internal reasoning for specific proposals qualify for the reduction.
2. Verifiable specificity. The disclosure must be sufficient for independent verification. General statements of participation are insufficient; specific accounts of who directed what, when, and why are required.
3. Reduced sanction, not immunity. Criminal immunity for the most serious conduct is not on offer; reduced lustration periods, avoidance of tax-status penalties for cooperating institutions, and retention of professional licensure are. The floor is that no truth-teller ends up worse than a non-cooperating actor at the same conduct level.

The reconciliation track is a compliance-forcing mechanism, not a moral offer. It creates asymmetric incentives: the first coalition partner or blueprint author to cooperate secures the best terms; late cooperators, less; non-cooperators face the full penalty within lawful limits.

8. Applying the Framework: Illustrative Placements

The following are illustrative applications of the framework. Actual placements would require the full evidentiary record, which does not yet exist for most items.

• Advocacy of Paris Accord withdrawal → Tier 1. Policy dispute, protected.
• Unitary-executive theory as applied to independent agencies → Tier 2. Contested legal theory. Courts adjudicate.
• Personnel screening for willingness to “test the limits” of presidential power → Tier 3, elevating to Tier 4 where the same criteria are combined with statements anticipating specific court orders will be defied.
• Schedule Policy/Career reclassification as policy proposal → Tier 2. Reclassification-then-firing of specific whistleblowers for reporting unlawful conduct → Tier 5 conduct by the specific actors.
• Withholding of Pillar IV operational plans from public disclosure while representing that Pillars I–III were the totality → Tier 4 anticipatory admission that the visible portion of the project understated its operational content.

9. What This Framework Refuses to Do

• It does not treat unpopular ideas as illegal.
• It does not treat association with the project as culpability.
• It does not treat Mandate content as evidence of anything on its own; blueprint text is evidence only in combination with subsequent conduct.
• It does not create new offenses; every accountability track above rests on existing statutes, existing constitutional doctrine, and existing transitional-justice practice.
• It does not distinguish based on the ideological direction of the underlying project. A comparable left-wing operational document would receive analogous treatment.

The framework’s design goal is that reasonable observers across the ideological spectrum, applying it in good faith, would place any given statement in the same tier — because tier placement is driven by structural features (character of the statement, evidence of subsequent conduct, chain of custody to named implementers) rather than by the reader’s agreement or disagreement with the underlying policy.

Sources

• Heritage Foundation, Mandate for Leadership: The Conservative Promise (April 2023)
• Center for Progressive Reform and Governing for Impact, Project 2025 Executive Action Tracker (February 2026)
• PBS NewsHour, “Tracking how much of Project 2025 the Trump administration achieved this year” (2025–2026)
• Congressional Research Service, Schedule Policy/Career: 2026 Final Rule, Legal Challenges, and Issues for Lawmakers (LSB11412)
• Government Executive, “Trump signs order moving thousands into Schedule F” (June 2026)
• The White House, Presidential Action on Implementing Schedule Policy/Career in the Excepted Service (June 2026)
• Economic Policy Institute, OPM Final Rule Analysis
• POLITICO, “37 ways Project 2025 has shown up in Trump’s executive orders” (2025)
• ProPublica, What You Should Know About Russ Vought, Trump’s Shadow President
• Forbes, Russell Vought Confirmation and Trump Officials with Project 2025 Ties (February 2025)
• Democrats.org, Vought “Second Phase” and “Blessed by Trump” Comments
• Brandenburg v. Ohio, 395 U.S. 444 (1969)
• NAACP v. Claiborne Hardware Co., 458 U.S. 886 (1982)
• Wisconsin v. Mitchell, 508 U.S. 476 (1993)
• 18 U.S.C. § 371 (conspiracy)
• South Africa Truth and Reconciliation Commission Final Report (1998), for the reconciliation-offer structure adapted in Section 7