THE RECORD
Every entry uses a three-layer format — The Facts the documented claim; Significance why it may matter in context; The Goalpost the strongest available defense or normalization frame. The layers are labeled so readers can inspect their different evidence burdens. Sequence can reveal patterns that a single headline cannot, but this mixed-review archive does not claim complete coverage or a gap-free account.
Support This ProjectFollow the evidence, then inspect the method.
Start broad, narrow to a person or public statement, keep source quality visible, and reverify every time-sensitive claim after the named cutoff.
📝 Final Seven Days in This Snapshot
Loading...
Who appears in the record?
Search officeholders, officials, judges, advisers, and political figures, then inspect the records and citations associated with each name.
Roster sources: congress-legislators · gubernatorial biographies · state.gov · treasury.gov. Record citations are graded separately inside each dossier.
POWER BECOMES POLICY HERE.
A presidential statement is not the end of a story. Follow it through the institution that writes the rule, moves the money, changes the staff, opens the case, deploys the force, or carries the order into public life.
THE TWITTER ERA
2009—2021 · before Truth Social
Twitter was not merely a place Donald Trump posted. It became campaign stage, presidential briefing room, pressure channel, grievance engine, and a minute-by-minute public artifact of power.
AI Opinion
These are labeled opinion essays, not neutral summaries. Five were originally written by Claude Opus 4 in April 2026 from a 4,250-entry snapshot and received a documented factual and epistemic review by ChatGPT/Codex Sol on July 17, 2026. The Truth Social essay is a separately preserved corpus-only run. Original authorship, review history, archive cutoff, and limitations remain visible on every essay.
Search The Record
Ask questions about the archive using a custom AI assistant trained on The Record's full dataset. The assistant can help you find entries, understand patterns across eras, compare how talking points evolved, and explore connections between events.
Opens in a new window. Powered by a custom GPT trained on The Record's complete master document.
Example questions you can ask:
"How did the 'fake news' talking point evolve from 2016 to 2025?"
"Show me every time Trump attacked the judiciary"
"What happened between Trump and Epstein in the 1990s?"
"How did Republicans go from 'Never Trump' to full loyalty?"
"Compare the family separation policy to the Jan 6 pardons"
"What did John Kelly say about Trump after leaving?"
By The Numbers
Term-by-term comparison · sourced where possible · honest about what we don't know
How to read this page. Every number below is labeled with a status badge. VERIFIED means the number is on the public record (court ruling, official disclosure, agency report, congressional record). ESTIMATE means it's a reputable third-party guess (Forbes, Bloomberg, Tax Foundation, CBO) — the actual number may differ. PROJECTION means it's a forecast for an ongoing term that hasn't finished. Where presidents inherit conditions (economy, debt, unemployment), we say so. We don't pretend a president caused a number just because it happened on their watch. We don't hide the numbers Trump's supporters point to either.
💰 Trump's Personal Wealth — What We Can Estimate
Trump has not released full tax returns since the 1970s. Every wealth number below is a Forbes estimate based on disclosed assets, valuation methods Trump himself has called "low" in some years and "fair" in others, and public filings. Real number: unknowable.
| Date | Forbes Net Worth Estimate | Status | Context |
|---|---|---|---|
| June 2015 (announces run) | ~$4.5B (Trump claimed $10B) | ESTIMATE | Forbes vs. Trump's own number diverged by ~$5.5B |
| Jan 2017 (took office, T1) | ~$3.5B | ESTIMATE | Drop attributed to soft NYC commercial real estate |
| Jan 2021 (left office) | ~$2.4B | ESTIMATE | COVID hit hotels, DC hotel sold, brand value dropped |
| March 2024 (DJT IPO) | ~$7.5B peak (briefly) | ESTIMATE | Truth Social (DJT) merger — paper wealth, lockup restricted |
| Jan 2025 (took office, T2) | ~$5.5B | ESTIMATE | DJT stock cooled; lockup expired October 2024 |
| May 2026 (at essay cutoff) | ~$6–8B range | ESTIMATE | Range reflects $TRUMP memecoin, World Liberty Financial, Gulf state deals — most not yet booked at fair value |
The Marianne caveat: Trump came in with somewhere around $3.5B (Forbes) or $10B (Trump's own claim). He's now estimated at $6-8B. Whether that means he made money in office, gained from external markets, or simply re-valued opaque assets is impossible to say without the tax returns he refuses to release. The DJT stock pump and the crypto launches happened during the political career. Every estimate above could be off by a billion in either direction — Forbes admits this. Take all "billionaire" numbers with that grain of salt.
📊 Term-by-Term Comparison: Trump T1 vs. Biden vs. Trump T2
Where the numbers are clean (counts of orders, pardons, etc.), they're verified. Where presidents inherit conditions (economy, deficit), we mark it. Trump T2 numbers are partial (term not yet complete) and labeled as projections.
| Metric | Obama (8 yr) | Trump T1 (4 yr) | Biden (4 yr) | Trump T2 (so far) |
|---|---|---|---|---|
| Executive Orders VERIFIED | 276 total (~35/yr) | 220 total (~55/yr) | 162 total (~40/yr) | ~210+ in ~16 months (~157/yr pace) |
| Pardons + Commutations VERIFIED | 1,927 (mostly drug offenses) | 237 (many connected to Trump or family) | ~4,000 (included preemptive family/staff) | ~1,500+ (Jan 6 mass pardon Day 1) |
| Cabinet Turnover (A-team, Brookings methodology) VERIFIED | ~71% over 8 yrs (~9%/yr) | 92% (highest in modern history) | ~62% (stable) | Tracking Hegseth Signal-gate, Ramaswamy DOGE exit, others |
| Days at Privately-Owned Properties VERIFIED | N/A (none owned) | ~430+ at Trump properties (taxpayer-funded Secret Service stays charged to Trump Org) | ~250 at Delaware/Camp David (no private revenue) | ~190+ at Mar-a-Lago / Bedminster (pace exceeds T1) |
| National Debt Added VERIFIED (presidents inherit fiscal trajectories) | +$8.6T over 8 yrs (rescued from 2008 collapse) | +$7.8T over 4 yrs (2017 tax cut + COVID spending) | +$8.5T over 4 yrs (COVID rebuild, IRA, CHIPS, infrastructure — debt $27.7T → $36.2T) | +~$2T projected first year (tax cut extension + tariff offsets) |
| Federal Workforce Cut VERIFIED | +0.5% growth | +2.8% growth (net add despite hiring freeze rhetoric) | +5.3% growth | -200,000+ jobs cut (DOGE — largest cut since post-WWII demobilization) |
| Tariffs Imposed VERIFIED | Minimal new | ~$80B cumulative (2018-19) (China 301, steel/aluminum 232) | Kept most T1 tariffs | ~$340B/yr projected (global, broad-based — Tax Foundation est.) |
| Tariff Cost Per Household ESTIMATE (Tax Foundation / Yale Budget Lab) | N/A | ~$200/yr | ~$200/yr | ~$600–1,800/yr (varies by income — bottom 80% pay ~$400–700, top 20% pay ~$1,000–1,800; Yale Budget Lab 4/2026) |
| Deportations VERIFIED | ~3M total over 8 yrs ("deporter in chief") | ~935K total (family separation policy) | ~1.5M total | ~620K in 16 months (plus high-profile El Salvador transfers) |
| Approval Rating Avg. VERIFIED (Gallup / Silver Bulletin) | 47.9% | 41% (lowest avg post-Truman) | 42.2% | ~37–39% approve / 58–62% disapprove (Silver Bulletin -18.9 net, 2nd-term low; NPR/Marist 5/6: 37/59; Reuters/Ipsos 4/28: 34%; -38 net among independents) |
📈 Economy: What Each President Inherited vs. What They Left
Presidents inherit economic conditions — they don't cause numbers on Day 1. This table shows the change during each term, not the absolute numbers.
| Indicator | Obama 8 yr Δ | Trump T1 Δ | Biden Δ | Trump T2 (so far) |
|---|---|---|---|---|
| S&P 500 Total Return | +182% (off 2009 lows) | +67% (includes COVID crash) | +58% | ~-3% YTD (tariff volatility) |
| Unemployment | 10.0% → 4.7% | 4.7% → 6.4% (COVID) | 6.4% → 4.0% | 4.0% → 4.3% (April 2026; initial jobless claims 211K week of 5/9; 37K+ layoffs first 10 days of May incl. Meta 8K, PayPal 4.7K) |
| Inflation (CPI YoY) | avg ~1.7% | avg ~1.9% (low until 2020) | peaked 9.1% (June 2022) (post-COVID surge, ended ~2.9%) | 3.8% (April 2026 print, BLS 5/12) highest since May 2023; energy +3.8% MoM, war-driven · core 2.8% · PPI 6.0% YoY (highest since Dec 2022) |
| Manufacturing Jobs | +899K | -178K (includes COVID) | +800K (IRA / CHIPS) | -90K so far (early tariff pass-through) |
⚡ Wartime Economy Snapshot — Week of May 9–16, 2026
The Iran war began February 28, 2026. These are the data points the week of May 9–16 made unavoidable. All VERIFIED against primary sources (BLS, AAA/EIA, Federal Reserve, BEA, U Michigan Survey of Consumers, exchange settlement data).
| Indicator | Pre-war (Feb 27, 2026) | May 9 (last update) | May 30 snapshot | Source |
|---|---|---|---|---|
| Gas (AAA national avg) | $2.94/gal | $3.45/gal | $4.628/gal (+57% from pre-war) | AAA / EIA |
| State records | — | — | WA all-time record $5.77; CA $6.15–6.16; HI $5.64 | AAA state |
| WTI crude | ~$72 | ~$96 | $106 (+11% on the week) | CME |
| Brent crude | ~$76 | ~$101 | $109 (+8.1% on the week) | ICE |
| CPI YoY | 2.4% | 2.4% (March print) | 3.8% (April print, released 5/12 — highest since May 2023) | BLS |
| PPI YoY | ~2.5% | — | 6.0% (April print, released 5/13 — highest since Dec 2022) | BLS |
| UMich Consumer Sentiment | ~74 | 49.8 | 48.2 preliminary May — lowest in 75-year survey history | U Michigan SCA |
| S&P 500 | ~6,150 | ~7,500 | 7,408.50 (Fri close, -1.24%) | CBOE |
| Dow Jones | ~43,400 | ~50,000 | 49,526 (Fri close, -537 pts, lost 50K handle) | NYSE |
| Nasdaq | ~19,200 | ~26,600 | 26,225 (Fri close, -1.54%) | NASDAQ |
| Fed Chair | Powell | Powell | Kevin Warsh confirmed 54-45 (5/13, closest modern Fed-chair vote); took over 5/15 | Senate roll call |
| 30-yr fixed mortgage | 6.18% | ~6.30% | 6.36% (5/14) | Freddie Mac PMMS |
| Iran war direct cost | — | ~$24B | ~$29B (Hegseth disclosure, 5/12) | DoD testimony |
| Hormuz vessel traffic | ~140/day | — | ~30 since Wednesday (5/14) | IRGC reporting |
| Trump disapproval — gas prices | — | 63% | 66% | PBS/Yahoo/Marist |
| Trump disapproval — Iran handling | — | 54% (March) | 60%; GOP disapproval climbed 15% → 22% | cross-tab averages |
| Generic ballot | D+1 | D+3 | D+5 to D+6 consensus (Morning Consult D+3, FiftyPlusOne D+5.4, Fox D+6) | tracker average |
How to read this block. Every column above is a number, not an interpretation. The interpretation — that wartime energy prices are bleeding through to core inflation, that consumer sentiment has reached a 75-year low, that the Fed chair was replaced by a 54-45 margin in the middle of a hot CPI print, and that the GOP base is showing its first crack on Iran — comes from reading the columns in sequence. The administration's position is that the costs are temporary and the trade-off is worth it. The columns are the receipt either way.
💸 The Trump Self-Dealing Ledger (Term 2)
Money flowing to Trump and his family while in office. These are mostly verified — they're public filings, public crypto wallets, or public Trump Org disclosures. The estimates exist where wallet ownership is inferred but not directly disclosed.
- $TRUMP Memecoin (launched Jan 17, 2025, 3 days before inauguration) VERIFIED
~80% of supply held by Trump-affiliated entities at launch. Trading fees paid to wallet controlled by CIC Digital LLC (Trump Org). Estimated take from launch + ongoing fees: ~$350M+ in disclosed receipts (per March 2026 financial disclosure). - World Liberty Financial (WLF) Token VERIFIED
Trump family entities hold ~75% of governance tokens. Foreign buyers (UAE-linked entities, others) bought roughly $400M+ in tokens during 2025. Sale proceeds flow to Trump-affiliated wallets per WLF terms. - Trump Media (DJT) Stock VERIFIED
Trump holds ~115M shares (~57% of company). Stock value at any moment ranges $3B–$5B paper. He has not sold (filings). - Gulf State Real Estate Deals VERIFIED
During the May 2025 trip: Trump Tower Jeddah announced (~$533M, partnership with Dar Global), separate $1B+ Saudi project announced, Trump Tower Dubai expansion, Saudi-funded LIV Golf hosting fees at Trump properties. Trump Org licensing deals, not direct sales — but cash flows during his presidency. - Mar-a-Lago Membership Fees VERIFIED
Initiation fee was raised to $1M (Jan 2025) up from $700K. Membership cap raised to ~750. Foreign nationals among new members. - Truth Social Ad Revenue + Subscription ESTIMATE
Modest by tech standards (~$10M/quarter). The asset value is the stock, not the ad business.
⚖️ Legal Judgments Owed (or in Appeal)
- E. Jean Carroll I (sexual abuse + defamation): $5M VERIFIED
- E. Jean Carroll II (defamation): $83.3M VERIFIED
- NY Civil Fraud Judgment (Engoron ruling, business valuation fraud): $454M (in appeal, accruing interest) VERIFIED
- NY Hush Money Criminal Conviction (34 felonies): Unconditional discharge — no fine, no jail, but felony conviction stands VERIFIED
- Federal Cases (DOJ January 6 + Mar-a-Lago documents): Dismissed by DOJ Jan 2025 after election win, per Justice Department policy on sitting presidents VERIFIED
- Georgia RICO Case: Stalled — DA Fani Willis disqualification appeal ongoing, 2026 status uncertain VERIFIED
- Total Confirmed Civil Liability: ~$542M (with interest accrual) VERIFIED
❓ What We Can't Score Honestly
True net worth. Without tax returns, every "billionaire" number is a third-party guess based on disclosed assets and valuations Trump himself has called both "low" and "fair" depending on the audience.
Dark money inflows. Donations to Trump-aligned PACs and super-PACs from undisclosed donors. We see the spending. We rarely see the source.
Foreign payments routed through hotels and licensing. Trump Org received foreign-government bookings during T1 (the lawsuit Congress lost on standing was over this). T2 disclosures are partial.
Crypto wallet flows. Public chains let us see addresses we believe to be Trump-controlled. Without verified self-custody disclosure, every dollar attributed could be off.
Family wealth transfers. Don Jr., Eric, Ivanka, Jared all receive licensing income, board fees, business arrangements. The family balance sheet is not the candidate's balance sheet — but the family is the candidate's primary beneficiary structure.
Counterfactuals. We can't say "the economy would have been X under President Y." We report what happened during whose watch, mark what was inherited, and let the reader draw causation if they want to.
🏛️ Cabinet & Senior Staff Turnover (Term 2)
Brookings tracker pegs Term 2 "A-team" turnover at ~20% one year in — already exceeding Biden's full four-year pace. Withdrawn nominations + forced resignations + outright firings combined. Cabinet alone is a narrower count; Brookings counts senior staff. VERIFIED for individual events.
| Date | Name / Position | Reason |
|---|---|---|
| Mar 27, 2025 | Elise Stefanik (UN Ambassador nominee) | Withdrawn — House margin too thin |
| Mar 2025 | Dave Weldon (CDC Director nominee) | Withdrawn — Senate vaccine concerns |
| May 2025 | Jared Isaacman (NASA Administrator nominee) | Withdrawn |
| May 1, 2025 | Mike Waltz (National Security Advisor) | Removed after "Signalgate"; renominated UN Ambassador Sept 2025 |
| May 2025 | Elon Musk (DOGE Special Government Employee) | Exited; statutory 130-day SGE limit reached |
| May 7, 2025 | Janette Nesheiwat (Surgeon General nominee) | Withdrawn — credential questions |
| Mar 5, 2026 | Kristi Noem (Secretary of Homeland Security) | Removed |
| 2026 | Pam Bondi (Attorney General) | Removed — Epstein file release fallout |
| Apr 2026 | Lori Chavez-DeRemer (Secretary of Labor) | Resigned |
| Apr 30, 2026 | Casey Means (Surgeon General nominee) | Withdrawn — replaced by Nicole Saphier |
| May 12, 2026 | Marty Makary (FDA Commissioner) | Resigned — vapes + mifepristone friction |
| May 15, 2026 | Jerome Powell (Federal Reserve Chair) | Replaced by Kevin Warsh (confirmed 54-45, narrowest modern Fed-chair vote) |
Sources: Brookings turnover tracker · U.S. News tracker
⚖️ Federal Litigation Against the Administration (Term 2)
More lawsuits in 16 months than any modern administration. Trackers count differently: Just Security counts every filing, Lawfare consolidates appeals. Democratic AGs alone filed their 100th lawsuit in April 2026.
- 753 cases tracked by Just Security as of April 9, 2026 ESTIMATE
- 316 active cases per Lawfare (consolidated appeals count) ESTIMATE
- 17 SCOTUS emergency stays in administration's favor; 2 SCOTUS affirmations against admin; 10 summary judgments for government, 10 against (early-term snapshot)
- 100th Democratic-AG-coalition lawsuit filed April 2026
Highest-stakes 2026 decisions:
- SCOTUS, Louisiana v. Callais (Apr 29, 2026) — 6-3 narrowing Voting Rights Act §2; Alito majority, Kagan dissent. Cascaded into FL/TN/AL/SC mid-decade redistricting within 2 weeks.
- 6th Circuit, ICE mandatory detention (May 11, 2026) — 2-1, struck down as unconstitutional. Third circuit to do so; four-way split now teed up for SCOTUS.
- Court of International Trade → Federal Circuit, Section 122 tariffs (May 7 → May 12, 2026) — CIT held tariffs unlawful; CAFC granted DOJ emergency stay; tariffs remain in effect for non-party importers.
- SCOTUS, Louisiana v. FDA / mifepristone telehealth (May 14, 2026) — temporary stay preserves mail and telehealth access; Thomas + Alito dissented invoking the 1873 Comstock Act.
Sources: Just Security tracker · Lawfare tracker · Ballotpedia SCOTUS emergency orders
👨⚖️ Federal Judges Confirmed (Term 2)
37 Article III judges confirmed through May 1, 2026: 0 Supreme Court (the three Term 1 picks remain), ~5 Circuit Courts of Appeals, ~32 District Courts. 19 nominations pending in early May 2026 (5 circuit, 13 district, 1 Court of International Trade). Cumulative Trump-appointed lifetime judges across both terms: ~271 (3 SCOTUS + 60 circuit + 205 district + 3 CIT) — Bloomberg Law notes the pace is slowing as vacancies are scarce after first-term reshaping.
Sources: Ballotpedia (May 7, 2026) · Wikipedia cumulative list
📱 Truth Social Volume & AI Imagery (Term 2)
NPR's May 8, 2026 analysis: 2,249 posts in the first four months of 2026 alone — averaging ~19 posts per day. Cumulative since Jan 20, 2025: ~8,800+ posts ESTIMATE (extrapolated from NPR's 4-month figure plus 2025 baseline; not independently audited).
- 93 posts of 1,500+ characters in Jan–Apr 2026; roughly half were all-caps endorsement screeds
- 16 posts in 90 minutes on May 2, 2026 (9 AI-generated images including Pritzker fat-shame + Iran drones "dropping like butterflies")
- 55 posts in 3 hours the night of May 11–12, 2026 — one every ~3 minutes — including AI Obama/Biden/Pelosi-in-sewage, AI $100 "Federal Victory Note," and repost calling for Obama's arrest
- WSJ reveal (May 12, 2026): Executive assistant Natalie Harp logs into the account and posts batches of pre-approved drafts; no internal review process
- Notable deleted posts: AI Trump-as-Jesus (Apr 13, 2026, 87% negative in WaPo poll); various others quietly removed
Sources: NPR analysis · Snopes 55-post spree · WSJ via TNR — Harp pipeline
🔍 Inspector General & Watchdog Removals (Term 2)
17–18 Inspectors General fired in one night — the "Friday Night Massacre" of January 24, 2025. The firings violated the 1978 IG Act's 30-day congressional-notice provision (as amended in 2022). Public Citizen counts at least 21 IGs total removed through 2026 across additional rounds. Beyond IGs: OSC (Office of Special Counsel) head Hampton Dellinger forced out via litigation; multiple career DOJ prosecutors removed (most recently 6+ in EDVA over the Comey second-indictment refusals, May 2026); ethics officers and watchdog appointees systematically replaced. Departments affected in the January 2025 mass purge: Defense, State, HHS, Labor, Interior, Agriculture, Commerce, Education, EPA, HUD, SBA, Transportation, VA, plus others.
Sources: Public Citizen tracker · Wikipedia: 2025 IG dismissals · Washington Post original report
🏢 Federal Workforce Cuts by Agency (Term 2)
Per OPM (Nov 2025) and Boston Globe analysis (Feb 26, 2026): ~317,000–352,000 employees exited federal service in 2025; net workforce reduction ~212,000 (~9% of 2.3M civilian workforce). About 123,000 took the deferred resignation buyout. DOGE's claimed $160B savings is disputed — CBS/Partnership for Public Service analysis says the cuts have cost ~$135B in lost productivity, rehiring, and severance.
| Agency | Approximate exit count | Note |
|---|---|---|
| Department of Defense | 60,000+ | Largest absolute headcount loss |
| Treasury (incl. IRS) | 30,000+ | ~7,000 IRS probationary cuts; partially reversed by court order |
| USDA | 20,000+ | Forest Service, FSIS heavily affected |
| USAID | ~7,000 | Effectively dissolved as a standalone agency; functions absorbed by State |
| HHS | ~2,000 direct + subagency cuts | CDC, NIH, FDA contract terminations layered on top |
| Department of Education | disputed; wound down via attrition | ~40+ direct firings cited; functions reassigned |
| FEMA (within DHS) | 14% reduction by Nov 2025 | Hurricane-season operational concerns flagged by GAO |
Sources: Federal News Network (OPM data) · Boston Globe analysis · Wikipedia mass-layoffs page
📜 Pardons & Commutations Breakdown (Term 2)
~1,600+ total clemency grants in Term 2 through May 2026. Per NBC analysis of disclosed cases, the grants have erased fines and restitution totaling $298 million; California Governor's office estimates the cumulative debt-erasure at ~$2 billion in victim repayment and taxpayer recovery (March 5, 2026).
| Category | Approximate count | Note |
|---|---|---|
| January 6 defendants | ~1,500 | Mass pardon, Jan 20, 2025 (Inauguration Day); plus 14 Oath Keepers/Proud Boys commutations |
| Fake-electors / 2020 plot | 77 preemptive | Proclamation Nov 10, 2025; incl. Giuliani, Meadows, multiple state-level participants |
| Individual non-Jan-6 pardons | 88+ named individuals (through Jan 2026) | More by May 2026; full count lags at DOJ Pardon Attorney |
| White-collar / fraud cases | ~70+ across both terms | >50% of individual pardons (NYT March 19, 2026) |
| Drug offenders | minor share | No consolidated count published |
Sources: DOJ Office of the Pardon Attorney · Wikipedia clemency list · NBC analysis (March 2026)
⚔️ Iran War — Cost & Casualties
War began Feb 28, 2026. Per Secretary Hegseth and Gen. Caine's testimony to Senate Appropriations on May 12, 2026: ~$29 billion direct U.S. cost (up $4B from the April 29 estimate). $24B for munitions/drones/aircraft replacement; additional ~$4B+ for damage to U.S. military facilities (not included in the $29B). 60-day War Powers Act deadline passed May 1, 2026; administration's position is that the April 8 ceasefire "paused" the clock — text not in statute.
| Category | Number | Reliability |
|---|---|---|
| U.S. direct cost | ~$29 billion through May 12 | VERIFIED (Hegseth testimony) |
| U.S. KIA | 13 | VERIFIED (DoD releases) |
| U.S. WIA | ~373 | DISPUTED — The Intercept (4/22) reported Pentagon scrubbed 15 WIA |
| Iranian military KIA | ~1,221 (HRANA / U.S.+IDF claims) | ESTIMATE — not independently verified |
| Iranian civilian deaths | ~1,500 (WaPo) / ~1,701 (HRANA) / 3,468 total per Iranian gov sources | ESTIMATE — wide range across sources |
| Hormuz vessel traffic | ~140/day pre-war → ~30 since 5/14 | ESTIMATE (IRGC reporting) |
| Brent crude | ~$75 → peak $109/bbl | VERIFIED |
| U.S. retail gas | $2.94 → $4.628/gal (+57%) | VERIFIED (AAA) |
Sources: NOTUS cost report · The Intercept (casualty scrubbing) · WaPo civilian deaths · Al Jazeera live tracker
💰 Tariff Cost Pass-Through by Income Decile (Term 2)
Yale Budget Lab's "State of U.S. Tariffs: April 8, 2026" calculates the burden as a share of post-tax-and-transfer income, broken down by income decile. The bottom decile pays roughly 3x the rate of the top decile as a share of income — the regressive ratio that has held across every credible model of the tariff incidence.
| Income decile | Burden as share of income (Section 122 expires) | Burden as share of income (Section 122 extended) | Annual cost per household (extended) |
|---|---|---|---|
| Bottom 10% | 1.1% | 1.9% | $813 |
| Middle 60% | ~0.7% | ~1.2% | ~$1,500–2,200 |
| Top 10% | 0.4% | 0.6% | $3,424 |
Section 122 tariffs were held unlawful by the Court of International Trade on May 7, 2026; stayed by the Federal Circuit on May 12, 2026. The "extended" scenario remains operative for non-party importers (which is most of the market). Average household impact: ~$1,200–$2,800/year depending on scenario.
Sources: Yale Budget Lab (April 8, 2026) · Budget Lab SCOTUS-ruling update
🚓 Deportations & Detentions (Term 2)
Cumulative deportations Jan 20, 2025 → mid-May 2026: ~270,000+ (pace ~400,000+ annualized). First 100 days of 2026 alone: ~65,000 removed, 66,000+ arrested. Interior vs. border split no longer disclosed monthly by ICE — the administration ended detailed disaggregation. ICE detention population: 60,311 as of April 4, 2026 (TRAC). Of those, 70.8% (42,722) had no criminal conviction.
El Salvador / CECOT pipeline: 9,000+ Salvadorans deported since Jan 2025 per Human Rights Watch (March 2026); only 10.5% had a U.S. conviction for violent or potentially violent crime; 280+ men transferred to CECOT in March–April 2025 with documented disappearances (per NILC, HRW). California reported 6 deaths in ICE detention facilities in May 2026 alone (CalMatters).
Sources: TRAC detention statistics · HRW El Salvador report · deportationdata.org · NILC CECOT tracker
📚 Where These Numbers Come From
Forbes 400 historical snapshots · Federal Election Commission filings · OPM federal workforce reports · Department of Justice pardon database · Office of the Pardon Attorney · BLS unemployment and CPI · BEA GDP · Tax Foundation tariff incidence studies · Yale Budget Lab tariff cost models · ICE deportation reports · Gallup approval polling · Federal Register executive order count · CourtListener court filings · SEC public filings (DJT) · Trump Org SEC disclosures · Trump campaign FEC filings · Trump White House press office daily schedule · Trumpgolfcount.com (cross-checked against pool reports). All numbers above are reproducible from public sources. If you find one wrong, email projectshadowqa@protonmail.com.
Evidence reviewed through August 16, 2026 · Numerical snapshots retain their source-specific dates and are not silently refreshed · No paywalls, no ads, no sponsors, no foreign money · The Record is self-funded
Methodology
How The Record is built, sourced, and maintained
1. Purpose and Scope
The Record is a chronological accountability archive documenting the political career of Donald Trump from pre-political history through its embedded August 16, 2026 evidence snapshot. It is not a live news feed, a blog, or an opinion column. It is a structured research dataset organized for pattern recognition across time; older records carry mixed review status. Every entry is designed to ask: what happened, when, and why might it matter?
2. The Three-Layer Entry Format
Every entry in The Record uses a strict three-layer format designed to separate fact from analysis from editorial interpretation. THE FACTS is the first layer — written in neutral, journalistic prose describing what happened, sourced to verifiable reporting. This layer reads like wire copy: no adjectives of judgment, no "clearly" or "shockingly," no opinion. SIGNIFICANCE is the analytical layer, explaining why the event matters in context — what pattern it fits, what precedent it sets, what power dynamic it reveals. GOALPOST is the most editorial layer, documenting the rhetorical shift or norm erosion — the actual talking points used to defend or normalize the event, and how those defenses compare to earlier stated positions. These three layers are always visible and always labeled, so readers can distinguish sourced facts from informed analysis from pointed editorial observation.
3. Sourcing Standards
Every entry in the national timeline is backed by at least one source citation. Sources are drawn from major wire services (AP, Reuters), national newspapers, broadcast journalism, government records, court documents, and specialized investigative outlets. At the August 16, 2026 review boundary, the embedded timeline contains 4,789 records and 6,825 citations; the preserved correction transform renders 4,796 records and 6,842 citations across 356 domains. Its source URLs use HTTPS and it contains no Google-search fallback links. The separate People dossier corpus is older and does contain disclosed search placeholders. Those links are not represented as verified citations, are visibly labeled in each dossier, and remain in a replacement queue.
4. Entry Types and Date Precision
The Record distinguishes between two types of entries. Event entries document a specific occurrence on a specific date — a speech, a firing, a court ruling, a tweet. These make up the vast majority of the archive. Context entries, marked with a purple border, document broader patterns, systemic developments, or periods that span weeks or months and cannot be reduced to a single date. Each entry also carries a date precision field: "day" for events with a confirmed specific date, "month" for events known only to have occurred during a particular month, and "year" for historical events where only the year is documented. This precision metadata is displayed to the reader so they know exactly how confident the dating is.
5. The Six Eras
The archive is organized into six chronological eras that correspond to distinct phases of Trump's political trajectory. Formation covers everything before June 16, 2015 — the family history, business career, early political flirtations, and the psychological and financial architecture that produced the candidacy. Campaign 1 runs from the golden escalator announcement through Inauguration Day 2017. Term 1 covers the entire first presidency. Post-Presidency spans from January 20, 2021 through November 14, 2022 — the January 6 aftermath, legal exposure, and the beginning of political rehabilitation. Campaign 2 covers the second campaign through Inauguration Day 2025. Term 2 covers the second presidency only through this archive's August 16, 2026 evidence cutoff. Each era has its own companion document with every entry from that period.
6. Tagging and Classification
The Record uses 30 automated topic tags powered by regex pattern matching against entry text. Tags include Russia/Mueller, Immigration, Courts & Legal, Economy & Trade, COVID-19, January 6, Elections, Ukraine, Media & Press, DOGE & Spending, Epstein, Culture War, Deaths & Consequences, Follow the Money, and others. Entries can match multiple tags simultaneously. In addition to topic tags, entries are classified as "Major" (significant enough to be highlighted in filtered views) or "Goalpost" (documenting a specific instance of norm erosion, rhetorical shifting, or moved goalposts). These classifications are editorial judgments made by the editor and are subject to revision.
7. AI Transparency
The Record is built with two AI systems: ChatGPT/Codex Sol (OpenAI) and Claude Fable 5. AI is used for drafting entry text from source material, analytical assistance in the significance and goalpost layers, code generation for the HTML app and document generators, and quality assurance auditing. The editor reviews, edits, and approves all AI-generated content before it enters the archive. AI-generated opinion essays (the "AI Opinion" section) are clearly separated from the factual timeline and explicitly labeled as AI-generated analysis. The archive does not use AI to fabricate events, invent sources, or generate unverified claims. Every factual assertion in the FACTS layer is traceable to a cited source.
8. The Pause Before Harm Protocol
This project operates under the Pause Before Harm Protocol (PBHP), an open-source harm-reduction framework for AI and human decision-making. PBHP asks: "If I'm wrong, who pays first — and can they recover?" It prescribes a 7-step decision flow: name the action, identify constraints, check for impermissible constraints, identify harms to the least-powerful stakeholders first, rate each harm, assign a risk gate, and act on the gate with logging. Applied to The Record, PBHP means: every entry must name who is harmed by the documented events, language must maintain "brutal clarity with zero contempt" (no mockery, no dehumanization), and the archive must be transparent about its own methods, limitations, and potential biases. The protocol is available at github.com/PauseBeforeHarmProtocol/pbhp.
9. Companion Documents and Master PDF
The Record produces 34 companion DOCX documents that provide focused views into the data: 6 era documents (one per era), 16 topic documents (Courts, Democracy, Media, January 6, Epstein, Foreign Influence, etc.), 5 AI opinion essays, and documents covering methodology, editorial approach, politicians, people mentioned, and internal audits. These documents are regenerated from the same source data as the HTML app, ensuring consistency. All 34 documents are compiled into a single master PDF (THE-RECORD-COMPLETE.pdf) that runs approximately 2,900 pages. Every companion document includes full attribution, contact information, and project links.
10. Corrections, Limitations, and Accountability
The Record is maintained by a single editor, Phillip Linstrum, a QA professional. This is both a strength and a limitation — it ensures consistent editorial standards but means the archive reflects one person's judgment about what events to include and how to frame their significance. The editor is personally accountable for every entry and welcomes corrections. If you identify a factual error, a missing source, a misattribution, or an unfair characterization in any layer, report it to projectshadowqa@protonmail.com or via facebook.com/plinst; no response-time SLA is promised. Submitted and resolved items are preserved in the public corrections ledger. The goal is accuracy and completeness, not advocacy — though the archive makes no pretense of neutrality about whether accountability matters. It does. That is the premise. Everything else is evidence.
What the archive still needs
These are disclosed limitations, not hidden footnotes. Counts are calculated from the data loaded in this dated archive build.
4,422 search placeholders in the People corpus still need direct article, transcript, court, or government URLs.
10,480 related profile records need an explicit role field: action by, statement by, action affecting, quoted by, relationship, or context.
The new chamber indexes 332 curated timeline entries. A complete tweet ingest still needs tweet IDs, timestamps, original/retweet status, deletion status, archive snapshot, and citation provenance.
The public ledger preserves submitted and resolved items; each published correction should retain original text, corrected text, reason, evidence, date, and affected entry permalink.
Politician cards now refuse guilt by association, but the underlying data should encode why each person is attached to each record rather than relying on the reader to infer it.
Major events should link forward to measurable outcomes recorded by the relevant cutoff: court disposition, policy implementation, people affected, cost, reversal, and status.
📱 Truth Social — External Dataset
This optional panel is separate from the curated August 16 archive. It starts with an embedded fallback sample and may attempt a browser-time retrieval from named mirrors; neither path creates an editorial currentness claim.
External dataset receipt
- Editorial relation
- Optional external research panel; not part of the curated national chronology.
- Embedded fallback
- Bundled sample; its post timestamps do not establish retrieval time.
- Runtime source
- Not requested yet.
- Retrieval marker
- No browser retrieval yet; no trusted publisher fetched-at or update time recorded.
AI-opinion layer attributed
A frozen, versioned prompt (RECORD-CORPUS-ANALYSIS v1.0) run by Claude over a named 32,754-post corpus, via the IaskAI run—figures attributed, not Record facts; internal variance shown (accusation 6,262–~6,838). Full write-up: AI Opinion → “The Truth Social Corpus.” Reproduction requires the same prompt, exact source dataset, and named model; this page does not independently recount the result.
About The Record
What This Is
The Record is a chronological research archive of Donald Trump's political career — from the earliest embedded family-history records through its August 16, 2026 evidence snapshot. Every entry uses a three-layer format: THE FACTS (the documented claim), SIGNIFICANCE (why it may matter in context), and GOALPOST (the argument used to normalize it). This is not a live news feed, and the deep archive is not uniformly reviewed.
Entry Types
Specific Event entries document a particular event on a particular date with verifiable facts. Context entries (marked with a purple border) document broader patterns, periods, or systemic developments that span time ranges. Context entries are labeled to distinguish them from single-day events.
Editor & Accountability
Maintained by Phillip Linstrum, a QA professional. Built with ChatGPT/Codex Sol (OpenAI) and Claude Fable 5. This project is self-funded and independent. The editor is personally accountable for every entry. AI-generated content (the "AI Opinion" section) is clearly labeled as such and separated from the factual timeline.
Methodology
Facts are drawn from major news reporting (AP, Reuters, NYT, Washington Post, CNN, BBC, court filings, congressional records, and official government documents). The significance layer provides analytical context grounded in documented evidence. The goalpost layer documents the actual rhetorical defenses used by supporters — not as straw men, but as observed talking points. Entries are tagged by topic and era, and filtered by automated keyword matching.
Corrections Policy
If you identify a factual error, missing context, or misattribution, please report it. Corrections are timestamped and transparent in the public corrections ledger. The goal is accuracy, not advocacy. Contact: projectshadowqa@protonmail.com · no response-time SLA · facebook.com/plinst
Pause Before Harm Protocol
This project is built through the Pause Before Harm Protocol (PBHP) framework — an approach to responsible AI use that prioritizes transparency, evidence, and accountability over speed or persuasion.
Companion AI
Have questions about The Record? Ask The Record Companion AI — a custom GPT trained on the complete master document. It can answer questions about any entry, era, topic, sourcing, methodology, or the Pause Before Harm Protocol.
Data & Technical
Evidence reviewed through August 16, 2026. Source data is maintained in structured JavaScript format with an automated build pipeline. 34 companion DOCX documents provide deep-dive analysis by topic.