The Financial Services Regulatory Authority of Ontario (FSRA) is trying to determine how auto insurance rating factors affect vulnerable groups given the general lack of data, a senior executive at the regulator said during an industry event Thursday.
Brian Sullivan, editor and owner at Risk Information Inc., asked panellists at the 2023 FSRA Exchange event how they go about the “delicate challenge” of analyzing factors such as race on rating.
“The short answer is very carefully,” Tim Bzowey, FSRA’s executive vice president of auto/insurance products, said at the 2023 FSRA Exchange event. “The slightly longer answer is that may not be any more illuminating.”
Bzowey reported that there are currently several studies underway in Ontario to try to first identify vulnerable groups. “Nobody’s ever done that work,” he said during the Auto Rate & Underwriting Regulation Reform: Fairness in Auto Insurance session.
“There’s not a whole lot of data. So, I think we’ve got to start there before we can start making any decisions about what are the kinds of things we would need to do…”
Race data is not information that the insurance industry typically collects or wants to collect for numerous reasons, added Roosevelt C. Mosley Jr., principal and consulting actuary with Pinnacle Actuarial Resources and president of the Casualty Actuarial Society. But while analyzing racial data as it relates to auto rates is challenging, it’s not impossible, Mosley said.
“We’ve gotten to the point, in the U.S. especially, [where] ‘I don’t have the data’ is not going to be a sufficient answer going forward,” he said. “We’re taking a step in the right direction. The answer is not going to be perfect but it’s going to continue to get better.”
Mosley said a few methodologies have tried to approximate what some of the sensitive information looks like — whether on an individual variable level or a more holistic level — and apply those methods to insurance data.
“Essentially what we’re finding is while there might be some areas we might want to poke out a little mire, there really isn’t a whole lot to see here,” he said. “Now, I’m not saying that’s going to be the case across industry.
“And those approaches admittedly… come with some level of error that we’re going to have to accept. But someone said earlier, ‘Let’s not necessarily let perfect be the enemy of good.’”
Sullivan said his organization held a forum last year to discuss whether insurance rates were accurately matched to loss costs, or if race or another protected characteristic was driving up prices, either due to intentional bias or unintentional error. They also examined social and historical failures in society that may lead to, for example, somebody living in a high-crime neighbourhood.
“I think this is part of the point of some of the research that we’re doing to try to get underneath and see if we can… come up with an answer,” Bzowey said, adding the social aspect question is difficult to assess.
“So, I’ll pick a reasonably non-contentious one,” he said. “Let’s just say for a moment that it was in Ontario’s interest that younger drivers got a better deal than they deserve based on the risk they present. That’s a policy decision. That’s one an entire industry could get behind if everybody had to play by the same rules.”
The industry is not there yet, but it’s “a lot better than the unintentional cost subsidization that goes on all over the place,” Bzowey said. “Or perhaps some of the things the industry complains about where you can’t actually get the right rate because of prohibitions or restrictions in the current rate regulation regime.
“So, I’d much rather we take the restrictions off and allow you to get comfort. You can get the right price, and then…on top of that [we could decide], ‘here are the groups or the people that we think from a policy perspective need more favourable terms.’”
Sullivan said once you start to play around with pure rating, “the law of unintended consequences comes and bites you.”
Mosley said society and industry agree pricing based on race or income, even if predictive, cannot be used.
“Let’s say I could find a characteristic or series of acceptable characteristics that would get me to the same place as using one of those prohibited characteristics… is that okay?” Mosley asked. “The wheels start turning a bit, but I think at some point, most people get to the place where they think the answer is still no because regardless of how I get there, I’m getting to a place where I say I won’t as a society.”
The question is “how close to that line can I get… before I scorch myself?” Mosley asked. “And that’s the part where regulators, where policymakers, are going to have to answer that question. I as a math guy can tell you how close you are, it’s going to be someone else to determine whether that’s acceptable or not.”
Feature image by iStock.com/LumiNola