We're not talking about car theft in general, but about the specific crime waves that occurred after the rollout of the less than secure Kias in the US and the Kias with the proper security measures in Canada.
If it did, that would point to a US and Canada crime trend correlation. If not, then you can't just say that the one static variable, city/county level policy and the independent variable, immobilizers, are the only factors.
You have different criminal populations, societal values, amounts of government aid, rehabilitation programs, etc that all play into the analysis.