42 Fatal Crashes Missing From Transdev's Record: The FMCSA Accountability Gap Nobody Is Talking About

In April 2025, a five-year-old named Lens Joseph was run over by his own school bus in Boston. The company that operated the bus was Transdev, one of the largest private operators of public transit in the country. When WBUR and ProPublica went looking for that crash in federal safety records, they found it listed under "CITY OF BOSTON MVMB," the city's Motor Vehicle Management Bureau. Nothing in the record showed Transdev was involved.

That is not a rounding error. That is a systems failure.

The newsrooms kept looking. What they found was not one missing crash but at least 60 fatal crashes involving Transdev-operated buses from 2016 through 2025, only 18 of them listed under Transdev's name. Of the other 42, 40 were filed under the names of transit agencies, cities or a company Transdev acquired, and two were not in the federal database at all. Three of those crashes resulted in criminal charges against drivers — including the driver in the Lens Joseph case, who has pleaded not guilty to felony involuntary manslaughter.

None of those cases appear in the federal record under Transdev's name, the record the government uses to evaluate carrier safety.

On July 29, 2026, Sen. Maria Cantwell, the ranking Democrat on the Senate Committee on Commerce, Science, and Transportation, and Sen. Ed Markey, a fellow Democrat on the committee, sent a letter to the head of the Federal Motor Carrier Safety Administration demanding answers. They requested the following by August 12: a list of all inspections and crashes in FMCSA records connected to Transdev, documentation explaining why the agency's database did not capture those 42 fatal crashes, and records showing what guidance the agency provides to law enforcement on how to report which company is involved in a crash.

The senators also asked about something broader and more structurally significant: why enforcement fell. The letter says FMCSA completed about 1,400 enforcement cases against motor carriers in fiscal 2025, compared with nearly 4,000 in fiscal 2024 — a 65 percent decrease.

That last question is the one worth sitting with.

FMCSA's enforcement function is the mechanism that converts crash data into accountability. If a carrier accumulates a safety record that crosses a threshold, the agency can intervene — issue penalties, mandate audits, pull operating authority. The whole system depends on two things working simultaneously: accurate data flowing in, and enforcement actions flowing out. The senators are raising questions about both ends of that pipeline at once. The data is incomplete. The enforcement response has declined. If both are accurate, the system is broken in both directions.

The strongest counterargument here is the one Transdev makes directly, and it deserves a fair reading. The company said in a statement responding to the senators' letter that it "complies with all applicable laws and regulatory requirements in every jurisdiction," that "all incidents are thoroughly investigated," and that it has "always followed federal regulations, as prescribed." That is a specific and substantive claim: if the crashes are missing from federal records, the company is asserting that the fault lies with the reporting process — with law enforcement's failure to correctly attribute the crash to the carrier — not with Transdev itself.

That is plausible as a partial explanation. FMCSA records depend on law enforcement agencies correctly identifying and reporting which company operated a vehicle involved in a crash. That is a process with known gaps, and the senators' letter explicitly asks what guidance FMCSA provides to law enforcement on exactly this point. The data gap may be a reporting architecture problem, not deliberate concealment.

The agency has also been warned before. When FMCSA reviewed Transdev in 2016, investigators found what they called "numerous crashes" missing from the company's safety record and wrote that "this self reporting is required for accurate evaluation by FMCSA and the accurate safety record of the carrier." The agency planned a "conditional" rating; Transdev won an appeal by arguing its drivers could not have prevented many of the crashes, according to the inspection report described by WBUR and ProPublica.

But here is what the innocent-explanation version cannot account for: 42 fatal crashes. Over a decade. Across at least 16 states. A pattern of missing data is itself a data point about the adequacy of a reporting system the agency is responsible for maintaining. The senators' letter makes this precise: "Without accurate and complete data, it is impossible for FMCSA to properly target enforcement on unsafe companies, because the agency lacks the critical data needed to conclude they are unsafe."

That sentence is the whole argument. A safety regime that cannot see its own inputs cannot produce reliable outputs. It does not matter whether the data gaps are the company's fault or the system's fault. The result is the same: a carrier moves through the oversight apparatus without leaving a legible footprint.

The Department of Transportation did not answer WBUR and ProPublica's questions about the letter. It issued a statement saying the agency "shares the Senators' commitment to the safe transportation of America's children" and is "currently investigating the matter." Republican leadership on both the House and Senate transportation committees did not respond to the newsrooms' requests for comment after the original investigation published in June.

The House side did produce one relevant voice. Rep. Rick Larsen, the top Democrat on the House Transportation and Infrastructure Committee, said accurate data is the prerequisite for any functional policy. "We can't save lives if we don't have accurate data," Larsen said.

That framing is correct, and it points to what should be watched next.

August 12 was the senators' deadline for FMCSA's document production, and it has now passed. We have not found a public FMCSA response. If the agency produces records that show a coherent explanation for the missing crashes and a defensible account of the enforcement decline, the story becomes one of reporting architecture and fixable process failures. If it does not, the story becomes something else: a federal safety oversight system that cannot account for its own gaps, at the same time enforcement cases fell 65 percent.

The crashes are not the interesting part. The record is.

Who owns that record? How long does it take an incomplete one to produce a policy response? What happens when the agency responsible for collecting it cannot explain what is missing?

Those questions had an August 12 deadline. Start there.