POSNER, Circuit Judge.
Cynthia Madden was killed when her car was struck by the defendant's train at a railroad crossing in Illinois. Her estate sued the railroad, basing jurisdiction on diversity of citizenship and claiming that her death had been caused by the railroad's negligence — specifically that the railroad crossing gates had descended, the warning lights at the crossing had begun flashing, and the locomotive horn had been blown, all fewer than 20 seconds (fewer than 15 seconds in the case of the horn) before the train reached the crossing, in violation of applicable federal safety regulations. 49 C.F.R. §§ 222.21(b)(2), 234.225.
A state may provide a remedy for negligence resulting from violation of federal railroad safety regulations. 49 U.S.C. § 20106(b)(1)(A). Illinois law, on which this suit is based, provides such a remedy, by making the violation of a safety regulation intended for the protection of the class of persons to whom the plaintiff belongs (and who received the type of injury the regulation was intended to prevent) prima facie evidence of negligence. E.g., Abbasi ex rel. Abbasi v. Paraskevoulakos, 187 Ill.2d 386, 240 Ill.Dec. 700, 718 N.E.2d 181, 185-86 (1999); Kalata v. Anheuser-Busch Cos., 144 Ill.2d 425, 163 Ill.Dec. 502, 581 N.E.2d 656, 661 (1991); Davis v. Marathon Oil Co., 64 Ill.2d 380, 1 Ill.Dec. 93, 356 N.E.2d 93, 97 (1976); Cuyler v. United States, 362 F.3d 949, 952 (7th Cir.2004) (Illinois law). The magistrate judge, presiding over the case by consent of the parties, ruled that the plaintiff had not made a prima facie case of negligence, and so he granted summary judgment in favor of the railroad, precipitating this appeal.
The accident occurred at 9:40 p.m. in May 2007 in the small town of Colona near Moline in northwestern Illinois. Madden had almost reached the crossing when her car stalled. She restarted it and drove onto the crossing, whereupon the car stalled again. Just then the crossing gates began to descend, the warning lights at the crossing began flashing, and the crossing bells sounded — unmistakable indications of an approaching train. But rather than rush out of her car Madden tried to restart it — at least that was the conjecture of an eyewitness, the driver of one of two cars stopped at the crossing behind Madden's car. Noticing that Madden wasn't getting out of her car, the driver got out of his car and started walking toward the crossing. He saw Madden open her car door when the train was only 45 to 50 yards from the crossing, get out, and start to run. The train struck the car, pushing it against her, causing injuries that proved fatal.
An event recorder alongside the tracks at the crossing recorded that the warning lights had begun flashing, the bells sounding, and the gates descending, all half a minute before the train entered the crossing. The interval was consistent with the witness's affidavit. He had watched Madden try to restart her car, had gotten out of his own car, and had walked toward her at the crossing — and all this after the warning lights had begun flashing. The event recorder had been inspected and found in working order 19 days before the accident. (There was another event recorder, in the locomotive, but because the locomotive had been built in 1997 its recorder wasn't required to, and did not, record when the horn sounded as the train approached the crossing. 49 C.F.R. § 229.135(b)(3).)
Three grown children of Madden — Michael, Jennifer, and Carrie Dillard — visited the scene of the accident either a few days, or a week or two (their testimony was in conflict), after the accident. They decided to time the interval between when the horn of an approaching locomotive is first heard and when the locomotive reaches the crossing. But the testimony they gave at their depositions about their efforts to determine the interval was inconsistent and confusing. Michael testified that he and his two sisters were at the accident scene timing trains for 20 or 30 minutes, but Jennifer testified that they were there for an hour and a half. And while she testified that they timed three different trains, Michael and Carrie mentioned only one. Jennifer further testified that they had conducted the tests at noon. But Michael testified that they had conducted them at dusk. He also testified that all three had started timing the approach of a train when they first heard the locomotive's horn, but Carrie said they didn't start timing until they could see the train coming. Jennifer testified that Carrie did not time a train, and that the train that she and Michael timed had reached the crossing between 12 and 13 seconds after they heard the horn. Michael testified that the train had arrived 13 to 14 seconds after he heard its whistle, but contradicted Jennifer's and his own testimony that all three had timed the train's approach, stating that his sisters had instead been "watching mine [his watch] with me for sure." But Carrie testified that it was 13 to 14 seconds after they saw (not heard) the train that it reached the crossing. Michael was unsure which track he timed. There are two parallel tracks at the crossing, and they are about 100 feet apart and each track has separate lights and gates.
He testified that the warning lights did not begin flashing until 2 or 3 seconds before the train reached the crossing, but Jennifer testified that they began flashing earlier, when the train's horn could first be heard, which was 12 to 13 seconds before the train's arrival at the crossing. Michael testified, contrary to the affidavits of both drivers who had seen the accident, that the gates came down "as the train was coming through" the crossing. That, like the failure of the warning lights to flash until the locomotive was within 2 or 3 seconds of zooming through the crossing, would be an incredible safety violation. Neither sister corroborated Michael's testimony.
Had Michael used a video camera instead of a watch to time the trains, he might have had better evidence of the interval between when the gates, lights, etc. signaled the train's approach, and when it arrived at the crossing. (Better evidence, but not irrefutable, given the possibility of editing a video.) As it is, the testimony of the three amateur "testers" was too garbled to prove anything — which is doubtless why the plaintiff's briefs are devoted almost entirely to trying to rehabilitate her two expert witnesses, railroad engineers all of whose proposed evidence the magistrate judge excluded after hearing their testimony at the Daubert hearing. The expert evidence that they wanted to give was, as we'll see shortly, hopelessly inadequate. Anyway their conclusion that the plaintiff's decedent hadn't had the full warning that the regulations required rested mainly not on the technical evidence that they planned to give but on their belief that the children had testified accurately. Vouching for a lay witness is not expert testimony, e.g., United States v. Vest, 116 F.3d 1179, 1185 (7th Cir.1997); United States v. Benson, 941 F.2d 598, 604-05 (7th Cir.1991); Nimely v. City of New York, 414 F.3d 381, 397-98 (2d Cir.2005); an engineer's training and experience do not make him abler than a juror to evaluate the consistency of lay testimony. For an expert to offer such an evaluation is to wrap the lay witness in the expert's prestige and authority — a disreputable tactic that the plaintiff's lawyer should not have countenanced.
An expert could properly testify that if Michael's testimony that the crossing gates did not begin to descend or the warning lights to flash before the train was within seconds of reaching the crossing was true, then the event recorder beside the tracks, which contradicted that testimony, must have been defective. An expert could also properly testify that what Michael or Jennifer or Carrie claimed to have seen happen could as a technical matter have occurred — but not that any of their testimony was accurate.
The attempt at vouching in this case, besides being impermissible, reflects poorly on the experts' judgment, for no one who had heard the children's testimony or read a transcript of it could think it worthy of belief, since it was a tissue of contradictions.
There is more wrong with the expert evidence. One of the experts (Paul Bodnar) testified that the removal of the locomotive horn after the accident for testing (rather than the horn's being tested while it was on the locomotive) was suspicious; maybe the railroad had taken the horn to a place where it could be repaired without a railroad worker's noticing. Much to that expert's surprise, in fact the horn was still on the locomotive when it was tested. Bodnar also testified mistakenly that the fact that the locomotive's recorder had not recorded the sounding of the horn violated federal regulations.
The other expert (James Sottile) refused to accept the evidence from the trackside event recorder that there had been a 30-second warning period before the collision with Madden's car. His reason was that the time shown on the recorder was standard time rather than daylight saving time even though daylight saving time had begun more than two months before the accident. But the fact that the recorder displayed the wrong hour did not affect the display that showed the 30-second warning period. He testified that the period shown on the recorder was from "an hour prior to the accident." No; undoubtedly it was the warning period triggered by the train that hit Madden's car. The accident took place an hour before the time recorded on the recorder only in the sense that daylight saving time dates time an hour earlier than standard time. The expert testified that the event recorder could have been tampered with, but admitted that he had no evidence of that.
The plaintiff's lawyer points out that three clocks recorded the time of the accident, and the clock in the event recorder, the clock in the dispatcher's office, and the clock in the locomotive all showed different times even after correction for the fact that the hour setting in the event recorder had not been changed to daylight saving time. But what mattered was not whether the clocks were properly synchronized, but the intervals between the activation of the crossing safety equipment and the sounding of the locomotive's horn, and the train's arrival at the crossing.
The lawyer insists that the railroad may have tampered with the event recorder or misrepresented the results of the inspection of the locomotive's horn after the accident. He asks us to infer that since the hour setting on the recorder had not been changed to daylight saving time maybe something else wasn't set properly, though there is no evidence of that. He failed to depose any employee of the railroad in an attempt to determine whether there was destruction of or tampering with evidence of negligence. (Willful destruction of such evidence could support an inference of negligence. E.g., Neace v. Laimans, 951 F.2d 139, 141-42 (7th Cir.1991) (Illinois law).)
In his briefs and at the oral argument, both heavily focused on the experts' suspicion of the defendant's records and testing procedures, the plaintiff's lawyer managed to lose sight of the fact that before he could gain any purchase by undermining the defendant's evidence, he had to present enough evidence to create a prima facie case — enough evidence that, were it all the evidence in the case, would allow a reasonable jury to conclude that it was more likely than not that the defendant had violated the law. See, e.g., Carroll v. Lynch, 698 F.3d 561, 564 (7th Cir.2012). Neither the children's testimony, reflecting their incompetent efforts to reconstruct the accident, nor the experts' worthless evidence, nor both bodies of evidence combined (0 + 0 = 0), would enable a reasonable jury to infer negligence on the part of the railroad. The judgment is therefore
AFFIRMED.
