The PeopleRank (or Whuffie or Karma or Facebook Points or eBay rating or whatever) is coming, no doubt.
This is one of those major shifts to society that one day will just be there and change everything.
Are you ready for it?
Are we?
This is one of those major shifts to society that one day will just be there and change everything.
Are you ready for it?
Are we?
I think a small team of folks with CS backgrounds plus an insurance company's claim data (or a bank's loss data) for a year plus a Hadoop cluster would make FICO obsolete, at least until they were regulated out of existence for being too darned efficient at what they do. (The basic theory is "Birds of a feather flock together": if most of your friends are losers, it is probably net-beneficial to the business to assume you are a loser, if most of our friends are solid debt-paying responsible citizens, it is probably net-beneficial to the business to assume you are a solid debt-paying responsible citizen.)
I think this would be so effective (prices risk appropriately, far cheaper to maintain than the current reporting infrastructure since folks helpfully update the social graph for you) you'd almost have to illegalize it. (Let me hum a few bars: "disparate impact".)
On the other hand, such an inefficient classification mechanism ought to create sufficient surplus for profit for those specialized enough to look for it. But there are other systemic inefficiencies that might prevent those niches from being profitable enough.
There already are insurance companies that only insure people with a university degree. Their rates are lower than those of other companies. They aren't prejudiced and they do not discriminate: they estimate risk based on solid information about a person and the mere fact that they can stay in business suggests there is indeed a good correlation between their estimates risk and the actual risk.
If they could determine who has many university educated friends, then they would perhaps want to extend their service to those people, because they have determined that those people share the characteristics of university educated people that cause the insured risks to be smaller. (Assuming these reasons that are independent of the reasons for not doing/finishing a university education, etc.)
If I know the chance of crashing my car and losing a leg is exactly 1 / 50000 (per year), and if I can insure against that at roughly that rate, I would still get insurance, because $20 bucks a year doesn't affect my life, but a million will help a lot when I do lose that leg.
In this case risk is still pooled and the insurance company can still function because the misfortune of the individual is paid for by the more fortunate.
There is still a need for insurance companies exist in a world with perfect information. Of course the insurance rates would shift. For example, when in a large group of people a few have a high chance of an expensive disease the insurance company wouldn't give the same rate to everybody. You'd still want to get insurance, though.
I am pretty sure that picking only people with a university degree is called discrimination.
Now, is it a good discrimination is a whole another matter.
Actually, it's not. When insurance is correctly priced, the payout for a group is close to its premiums, no matter how the group is defined. (Yes, including factors that weren't considered by the insurance company.)
If you don't price that way, some group is getting a subsidy and another group is paying one. An insurance company that figures out how to identify the latter can offer its members better rates than a company that doesn't. Or, if the smarter company can figure out how to avoid the former, it can have higher profits (which attracts competitors).
Insurance companies like to make money. People like to pay the minimum for their insurance. These two facts have economic consequences.
It sounds like you haven't quite decided whether this would be less accurate than FICO (maybe! But there's no reason to be outraged that some random person's prejudice is costing them money) or more accurate than FICO (in which case your argument seems to be "Yes, this kind of prejudice is accurate--or, at least, more accurate than the alternative. But that makes me feel bad, so misallocate capital and make more bad loans.")
The world economy just went through a couple very bad years caused by making wishful-thinking loans instead of unpleasant-truth loans. I hope that doesn't have to happen again.