Numbering and Numbing the Risks
Looking at this issue reminds me of how things have changed over the years. I frequently listen to a “Nostalgic” music station—yes, even in France these exist—and they play Ray Charles and the Beach Boys, and recently the song from “Hair” about this being the “Age of Aquarius, “Love will steer the stars” and all that. Didn’t happen, did it? Instead we got computers and they’ve changed everything. The little gadgets can calculate more in a few seconds than most mathematicians could in a lifetime. We’ve therefore become ever more dependent on numbers—real and imagined.
The modern foundations of probability theory were laid down around 70 years ago, and until fairly recently only those possessed of advanced mathematical talents could use them. The computer changed that. For better or worse calculating the chances of something happening—or not happening—has become so easy that it now dominates the industry. It’s practically synonymous with risk management.
Why else would homeowners on Cape Cod be unable to obtain insurance, even if they haven’t had a hurricane in 30 years? Well the numbers say they might have one. Why has the cost of liability insurance for organizations that deal with children gone up by 30 percent to 40 percent in each of the past two years, even if they’ve had no claims? Well they might have one. Why can’t we price terrorist risks? No accurate models. Oddly enough even tort reform, which Andy Simpson brings us up to date on, is driven by it. Class actions, form shopping, which cases are settled and which ones are tried, increasingly depends on calculating where the chances of success are greatest.
The industry has gone from looking at specifics—who is the client, what do they do, where, how, what’s their track record—to putting them in boxes and looking at the numbers. Part of this is driven by the desire to “improve underwriting standards,” which has become an industry mantra, and part of it is due to the fact that with a computer it’s easy to plug in a set of numbers, and get a set of answers. You don’t have to think, and above all you don’t have to make hard decisions that you might get blamed for later. While most financial service businesses—including insurance—admonish managers and employees to “know their clients,” the fact is that they increasingly rely on whatever the computer spits out. How else to explain the current fuss over credit scoring? Oh, I know it’s a good barometer of future loss potential, etc., but it also gives agents and carriers a way to avoid actually looking at the person who wants to buy a policy in favor of looking at what the computer says.
The Age of Aquarius might have departed even before it arrived, but computers are here to stay, and in all fairness they’re just tools, and aren’t really at fault. That comes from an excessive reliance on them, and the willingness to subordinate one’s own expertise to what the numbers say. Maybe a bit more synergy could develop a system that would still take advantage of their number crunching power, but use it in partnership with old style personal involvement.