Credit Insider Guide to Credit Score Victims
Section #2: What Happened?
So what happened? What happened was that a wacky group of Minnesota statisticians at Fair Isaac Corporation (FICO®) in the late 1950s decided to have a look at how historical variables (what a consumer did in the past) correlated with future behavior (what happened later). Not surprisingly, these brainiacs quickly found that the best way to predict whether a consumer is going to become seriously delinquent is to look at how he previously handled his accounts. So the "FICO® Score" was born.
Now, all of that may sound fine, or even ingenious, but there are several serious problems afoot, and the first is scientific in nature. In statistics, results are measured by what us eggheads call the "R-value" -- whether a particular result just occurred because of dumb luck or whether it was real life stuff ("beyond chance" to use the parlance). The kicker is that you can get a VERY significant "R" when just a minority in your studied population tests positive. So, with credit scoring, if say 15% of people with scores in the 650-700 range are later found to be seriously late with their bills, then Fair Isaac will scream "Eureka!" and proclaim wild success with amazingly significant "R" but only at the expense of the other EIGHTY-FIVE PERCENT of the consumers in that group who would never even think of paying their bills late. Even worse, using this same example, if only a fifth of the delinquent 15% subset of this group actually defaults and never pays back the borrowed money (and the actual number is probably far less), then guess who in this 650-700 range gets blamed for that? Yep, the other 97%.
Next, consider what happens to the rest of the innocents in the 650-700 range. They simply bleed higher interest payments for mortgages and car loans and credit cards. They provide a terrific excuse for Citibank and Chase and Capital One to advertise wonderful interest rates but then actually offer much less favorable (and much more profitable) terms later on. Such consumers literally pay and pay and pay, all because of a cockamamie number that can predict statistically "significant" trends for large groups but so often tragically fails at the level of the individual person.
Another issue is strictly psychological. Numbers are seductive, and people love to be seduced. They want to know their height, their weight, their IQ, how high they can jump measured in centimeters, how many paint-balls can be hurled at an opponent, you-name-it. If a single number can provide the "right" answer, then we clamor for it. And of course, when it comes to considering applications for credit cards, just imagine how much easier it is now to simply use FICO® Scores! Human beings are no longer required to read through credit reports to forge intelligent decisions. Instead, computers can electronically procure the FICO® Scores and render instant judgments. Sadly, the problem of false positives (all those innocent folks in the 650-700 range mentioned above who are automatically judged by the worst behavior of their group's bad apples) is ignored.
Basically, with credit scoring, our society has traded the human problem of rampant subjectivity, where anything is ripe for discussion, for the mechanistic problem of impersonal automation, where nothing about a consumer's individual situation is considered. Clearly the first problem needed addressing, but couldn't smart people devise a better solution than this?