Articles traced back to the evidence

TruthBased.org covers a variety of subjects, including politics, economics, health, history, physics, and psychology. Each article traces its claims back to the strongest evidence available, whether that is official statistics, a peer-reviewed study, or a systematic review of many. Each states what the evidence supports, presents the strongest opposing case, and says what would change the conclusion.

Political Fact Checks

Claims politicians make, measured against the record

Do Voter ID Laws Reduce Turnout?Thirty-six states ask for ID at the polls. The largest study, on 1.6 billion voter records, finds no turnout effect and rules out a drop above half a point. Single-state studies find drops. The 10.3-point figure everyone quotes was taken apart in the same journal that published it.
Is Voter Fraud a Real Problem?Heritage’s database lists 1,619 cases from 1982 to 2025, about 7 for every 10 million ballots cast in federal general elections alone. Fraud has forced one congressional election to be re-run; nothing found comes near a presidential result, and the kind voter ID is aimed at is among the rarest.
Do the Richest Americans Pay Less Tax Than Everyone Else?The top 1 percent paid 26 percent of the income they reported in 2023, taxpayers as a whole 14 percent, and the 400 richest families about 8 percent once gains on unsold stock were counted as income. Which figure answers the question depends on what counts as income, and the data cannot settle that.
Is There Really a Border Crisis?It depends what you count. Counted as encounters, which are events, FY2021 to FY2023 were the three highest years on the Border Patrol's series; counted as attempted crossings, on DHS's catch-rate estimates, the years around 2000 were probably higher.

History & Myth-Busting

Things everyone "knows," tested against primary documents

Was Hitler Actually a Socialist?The party name, Hitler's own words, privatization data, the Night of the Long Knives, and what happened to actual socialists.
Did the U.S. Political Parties Switch?Roll-call votes by region, Black voter data, congressional seats 1960-2020, and the Southern Strategy in the architects' own words.
What Does the Record Actually Show About Columbus?The flat-Earth myth debunked, Columbus's journal entries, Hispaniola data, and how Washington Irving built the myth.

Science & Perception

Viral phenomena explained with primary source neuroscience and physics

The Dress: White & Gold or Blue & Black?The same pixels. Two completely different colors. The neuroscience of color constancy, Bayesian inference, and what it means for shared reality.
The Mandela Effect: Mass False Memory or Something Else?Berenstain or Berenstein? Did the Monopoly man have a monocle? Why millions share the same vivid, confident, wrong memories, and what that reveals about the mind.
The Mpemba Effect: Does Hot Water Freeze Faster Than Cold?Aristotle noticed it in 350 BCE. A Tanzanian student rediscovered it in 1963. Sixty years of modern physics later, the answer is still contested, and more interesting than either side admits.

Science & Media Accuracy

What studies actually say vs. how they get reported

Is Climate Change Real and Human-Caused?The IPCC puts the warming since 1850–1900 at about 1.1 °C on four independent records and the human share at all of it; 2024 was the warmest year measured, CO₂ is at its highest in 800,000 years, and the ocean, sea level and ice agree. What coverage gets wrong, both ways.
How Dangerous Is Nuclear Power?Deaths per terawatt-hour for every major energy source, what Chernobyl and Fukushima actually killed, and why the strongest case against nuclear power is about cost rather than safety.
Do Trans Women Retain an Athletic Advantage After Transition?Testosterone suppression takes haemoglobin back to cisgender female levels in about four months, reduces muscle by about a twentieth in the first year, and does nothing at all to the skeleton. The largest synthesis to date finds strength and aerobic fitness comparable, at low certainty, while most federations have banned participation anyway.

Economics & Policy

Contested economic claims vs. BLS, CBO, and peer-reviewed data

The Gender Wage Gap: What the Data Actually ShowsThe raw 77-cent figure vs. the adjusted gap. What each statistic actually measures, what causes the residual gap, and what neither side in the debate gets right.
U.S. Inflation (2021–2026): Causes, Drivers, and International ContextThe CPI path from 0.3% to 9.1% and back, decomposed into supply-chain, stimulus, and energy effects, compared internationally, with the Federal Reserve's response.
U.S. Federal Spending: What We Spend and What It CostsThe FY2026 budget by category, including Social Security, Medicare, defense, and net interest, and the debt trajectory CBO projects.
Is College Still Worth the Cost?Wage premiums, tuition, debt loads, graduation rates, return by major and institution, and the AI question, from BLS, Census, and College Scorecard data.
Do Tariffs Help U.S. Manufacturing?Steel, tires, China, and the 2025 escalation, with each episode measured by jobs saved, jobs lost downstream, cost per job, consumer prices, and the trade deficit.
Does Raising the Minimum Wage Kill Jobs?Employment, poverty, prices, small businesses, and automation, drawing on Card and Krueger, CBO, Dube, and Neumark.
The National Debt: What Both Parties Actually DidDeficit spending by administration, by Congress, and by policy area, from CBO and Treasury data.
Job Creation Claims: Presidents vs. The DataEach administration's jobs record against BLS data, with context for policy lags, inherited recessions, and the Blinder-Watson finding on luck.
Communism and Capitalism: What Do the Measured Outcomes Show?Divided Germany and Korea, the China and Vietnam reform breaks, famine ranges and global poverty. China's income gains came after the reforms and its life expectancy gains almost entirely before them.

Health & Medicine

Medical statistics and health claims vs. peer-reviewed evidence

Are Supplements Worth It, or a Waste of Money?Six in ten American adults take a supplement. The largest trials find little or no benefit for a healthy adult who is not short of anything; for a short, specific list of people, supplements are treatment rather than insurance.
Vaccine Safety & Efficacy: What the Data Actually ShowsStudies covering more than 23 million children across 138 studies, the autism question, the real but rare adverse events, COVID vaccine trials, myocarditis rates, VAERS and V-safe, and the childhood schedule debate.
Is Ultra-Processed Food Bad for You?Two controlled feeding trials, the cohort studies of death and disease, and the kinds of ultra-processed food most consistently linked to harm.
Social Media & Youth Mental HealthCDC YRBS, NHS Digital, the ABCD cohort, meta-analyses, and natural experiments on adolescent mental health, with effect sizes stated and the Haidt and Odgers positions compared.
How to Help Someone With an Eating DisorderFor families and friends. The emergency signs, what to say, the reassurance that backfires, meal support, and an honest account of how little of this advice has actually been tested.
Anorexia Nervosa: What It Is, and What Actually HelpsOne of the highest death rates of any mental illness, and the version where the person is not underweight is the more common one. Heritability, the starvation experiment that reproduced the psychology in healthy men, what treatment the evidence supports, and how to help someone.
How to Help Someone With DepressionFor families, partners and friends. Asking someone directly about suicide has not been found to raise their risk, and a written safety plan has more behind it than a promise. Most everyday advice on what to say and do has not been tested; this page says which is which, and lists helplines for supporters.
Depression: What It Is, and What Treatment Actually DoesTwo people can be diagnosed with major depression and share no symptom. The 227 ways to qualify, why the 15% lifetime suicide figure was a counting error, what antidepressants and therapy do once publication bias is taken out, and the fact that treatment quadrupled while prevalence stayed flat.
What Exactly Creates a Sociopath?The warrior gene, childhood maltreatment and brain damage are each offered as the cause of sociopathy. This traces what twin studies, MAOA research, maltreatment data and brain imaging each measured, and how much any of them explains.
What Is Gender Dysphoria, and What Does the Evidence Actually Show?Adults. What the diagnosis is, the distinctions people get wrong, the causes put forward, and what adults who have been through treatment report. Every figure with its population and its certainty.
Can Sociopaths Be Fixed?Companion to the causes article. What has been tried for antisocial personality disorder, what the trials measured, and why nobody yet knows whether it can be treated.

Crime & Justice

Policing, punishment, guns and political violence: what the data can and cannot show

Did Police Funding Cuts Cause the 2020 Crime Surge?Firearm homicide rose 46% from 2019 to the 2021 peak. Police budgets in 264 cities were mostly not cut, the four-city study of police pullback found no clear rise in violent crime, and the gun-buying surge and pandemic disruption fit the timing better.
Does the Death Penalty Deter Crime?No one has shown that it does. The National Research Council found the research not informative either way, and the 2003 study behind the claim of 18 murders prevented per execution reversed when re-run. Capital cases cost more in six state studies, and 203 people have been exonerated from death row since 1973.
Is Mass Incarceration Mainly a Product of the War on Drugs?Releasing every drug prisoner in America frees about 196,000 people and leaves more than a million inside. But drug crimes were about a third of prison admissions, and Black and Hispanic state prisoners are less likely than white ones to be held on a drug conviction.
Which Gun Policies Actually Have Evidence Behind Them?RAND assessed 18 policies against 8 outcomes. Of 144 pairings, 18 showed an effect, 50 were inconclusive and 76 had no qualifying study at all. Four reach the top tier, and two of those find a policy increases homicide.
Why Does the U.S. Have More Gun Deaths?The U.S. firearm death rate, 10.4 per 100,000 on the Global Burden of Disease’s 2019 rates, is nearly five times France’s, the next highest peer, and the gap is half suicide and half homicide. Without a gun, Americans are killed at 1.2 times the Canadian rate; with one, at six and a half times it. CDC, GBD, Statistics Canada, the ONS, the Small Arms Survey and RAND.
Which Side Commits More Political Violence?Right-wing attackers have killed more people than left-wing ones in every count of a decade or more, from CSIS, the Cato Institute, New America and ADL, while in 2025 CSIS counted more left-wing than right-wing attacks and plots, and more people killed by left-wing attacks; what each count includes and the filing choices that move it.

Groups & Crime

Claims that a particular group commits more crime, or suffers more of it, checked against the records that exist

Is It True That Trans Women Are More Likely to Be Sex Offenders?The data available does not show that trans women are more likely to be sex offenders, or less likely. Two versions of an infographic put trans women in England and Wales at 1,916 and 1,177 per million in prison for sexual offences. Both divide a prison headcount by a census count that the Office for National Statistics says should not be used this way.
Are Trans People Overrepresented as Mass Shooters?No. On the Gun Violence Archive, the only database big enough to test it, trans people carried out 4 of 5,748 US mass shootings from 2013 to September 2025, 12 times fewer than their share of the population would give; the Crime Prevention Research Center’s 6.2 times rests on one confirmed case.
Are LGB People More Likely to Be Sex Offenders?Companion to the victimisation article. The claim is about gay men and children, and no official statistics in the US or in England and Wales give an offending rate by orientation. The three small studies both sides cite, the 10-to-20-times figure and what it counted, and the strongest case the other way.
Do Violent Offending Rates Differ by Race?The last homicide offending rate by race that BJS published is 34.4 per 100,000 against 4.5, and for serious nonfatal violence, victims’ descriptions of white and black offenders match the arrests. Where controls are applied the gaps shrink, some to nothing measurable.
Are the Racial Violence Statistics Being Misread?Yes, both ways. The FBI table people screenshot is a count with no population figure that gets read as a rate, and the offender was unknown for 27.8% of the victims in its file, not half. The 84% grooming gangs figure has no source that can carry it, and neither does the national claim the other way.
Do Immigrants Commit More Crime?No: 674 undocumented and 303 legal immigrants incarcerated per 100,000 aged 18 to 54 against 1,195 native-born in Cato’s estimates from the 2024 American Community Survey, Texas arrest rates say the same, and why Europe differs.
How Often Are Transgender People Victims of Crime?Companion to the sex-offender article. What probability surveys in three countries find, the prison surveys that put transgender inmates at eight to thirteen times the victimization rate of other inmates, the single English victim figure, and why homicide counts cannot become rates.
Are LGB People More Likely to Be Victims of Crime?American survey data shows a large gap in violent crime, and most of it survives adjustment. The British analysis in which the gap vanishes covered crime of every kind and not violence. Bisexual people have by far the highest rates, and neither country systematically records the sexual orientation of murder victims.

Religion & Society

Measurable claims about religiosity and social outcomes

Does Religiosity Correlate With Better Social Outcomes?Crime rates, social trust, wellbeing, charitable giving, education, what Pew Research, Gallup, Harvard longitudinal cohorts, and dozens of meta-analyses actually find when comparing more and less religious populations. The answer depends on whether you measure individuals or societies.
Does Religion Cause Violence?Since 1975 a growing share of armed conflicts has been fought over religious demands, and religious terrorist groups have been deadlier; but more religiously diverse countries have been no more likely to have civil wars once income is taken into account, and the popular death tolls for the Inquisition and the witch trials are far above what historians count.

International Comparisons

How U.S. outcomes compare globally

Most Educated Countries in the World47 countries ranked by tertiary attainment using OECD Education at a Glance 2025. What each top-third nation specifically excels in. The paradox where more years of schooling doesn't mean better outcomes. Gender gaps, spending efficiency, and the bottom third.

Source standards and how conflicts between sources are handled are described on the methodology page.