Home → Methodology → Corrections
Our methodology page says nothing gets quietly edited. This is that page. It includes corrections to our own published work, which are the ones that matter.
Anyone can promise accuracy. The test is what happens when you get something wrong — whether you correct it in public with a date on it, or fix it silently and hope nobody noticed.
Three of the entries below are errors we published. One of them was live on 24 pages.
We published Louisiana’s rules without picking up Act 510 of 2025, which materially changed them. The page was roughly a year out of date on the two figures a borrower would actually use.
The page now leads with the change and cites OFI directly. Louisiana →
We ran a badge reading “Law unchanged since 27 July 2026” on 24 jurisdiction pages. That date was when we verified the page, not when the law last changed. Nebraska’s law changed in 2020, Hawaii’s in 2022, Illinois’s in 2021, Louisiana’s in 2025.
It was a false statement of fact, repeated across half the site.
The irony is that we built that badge to stop verification dates looking stale. The fix we reached for replaced a stale-looking true claim with a fresh-looking false one, which is worse. A date that ages honestly beats a date that lies.
Every jurisdiction page now carries two claims, both true and each about a different thing: Verified against the state regulator (a person traced every figure) and Sources checked (an automated daily link and content check). Where a law’s change date is actually known, it appears in the page body with a citation.
An index-building script regenerated state-california.html and state-texas.html every time it ran. It ran around twenty times. Each run replaced those pages with an older version that had no inline source links.
The result: the two pages carrying our strongest verification claim were the two with no evidence on them. The badge survived. The citations did not.
Found by an automated audit, not by reading the pages — they looked finished. Both have been rebuilt from DFPI and OCCC directly, and the script no longer touches them. California → · Texas →
Several consumer guides describe Minnesota as having adopted a 36% cap effective January 2024. We were on course to repeat it.
The distinction has practical bite. A borrower told the ceiling is 36% would conclude a 45% loan is unlawful. It is not — but it does carry an underwriting duty the lender may not have discharged, which is more useful to know. Minnesota →
We found published guides disagreeing about whether Nevada has a rate cap, and published a note saying we could not resolve it. That was the right call at the time but the wrong stopping point — the answer was available.
The Supreme Court of Nevada states it directly: Nevada does not have a usury law, so there is no statutory cap on interest rates. The guides listing a 36% cap are simply wrong.
The same case also decided something more useful: the 25%-of-income limit includes principal plus interest and fees, not just the amount borrowed. Nevada →
Both pages originally cited a legislative research compilation and secondary summaries rather than the statute. We flagged that on the pages at the time, but flagging a weak source is not the same as fixing it.
Rebuilt from the Official Code of Georgia and, for North Carolina, the Attorney General’s Senate testimony and the former Commissioner of Banks’ own published account. Both pages are substantially stronger as a result — the Georgia rebuild surfaced that the state codified the true-lender test in 2004, seventeen years before anyone else. Georgia → · North Carolina →
Not corrections to our work. Recorded because they explain why this site cites statutes and regulators rather than other websites, and because anyone relying on these sources should know.
A state legislative research office compilation of payday lending regulation — a government source, and a good one when written — is now materially wrong about Colorado, Ohio and Nebraska, purely because those states changed their laws after it was published.
It gives Colorado’s pre-Proposition 111 rules, describes Ohio before HB 123, and lists Nebraska at $15 per $100 — roughly 460% — when Nebraska capped rates at 36% by ballot initiative in 2020 with 83% support.
A 30% error rate, from age alone. We nearly used it to build ten state pages in one sitting. Had we done so, three would have been confidently wrong, and the Nebraska error would have told borrowers they could lawfully be charged twelve times what the law now permits.
Working through 52 jurisdictions, the same states appear with contradictory figures across published compilations:
Connecticut listed at 30.3% by one source and 12% by another. Georgia at 16% by one and 60% by another. Nevada at 36% by one and no cap by several. North Carolina described simultaneously as a full ban and as having fee caps and term rules.
We verified Connecticut, Georgia and Nevada directly and found all of those figures misleading in different ways — some describing thresholds rather than caps, some describing repealed law, some simply wrong.
This is the reason every figure on this site carries a marker linking to the regulator, the statute or a court. Click any of them.
Email [email protected]. We will check it against the primary source, correct it if we have it wrong, and add an entry to this page with the date.
That applies to lenders too. If a lender page misstates something, send the correction with a source and we will publish it — and we will publish the fact that you asked. What we will not do is take it down.