The Federal Transit Administration issued two related special directives on September 24. Special Directive 26-3 requires the Chicago Transit Authority to correct deficiencies in its internal safety-event investigation process and near-miss tracking. Special Directive 26-2 requires the Illinois Department of Transportation to address deficiencies in its oversight of safety-event investigations at CTA's rail system.

The pairing is the story. One directive reaches the operating agency's investigation and learning process. The other reaches the state oversight function that must evaluate whether that process is adequate. FTA's directive index explains that the agency may use a special directive when it identifies a violation or pattern, a substantial safety risk, or a public-interest need for immediate compliance.

A completed report is not the same as a complete investigation

An investigation program has to do more than close files. It should preserve evidence, establish a defensible sequence, examine human and technical factors, identify organizational contributors, and connect findings to corrective action. A near-miss program adds another demand: employees must recognize reportable events, have a practical way to report them, and trust that reports will be analyzed without becoming a shortcut to blame.

When near misses are recorded separately by operations, safety, maintenance or control, patterns can remain invisible. The same problem appears when investigation quality is measured only by timeliness. A fast report that does not test causes or verify corrective action can create paperwork without reducing risk.

What agencies and overseers can audit now

• Definitions: Are safety events and near misses defined consistently across departments and contractors?
• Evidence: Do investigators preserve video, data logs, radio traffic, work records and witness information before it is lost?
• Competency: Are investigators trained and periodically assessed for interviewing, causal analysis and report writing?
• Independence: Can the investigation challenge operating assumptions and supervisory decisions?
• Corrective action: Does every accepted action have an owner, due date, evidence standard and effectiveness review?
• Oversight depth: Does the state overseer sample investigation quality and near-miss trends rather than reviewing only completion dates?

The public FTA summary pages do not reproduce every requirement in the directive PDFs. CTA, IDOT and other agencies should use the official directives—not a secondary description—to determine obligations and deadlines.

For the wider industry, the lesson is immediate. Investigation and oversight are separate safety defenses, but they depend on the same evidence. If near misses are not captured, if causal analysis is thin, or if corrective actions are not tested, both defenses weaken together.