Ever since the 1995 CRA rule was published bank regulators have mandated certain “community” and “market” benchmarks as the basis for rating bank performance under the CRA regulations. Most of that data is in the public domain and is transparent and free. For example, the FFIEC annually publishes over 1,200 fields of population and housing demographics plus mortgages reported under HMDA and small business and small farm loans reported under CRA. But when it comes to one key demographic variable, business demographics, the data is not published by the FFIEC and must be purchased from a commercial source, thereby creating a monopoly for the business demographics source.
The reader may think that this single demographic variable can’t be that important in comparison to the more than 1,000 demographic fields of data published annually. But that inference is misleading because, although the FFIEC data contains more than 1,200 fields of demographics, the agencies rely, for CRA performance evaluations, only on the distribution of families by income class (“borrower characteristics”) and the distribution of owner-occupied by census tract income class (“geographic distribution test”). The business demographics are used for the “geographic distribution test” (business distribution by tract income class) and “borrower characteristics” test (businesses with $1 million or less gross annual revenue). In other words, business demographics constitute 50%-67% of the demographic variables that are the normal standards for a CRA performance evaluation for banks that extend residential mortgages and small business and/or small farm loans.
For a bank that self-monitors its CRA performance, access to the business demographics is absolutely required to complete the analysis (as strongly recommended by regulators). But unlike all the other performance benchmarks, the business demographics are sole sourced thereby creating a monopoly on an important performance parameter. This can be expensive for banks and even more so for the public when community groups want to evaluate a bank’s performance (rather than wait 3-4 years between every CRA exam).
The agencies have not indicated that they have done anything to evaluate the integrity of the business demographics used as critical benchmarks for CRA performance evaluations. When comparing the FFIEC-approved business demographics to the “Census County Business Patterns” data, there appear to be large unexplained differences in the count of businesses. This could be attributable to establishments that employ people and businesses that are self-employed people. This latter group tends to be at home and far more numerous than employer establishments. Also, self-employed businesses tend to use non-commercial finance to fund business activity. Therefore, using them as a reference point for benchmarking small business loans could be extremely misleading. The regulators should disclose what they have done to assure the integrity of the data and its appropriateness for use as a benchmark for small business loans.
Finally, if the current source of the official business demographic data continues, the agencies should acquire the demographics and make them part of the public domain, accessible to banks and the public. It’s time to end the data monopoly created by the policies of federal agencies.
