The numbers are a little too good ...
In the project to build a new Bay Bridge, there are more problems. Beyond the allegations that a bunch of the welds are defective (as if that were not problem enough), there are concerns about the safety record for workers on the bridge. Specifically, the worry is that the safety record is too good. The concern is that, for the number of people working on a project like this, statistically, there should have been more injuries than were. Because, you know, accidents happen. When many fewer accidents than you'd expect are reported, the agency to which you're reporting them gets ... suspicious. It's not impossible that you could have so few injuries. Maybe you've got a really lucky work site. But maybe ... you're underreporting the injuries.
And why would you underreport injuries? Could it be because you've actually incurred more injuries than you ought to (because your work site is not as safe as it should be)?
I find it intriguing that falsifiers are often like comic book bad guys -- they get caught because of overreaching. They get greedy. In making up the numbers, they paint too rosy a picture of the reality they're making up.
And those meddling kids always manage to figure it out!