How a claim earns its place.
The standard, in one line: every statistic, ranking input, daily fact, and trivia answer on this site traces to a real, citable primary source — checked, dated, and adversarially verified — or it is marked data pending and shown to no one as fact. That restraint is not a limitation of the work. It is the work.
Black Intervention is a fact-based intervention. The premise is simple and unfashionable: the most important questions facing Black America deserve to be argued from the record, in broad daylight, with the sources attached. What follows is that record's operating standard. Read it as a journalist or academic would read a newsroom's stylebook or a research lab's protocol — because that is what it is. If a single claim on this site fails the test described here, it is a defect, and we want to know.
We cite the people who collected the data, not the people who summarized it.
A claim is only as good as its origin. We go to the institution that produced the measurement — the census enumerator, the test administrator, the crime-data submission — never to an aggregator's retelling of it. The recurring primary sources behind this site are:
The core sources
- FBI Crime Data Explorer — reported offenses and clearances, used for all safety and crime metrics.
- U.S. Census Bureau / American Community Survey (ACS) — population, income, poverty, homeownership, and demographic shares.
- Bureau of Labor Statistics (BLS) — employment, unemployment, and wage data.
- NAEP / NCES — student proficiency and the federal education statistics behind every school ranking.
- CDC & County Health Rankings — health outcomes and mortality.
- City ACFRs (Annual Comprehensive Financial Reports) — municipal fiscal health, audited.
- The Lincoln Institute property-tax study — comparative property-tax burden.
The sources we refuse
We do not cite or link to crowd-edited encyclopedias or content aggregators. Not as a source, not as a convenience, not "just for the date." It is a tertiary, editable surface — exactly the kind of secondhand authority this project exists to replace. Every figure here resolves to the body that originally measured it, with the dataset, year, and table identified so a reader can reproduce the number themselves.
Every figure carries its vintage. Data has a date; a 2019 poverty rate and a 2023 poverty rate are different facts. We label the year and refresh against the latest release rather than letting a number quietly go stale.
We tell you how strong the evidence is — not just what it says.
Honesty about certainty is itself a form of accuracy. Every substantive claim on this site carries a status, surfaced in the interface so you are never guessing how firm the ground is:
Settled
Where the evidence is lopsided, we say so plainly. A claim is marked settled when the primary sources converge and serious dissent is absent or marginal. Settled does not mean "popular" or "comfortable" — it means the record is one-sided, and we will not manufacture a false balance to soften it.
Contested
Where serious, credentialed researchers genuinely disagree, we label the claim contested and show the disagreement — the competing findings, side by side, with each side's sourcing. We do not resolve a live scholarly dispute by fiat or pick the answer that flatters a narrative. The reader sees the fault line and the evidence on both sides of it.
The distinction is deliberate and load-bearing. Collapsing contested questions into false certainty, or muddying settled questions with false doubt, are the two failure modes that erode trust in data journalism. This site refuses both.
Two independent authorities, and a screen for every "fact everybody knows."
A trivia game and a daily fact are only worth playing if the answers are right. The bar here is higher than for ordinary editorial copy, because a quiz answer is asserted as settled by definition:
Two independent sources
Every trivia answer and every Fact of the Day must be confirmed by two independent authoritative sources — not one source quoting another, but two that arrived at the fact separately. A single citation is never enough to ship a fact as settled.
The myth-screen
Every surprising, counterintuitive, or commonly-mistaken claim is run through a myth-screen before it can publish: a deliberate check for the popular version that turns out to be wrong — the misattributed quote, the off-by-a-decade date, the invention credited to the wrong person. If a claim is famous and shaky, it is screened until it is verified or dropped.
Nothing auto-promotes
Machine-generated or candidate questions enter a holding state for human review. Nothing is auto-promoted into the live pool. A fact is live because a person verified it against the standard above — never because a pipeline produced it.
We score results an office can influence — never party, never ideology.
The site publishes four original rankings: Black City Leadership, Black-Student Academic Outcomes, Congressional Representation, and the Best Metros for Black Economic Mobility. All four obey the same constitution:
Outcomes only
We score what the office or place actually delivers — safety, income, proficiency, mobility, fiscal health — never a party label, a vote pattern, or an ideology. The question is always "what is the result," never "whose side is this."
Indexed to the national average
Every metric is indexed so 100 = the U.S. national average, making places comparable on a common scale. For the City Leadership ranking, each criterion is a 0–100 points score where 50 = the national average, rolled up across Governance, City Outcomes, and Equity & Accountability into a single Score %.
Lower-is-better, inverted
Crime, poverty, and other measures where less is better are inverted, so a safer or fairer city scores higher, not lower. The scale always points the same direction: more points means a better outcome for the people who live there.
Best ranks first
The best always ranks first. There are no gimmick orderings and no editorializing in the sort. Rank 1 is the strongest measured outcome on the page, full stop.
Anti-distortion guards
A ranking is only as honest as the data it refuses to fake. Guards prevent a place from climbing by missing its worst metric: a city must have crime data to be ranked; a district must have NAEP proficiency to be ranked. Nothing scores high because its weakest number is conveniently absent. Each ranking ships its own methodology breakdown and a downloadable CSV, so the inputs are inspectable — and contestable — line by line.
When we do not have the number, we say so. We never invent it.
This is the rule the whole project is built to protect. Where a figure is compiled and cited, it is shown. Where a figure is not yet confirmed, it is shown as data pending — visibly, in the interface — and it is never filled in with a guess, an estimate dressed as a measurement, or a plausible-sounding placeholder. A blank we are honest about is worth more than a number we cannot defend. There is no circumstance under which fabricated data is acceptable on this site.
What this method can and cannot promise.
Rigor includes admitting the edges of rigor. Public data lags reality — a metric can be the most recent official release and still describe last year. Sources revise their own figures, and when they do, we follow the revision rather than the version that read better. Indexing to a national average is a comparison, not a verdict on a place's whole character; a number is a measurement, not the entire truth of a community. We name these limits on purpose. A standard that hides its own boundaries is just confidence with better lighting.
If something here is wrong, the standard requires us to fix it in the open.
This record was built in broad daylight, and corrections happen the same way. If you can show that a figure is outdated, a source is misread, a "settled" claim is actually contested, or a trivia answer fails the dual-source and myth-screen test, bring the primary source and we will verify it against the standard on this page. Substantiated errors are corrected and logged — not quietly edited away. The goal was never to be unchallengeable. It was to be checkable, and to deserve the trust that survives the check.
“Truth in Service of Freedom.”