The $91 million dollar word.
It's "if."
On May 9, 2014, a man working as a stonemason was injured in a fall from a scaffold or staging while repairing a bridge over the Charles River.
On December 1, 20251, a Massachusetts judge awarded him $90,971,6122. The sum was awarded not as damages for his physical injuries — he had already been compensated for those — but because of the way his claims had been investigated3, and subsequently handled, by the general contractor’s insurers.
Early in their investigation, an internal report from one of the insurance companies noted that a certain foreman said that the fall had not been caused by defects in the fall protection system, or the scaffold — items for which the general contractor would be responsible — but that the mason placed a single plank between the bridge and staging in violation of worksite rules, and was crossing that plank when he fell.
Based on the statement, the insurers concluded that the mason had been negligent, and his negligence caused the fall.4 The insurers declined to revisit this conclusion, even though:
Meeting notes from shortly after the accident reflected an entirely different cause, i.e., that the fall protection system became caught, causing the mason to stumble and fall from the platform;
An incident report prepared by the mason and his foreman reflected that the mason stumbled when he reached the end of his fall-protection lanyard;
A handwritten report reflected a fall from scaffolding, not a plank;
The foreman did not make the same statement when reinterviewed by counsel, nor when testifying under oath at his deposition;
The mason testified consistent with the handwritten incident reports and meeting notes.
Because the insurance companies could not let go of their favorite “fact” — that the mason placed the plank himself, and caused his own injury — they did not settle with the mason, prior to suit, or during suit, or even, as it became more and more obvious that the “plank” defense was doomed.
The general contractor lost at trial. Of course. The defense was based on a “fact” that could never come into evidence. And the insurance companies lost the subsequent suit concerning the investigation and settlement of the claim.
There are likely some interesting things to be said about incentives and culture and group decision-making, here. But Bullshit Hunting is not a place to stash and dissect concepts I vaguely remember from Social Psychology.5
It is about investigation6. Or research. And on those topics, the take-away is:
Facts are sticky.
So goddamned sticky.
“Good” facts — particularly when presented to someone with a personal, or financial, or professional interest in them — are the stickiest kind. Stickier than long-forgotten lollipops melted to crud under a booster seat. Stickier than the muck that served as a metaphor for clinical depression in what was intended, somehow, as a children’s movie.7
Sticky enough to be dangerous.
What to do, then, apart from only taking on projects where no favorable information will ever be encountered, or holding anything even slightly hopeful so close to your chest that no one will ever be hurt by it?
Stay away from “facts.”
Or, slightly more professionally put, scrupulously maintain the distinctions between “information,8” “inferences” and “facts.”
Information is the raw material, gathered in the course of research or investigation. Statements, documents, things said, positions taken.
Information would include:
Mr. X was a foreman.
Mr. X was present during the incident.
Mr. X participated in at least two meetings concerning the incident.
Mr. X gave at least three statements. The dates, circumstances, and substance of these statements were…
Inferences are conclusions reached — or that could conceivably be reached — from the information presented, if the information is credited9.
There is a way to make make it completely unmistakable that an analysis is inferential. In fact, a single word: If.
For example:
“If the stonemason’s injury occurred as described, the fall-protection system could be implicated.”
“If the foreman’s second statement is true, the stonemason may bear some degree of fault…”
“If” is also key in placing information in the context of how it may, ultimately, be used:
“If the foreman testifies at trial consistent with the second statement, he could be impeached with the accident reports and meeting notes.”
“If” can also be used to express how information can be tested: Who or what would be expected to corroborate the information? What might disprove it?
Compared to an inferential analysis phrased in “if,” even a properly sourced is/because statement is static, a cognitive dead end:
“Based on the foreman’s statement, stonemason is comparatively negligent because he placed a plank between the bridge and the scaffold.10”
Peerless Ins. Co. v. Rooney, No. 2284-CV-00652-BLS2. Decision available here.
Plus attorney’s fees, in an amount to be determined.
I am not going to get deep into the laws governing claims handling in Massachusetts, but as a very cursory statement: When a claim is reported to an insurer, the insurer has a duty to investigate. If the investigation determines that “liability is reasonably clear” — meaning it is reasonable to believe that the insurance carrier is required to pay the claimant — the insurer must try to settle the claim.
Massachusetts follows what is called a modified comparative fault regime; if the injured person is more than 50% negligent, the injured person cannot recover. If the injured person is less than 50% negligent, their award is reduced to account for their degree of fault.
The insurers reasoned here that the mason would be found more than 50% at fault for his injuries, and therefore, the contractor would not be liable.
Confirmation bias, groupthink, risk aversion, etc. I think I got an A.
Sometimes dick jokes.
Or data, or research, or documents. “Information” seems broadest, so that’s what I am using here.
I like the term “credited” rather than “credible,” because it reflects an analytical choice.
“Too conclusory,” is what would be marked in red pen, if this were an assignment in law school.