How Credit Scoring Came into Being
The question remains: How did one little number come to have such an outsized effect on our lives?
Credit scoring has been in widespread use by lenders for several decades. By the end of the 1970s, most major lenders used some kind of credit-scoring formulas to decide whether to accept or reject applications.
Many were introduced to credit scoring by two pioneers in the field: engineer Bill Fair and mathematician Earl Isaac, who founded the firm Fair Isaac in 1956. Over the years, the pair convinced lenders that mathematical formulas could do a better job of predicting whether an applicant would default than even the most experienced loan officers.
A formula wasn’t as subject to human whims and biases. It wouldn’t turn down a potentially good credit risk because the applicant was the “wrong” race, religion, or gender, and it wouldn’t accept a bad risk because the applicant was a friend.
Credit scoring, aided by ever more powerful computers, was also fast. Lending decisions could be made in a matter of minutes, rather than days or weeks.
Early on, each company had its own credit-scoring formula, tailored to the amount of risk it wanted to take, its history with various types of borrowers, and the kind of people it attracted as customers. The factors that fed into the formula varied, but many took into account the applicant’s income, occupation, length of time with an employer, length of time at an address, and some of the information available on his or her credit report, such as the longest time that a payment was ever overdue.
These calculations took place behind the scenes, invisible to the consumer and understood by a relatively small number of experts and loan executives.
The cost to develop and implement these custom formulas was—and still is—considerable. It was not unusual to spend $100,000 or more and take 12 months just to set one up. In addition, not every creditor had a big enough database to work with, especially if the company wanted to branch out into a new line of lending. A credit card lender that wanted to start offering car loans, for example, might find that its database couldn’t adequately predict risk in vehicle lending.
That led to credit scores that are based on the biggest lending databases of all—those that are held at the major credit bureaus, which include Equifax, Experian, and TransUnion. Fair Isaac developed the first credit bureau-based scoring system in the mid-1980s, and the idea quickly caught on.
Instead of basing their calculations on any single lender’s experience, this type of scoring factored in the behavior of literally millions of borrowers. The model looked for patterns of behavior that indicated a borrower might default, as well as patterns that indicated a borrower was likely to pay as agreed. The score evaluated the consumer’s history of paying bills, the number and type of credit accounts, how much available credit the customer was using, and other factors.
This credit-scoring model was useful for more than just accepting or rejecting applicants. Some lenders decided to accept higher-risk clients but to charge them more to compensate for the greater chance that they might default. Lenders also used scores to screen vast numbers of borrowers to find potential future customers. Instead of waiting for people to apply, credit card companies and other lenders could send out reams of preapproved offers to likely prospects.