The record, as it stands.

The 2026 claims checked against the record: Maryland’s “500,000 fake ballots” was a vendor printing error the state says could not produce a single duplicate vote, and the documented mail-fraud rate runs under one case per million ballots against 46.8 million cast. Five independent checks of 2020 — including the recount Trump’s own allies ran — reached the same answer. With the honest kernels quantified: a 1.2% rejection rate, NC-09, and uneven state safeguards.

Evidence base: EAC EAVS 2024 · PNAS ×2 · GA SOS audits · Lost, Not Stolen · state statutes · primary sources

On June 30, 2026, the Supreme Court ruled 6–3 that US-born children are citizens regardless of their parents’ status. Hours later came the claim that ordinary legislation could undo it. The 1898 Court answered that exact question — Congress has “no authority… to restrict the effect of birth” — and the one time Congress legislated on birthright citizenship, it expanded it. Every quotation verified against the primary documents.

Evidence base: Trump v. Barbara slip op. · Wong Kim Ark (1898) · Constitution Annotated · CRS · primary sources

The 2026 Trustees Report says benefits don’t go to zero — 78–83% keep flowing after the trust funds run dry in 2032–2034 — but the 75-year gap just hit 4.42% of payroll, the largest in decades, and even scrapping the payroll cap entirely closes only half to two-thirds of it. The doom, the denial, and the one-weird-trick fixes, all scored against SSA’s own actuaries.

Evidence base: 2026 Trustees Report · SSA OACT provision scores · CBO · CRS · primary sources

CBO’s own analysis of the 2025 tax law: the poorest tenth lose ~$1,200 a year while the richest tenth gain ~$13,600 — but the safety-net cuts cover only about a quarter of the tax cost. The tax cuts were paid for by borrowing, and the “no one loses coverage” defense fails against CBO’s own 7.5 million. A left-coded claim and its right-coded rebuttal, both checked against the same documents.

Evidence base: CBO distributional analysis · JCT tables · Senate roll call 372 · NEJM · primary sources

The typical (median) account gained about $6,000 last year — the claim is off by five times and describes only the top decile or two. Built from the account custodians’ own data, with the market-vs-paycheck gap measured (+24% S&P, +6.5% accounts, −0.7% real hourly pay) and the doom version checked just as hard. The portable lesson: whenever a politician says “typical,” demand the median.

Evidence base: Vanguard How America Saves · Fidelity Q1 2026 · Census P60-286 · BLS · primary sources

A sourced fact-check of the claim that undocumented immigrants are collecting or draining Social Security. Federal statute has barred payment to anyone not lawfully present since 1996; SSA’s own actuaries found unauthorized work a net positive of roughly $12 billion to the program in a single year, and CBO projects the recent surge population pays in at roughly 348 dollars for every dollar paid out. Checks the 2025 “150-year-olds” and “millions of SSNs” claims against the government’s own audits — with the kernel of truth (real SSN misuse at scale) stated plainly, not hidden.

Evidence base: 8 U.S.C. §1611 · 42 U.S.C. §402(y) · SSA Actuarial Note 151 · SSA OIG A-06-21-51022 · CBO 60569 · CRS · primary sources

A sourced fact-check of the claim that drawing strength from a broad population — across class, origin, and race — is a modern invention. The army that saved the Union was one in four foreign-born, ten percent Black, and nearly half farmers; the wealthy were the ones who could buy their way out. Traces the breadth of who served under Grant through to the Grand Army of the Republic, with charts, a Claims vs the Record panel, and the honest caveats stated plainly.

Evidence base: NPS · National Archives · American Battlefield Trust · Gould 1869 · McConnell · primary sources

An interactive map of every documented, court-verified voter fraud case tied to a US federal election cycle from 2000 to 2024 — geolocated by state and county, tagged with the party of the perpetrator, and counted by ballots actually affected. Paired with a Claims vs Reality panel that checks the loudest fraud claims against the documented record.

Evidence base: Court records · prosecutions · state audits · primary sources

A primary-source audit of what the public data actually shows about immigration status and crime. Texas is the only state with a comprehensive longitudinal offender-status dataset; federal data (USSC, BOP, ICE ERO, GAO) covers the federal slice; forty-nine other states publish nothing usable. Paired with a Claims vs the Documented Record panel that tests eight of the loudest assertions from both directions — with every think-tank source labeled by its institutional priors.

Evidence base: USSC · BOP · ICE ERO · Texas DPS · GAO-24-107598 · peer-reviewed PNAS · primary sources

A data-driven case study on the persistent unemployment and underemployment gap between veterans and the broader labor force. Built from live BLS data, Phase 1 brief through full analytical deliverable.

Evidence base: BLS series LNU04049526 · LNS14000000 · 2023–2026

After-action review of the website publishing process, built on Six Sigma methodology. Identifies defect categories, measures publishing KPIs, and documents the improvements that moved the process from unstable to predictable.

Evidence base: Publishing logs · KPI tracker · AAR format

A working document accompanying the Reconstructing Corporate Personhood book. Maps the internal contradictions the decision introduced into the corporate-personhood line of cases, case by case, without policy advocacy.

Evidence base: Supreme Court opinions · common-law precedent · secondary legal scholarship

A full analytics case study on veteran retention in civilian workplaces. Six delivered phases, retention logistic-regression model, feature importance analysis, executive summary, and a documented model card. Build v0.13.

Evidence base: Phased analytics pipeline · documented assumptions · model card