Peeling the credit score onion
By Robert Gellman
DM News
December 8th, 2006
Most people are familiar with credit scoring. A credit score is a method of assessing risk used by a lender considering extending credit to an individual. A well-designed credit score is objective and based on actuarially sound standards.
The brand-name credit score from Fair Isaac Corp. is known generally as a FICO score. This score appears to have the lion’s share of the market. See If used properly by a lender, a FICO score should result in faster credit decisions and fairer standards.
The FICO score derives solely from information from credit reports. Factors are payment history, amounts owed and length of credit history. Wealth and income are not considered, presumably because reliable information is unavailable.
The FICO score is not a simple or single number. The three major national credit bureaus may have different information on the same individual, so each bureau’s report can produce a different score. The differences potentially add an element of subjectivity and manipulation to a lender’s decisions. Will the lender use the highest score, lowest score, middle score or average score? It can make a huge difference.
Any credit bureau, merchant or lender can develop its own credit scoring system, and some have. The three major credit bureaus offer to sell consumers a “Vantage” score based on standards that differ from FICO’s. Of course, if lenders don’t rely on the Vantage score, the score is not relevant to a consumer seeking credit. A proliferation of meaningless scoring systems with inadequate disclosures may only exploit consumer fears and pocketbooks. This could be a fertile area for crooks.
Consumers can readily learn about credit scores and take steps to improve their scores. A good reference book on the topic is “Credit Scores & Credit Reports”
But we are not finished peeling the credit score onion. Insurers often use credit scores to assess risk in auto and homeowners insurance. No one knows why people with low credit scores are more likely to make insurance claims, but studies appear to support the connection.
The result is that credit scores determine insurance prices. Partly because of the lack of an understandable link between a credit score and driving ability, the use of credit scores for insurance is controversial. Many states regulate how insurers use credit scores, though the regulation is often light.
So a credit score is also an insurance score. Other types of scores, not derived from credit scores, are also in use. Scoring systems assess the collectability of debts. Scores predict the likelihood that a consumer will seek bankruptcy protection. ID Analytics is one of several companies that offers an ID score to calculate the risk associated with an identity, with a goal of distinguishing fraudsters from legitimate customers.
Presumably, if someone can find a correlation between credit scores and other consumer characteristics, the credit score might be used in other ways. Suppose someone finds a statistical correlation between a credit score and health expenditures. The credit score might be a proxy for a health cost score. That possibility is pure speculation, but some health scores are already in use. Medicare has used something called a resource utilization group score to assess residents in skilled nursing facilities. The extent to which health scores are in use in the health or employment sector is unknown, but this is a likely area of scoring expansion. Imagine how a single number that assesses your health costs might be used and abused.
I predict greater use of scoring by businesses, and maybe governments, to make decisions about consumers in the next decade. The existing system of renting lists selected by crude characteristics could be replaced by scores built from the ground up for each household based on actual consumption data and good statistics. Targeted marketing might be more nuanced than ever. The list business could change dramatically.
Is more scoring good or bad? I worry about the effect on privacy of increased aggregation of personal data needed to fuel consumption scoring. I worry about scoring providing cover for unfair or discriminatory practices. I worry about the use and misuse of scoring by government to decide who gets on an airplane or for other purposes. I worry about health scores affecting employment practices. I worry that Americans might be pressured to live for their scores, lest they be unable to achieve fair treatment in the marketplace.
I worry especially about a lack of transparency, accountability and recourse for consumers. If scoring creates problems for consumers, I predict that a Fair Consumer Scoring Act will pass in California within a decade, probably over the violent objection of the yet-to-be-formed Consumer Scoring Trade Association. A federal law will follow shortly thereafter.
As for worrying about the effect of consumption scores on the list business, I leave this as an exercise for DM News readers.