On January 17 I&T posted a story about an exchange that took place at the Property/Casualty Joint Industry Forum between State Farm’s CEO, Edward Rust Jr., and an industry analyst, Brian Sullivan. Mr. Sullivan said, “It's impossible for anyone to look at the data and say there won't be fewer accidents than before. The technology is getting better and drivers are getting safer. I think this business is shrinking: Fewer accidents means fewer exposure." And Mr. Rust responded, "I don't see the risk being mitigated so much that the premium falls significantly," Rust added. "The cost to repair a vehicle that has been in an accident is much greater. It's not your Grandpa's Olds." I will judiciously say that both Mr. Sullivan and Mr. Rust are correct—but the real question is the timeframe during which each of them is correct. This year and next year and maybe the year after, there won’t be much technology-driven reduction in auto losses (and associated drops in premium). But inexorably collision avoidance technology is going to get better, and even more importantly, it will become more pervasive among the vehicles on the road. And while insured losses depends on severity (i.e. the cost to repair partial losses or replace total losses), it also depends on frequency. As collision avoidance technology (and automated traffic law enforcement, and yes eventually driverless cars) advances, frequency will drop. And in all likelihood severity will also drop—for example when an automatic braking system reduces the speed at impact from 15 mph to 5 mph. So losses will drop and insurance premiums will follow. The big questions are how much and how soon.