Loading…
Everyone described in individual cases is accused, not convicted, and this site does not name private individuals. Each post is linked so you can read it in full. We have no connection to the account, and no view on its motives; we check claims, not people.
What kinds of claims are in the feed?
Posts by type
Only some of those can be checked against data: statistical claims (a number or trend) and the pattern behind the case reports. Policy commentary (spending, surveillance, immigration, and so on) is argument rather than fact and is outside this site’s scope.
Claim check: statistical claims we could test
Homicide
Specific crimes
Policing, traffic and budgets
Statistical claims we could not test
These posts make factual claims, but they need data this site does not have. “Not tested” does not mean “false.”
The case reports: what they show, and what they can’t
Case reports are the account’s signature content: a specific person, a record of prior cases, a release, and a new alleged offense. They are vivid, and some describe real failures. Whether they describe a pattern is a different question.
Who the post faults
Case reports by the decision the post blames. A post that implies blame without stating it is counted as “neither stated.”
Prior cases, where a post gives a number
One bar per post that states a count (cases, charges or convictions; the units vary).
All case reports, de-identified
Offense and release status are as stated in the post; we have not independently verified court records. Each row links to the post.
How often are released defendants arrested again?
The whole “revolving door” argument turns on one number nobody has published for Travis County: of the people released before trial, how many are arrested for something serious while their case is pending? Research from other places gives a sense of scale. No source reports this rate for Texas or Travis County.
What studies elsewhere found
Definitions differ, so read the scope column before comparing rows.
- In large urban systems, roughly one in six released felony defendants is rearrested for something before their case ends, and a few percent for a violent offense. So a feed of cases where a released person was arrested again will always have material, in any city.
- That does not show Travis County is typical. Without Travis’s own rate there’s no way to tell whether it is higher or lower than these benchmarks.
- Texas does collect a count of cases where a defendant “committed an offense” while on bail (statewide, FY2025: about 9,700 district-court cases), but with no denominator, and Travis reports zero.
Reading the cases fairly
- APD’s public arrest file has no person identifier. It records the case number, charge, and the arrestee’s race and gender, but nothing that links one person’s arrests together. “Arrested N times” cannot be counted from open data.
- The state’s court report has a field for offenses committed while on bail, but Travis County’s entries for it are zero in every year, which almost certainly means “not tracked,” not “none.”
- Case-level court records exist (the Travis County District Clerk holds them) but are not published as a bulk, person-linked dataset.
What would settle it: for every person arrested for a violent felony in a given year, follow their prior arrests, release decisions (and who made them), and what happened next. Until that is public, claims in either direction rest on examples.
Questions worth asking of any claim like this
- Compared with what? A rise or fall means little without a baseline year and a national or peer comparison.
- Counts or rates? Austin’s population grew about 25% between 2010 and 2025.
- Cases, arrests, or convictions? “18 cases” is not “18 convictions”; cases include misdemeanors, dismissals, and charges bundled together.
- Who decided? Bond is set by magistrates and judges, probation is revoked or continued by judges, and charging and plea decisions belong to prosecutors. Blame depends on which decision failed.
- How were the examples chosen? Vivid cases are chosen because they are vivid. They show that something can go wrong, not how often.