„Scoring procedures“ and „score values“ have become indispensable in the business world. In most cases, they are used to assess creditworthiness. And credit is used much more frequently than many people realize.

Credit is more common than initially assumed

Anyone who takes out an installment loan with a bank is well aware of the fact that they will only receive the loan from the bank if they are creditworthy. But if he agrees to pay in installments with a retailer, the same applies. And anyone who signs a cell phone contract also takes out a loan. After all, he always pays the bill after the fact.

Creditworthiness is an important aspect

Anyone granting credit faces the problem of reliably assessing the creditworthiness of their contractual partner. A mere gut feeling is a bad advisor here. What is needed are rational criteria that can be objectively verified.

Score values enable objective statements

This is where score values come into play. They are linked to facts that allow conclusions to be drawn about creditworthiness. Here is an example: Someone who is in permanent employment is less likely to become insolvent than someone whose employment is temporary. In a serious case, the lender must use statistical methods to prove whether this statement is true.

They must be based on meaningful facts

From the perspective of data protection, it is first important which facts may be included in a score value. These must always be factors whose suitability can be verified. Practical example: previous loans that the borrower has properly repaid. However, general experience may also be taken into account. It is conceivable, for example, that homeowners repay loans more reliably than people who are not homeowners. However, it could just as well be the other way around. The reason for this could be that homeowners have less money at their disposal because of the burdens imposed by the house. Whatever the case, it must be possible to justify such statements by statistical means.

Discrimination is prohibited

It is extremely controversial whether it is permissible to take into account in which district or street someone lives. Such an approach can quickly lead to impermissible discrimination. This is particularly clear from the following example: Several people who have not properly repaid loans in the past live in a certain house. Someone new moves into this house. The inference that this person will also not properly repay loans would inadmissibly discriminate against them.

A score value is a point value

Scoring points are assigned for each individual characteristic of creditworthiness. The sum of these evaluation points is the score value. The lender decides how high the score must be in order to still grant a loan. Each lender may apply its own standards here. Which risk they are willing to take and which they are no longer willing to take is part of their business policy.

Credit agencies calculate it as a service provider

Very few lenders calculate score values themselves. They generally lack the expertise to do so. They therefore call in credit agencies as service providers. The SCHUFA is very well known in this context. However, there are also smaller credit agencies that only work for certain industries, for example.

The case law on credit agencies is detailed

In principle, the business model of credit agencies is fine under data protection law. However, they must observe a large number of principles that result from court decisions. This applies, for example, to how long negative facts may be taken into account. Here, too, there is something like a right to be forgotten. However, this oblivion must not begin too soon. Otherwise, it jeopardizes the legitimate interests of creditors.

Data subjects have a right to information

Anyone affected by a score value can demand information about the score value. They can also request information about the facts used to determine the score. However, the calculation method in detail is considered a trade secret. A data subject cannot request information about this. Score values are generated millions of times a year. Measured against this, there are gratifyingly few justified complaints.