Home > Regulation & Examinations > Bank Examinations > Side by Side: A Guide to Fair Lending
Side by Side: A Guide to Fair Lending
Planning the Test
Selecting the Testers
Training the Testers
Providing Tester Identities
The first step in the planning process is to select a test manager knowledgeable in fair lending and capable of supervising all activities before, during and after the test. Planning should include:
The number of tests that should be performed to determine if a financial institution may have discriminatory lending practices that are illegal will vary according to the institution's size and loan volume. For example, if testing a multi-billion dollar institution operating in a large metropolitan area, several tests for each branch may be necessary. However, if the financial institution is small or has a low lending volume, more than a few tests may be impractical. The aim is to test a representative sample of an institution's lending staff over time.
Institution size and loan volume also should determine the frequency of test-ing. For example, all the testing for race discrimination in a financial institu-tion that typically has few minority loan applicants should not be performed in one week nor should the number of tests each week exceed the institution's weekly average of loan applications.
Properly trained, almost anyone can be a tester. The tester pool should reflect the general population in terms of race, color, religion, sex, national origin, age, marital status, and disability. Potential testers may include, retired individuals, teaching professionals, or graduate students among others. For example, the Philadelphia Commission on Human Relations hired as testers recently laid-off, experienced professionals through the area unemployment office. Typically, such recruits present themselves credibly and have good interpersonal and writing skills. Intelligent, dependable individuals who communicate well and think quickly are ideal tester candidates. Hiring testers should be like hiring any good employee.
Independent testing contractors will generally already have a pool of testers. If your institution retains an independent firm, verify that the
contractor's testers are representative of the protected classes you may want to have reflected in the tests.
Training is a crucial component of any testing program. The amount of time needed to train testers adequately will vary depending upon the experience level of the trainees and of the trainers. Testers should be advised that testing may be conducted even where there is no prior evidence of illegal discrimination. The following issues also should be reviewed with them:
In addition to this basic type of information and instructions on confidentiality, objectivity and accuracy, trainees also should be provided with necessary background information, a tester "identity", and instruction on completing Test Report Forms.
Testers should be presented with a tester identity before the test. Ideally, a tester would use his or her actual background. However, since it is unlikely that two testers will have similar enough backgrounds, fictitious identities are used for uniformity. As illustrated in Appendix A to this guide, the identities generally include information and instructions needed to complete the test; various personal and financial characteristics that the tester assumes for the test; and information about the lending institution to be tested. To maintain objectivity, however, a tester should not be provided with information about the identities of the other testers or the purpose of the test.
Items that might be included in an identity are:
As customer service personnel may ask why an applicant selected the particular financial institution, credible reasons for the choice and other pertinent information specific to the financial institution also should be provided to testers.
Identity information should be sufficient for testers to answer adequately most questions asked by a lender. The protected class tester's financial situation and employment record would usually be slightly better than those of the control tester. These differences should only be slight so as to suppress all socioeconomic factors except one, status in a protected class.
Identities should be varied to test for different types of illegal discrimination. For example, to test for discrimination based on receipt of public assistance income, one tester's identity should include that income source and the other tester's identity should not. Other information would generally remain the same. Individuals of the same race should be used when race discrimination is not the testing focus.
Testers should thoroughly familiarize themselves with the information contained in their identity form and consider the following questions:
If answers to any of the questions are yes, then a new tester should be assigned. In cases where the tester is known to institution personnel, or otherwise uncomfortable with the assignment, the test may be compromised.
|Last Updated firstname.lastname@example.org|