> But the fact that they can ease up on background checks and see an immediate large bulk of car thefts is damning evidence that the background checks were truly having a real impact on preventing theft.

No, it isn't “damning evidence”, it's at best a suggestive indication which is easily consistent with other alternatives.

In fact, there is extremely compelling evidence that the old background check policy was not what was protecting them in the article: “By midweek the company suspended service in Chicago altogether, an acknowledgment that it couldn’t figure out how to distinguish legitimate customers from the group of thieves.” The obvious first guess of how to distinguish legitimate customers from thieves would be exactly the method they had been using right before the event that has supposedly prevented similar evebts—suspend every new account that was checked until they agree to a background check and it is completed and, if the old method was really working, everything would be fine. This is so obvious that if they just threw up their hands and couldn't figure it out, it means either they didn't believe the system was doing it or they didn't believe the problem in Chicago represented a general problem that should be protected against, just a local aberration; either way, it doesn't really seem consistent with the idea that even they believe there is some widespread general problem that the background checks were preventing.