A Final Pet Peeve: The Right to Consumer Financial Industry Data

Thank you to the Credit Slips team for allowing me to use their soapbox for the last few weeks.  I leave you with a final pet peeve: Why does the government have to rely on commercially-collected financial industry data sets or voluntary surveys of financial firms to discover the effects of policies the government has put in place? This is just embarrassing. The U.S. government has so little power over the financial industry – an industry that only exists by virtue of the full faith and credit, payments systems, FDIC insurance, etc. provided by the U.S. government – that it cannot demand data from banks and financial firms, but instead must ask politely for voluntary survey answers or search the data market and pay for information like a commoner? 

The CFPB fancies itself a “data-driven agency” but is subject to budgetary constraints in obtaining that data. Worse, it can only obtain the data the market chooses to provide, which is often a bunch of incomplete data sets that cover performance of only a sample of any particular financial product or that consist of voluntary unverified survey responses of industry members. Even more galling, some of those data purchases come with use restrictions. For example, it appears that the CFPB's recent report on student loans was based on data provided voluntarily by lenders, data which was stripped of identifying information before it was shared with the government not merely to protect individual borrowers but to prevent identification of any particular lender within the data. 

Other government agencies are often in the same position. The Government Accountability Office, for example, often relies on data in whatever form industry chooses to sell or voluntarily share it (see, e.g., GAO’s report to Congress on the potential impacts of Dodd-Frank’s mortgage provisions, which forecasts the effects of Dodd-Frank's loan structure and underwriting provisions discussed in one of my prior posts by applying them to mortgages originations between 2001 and 2010, relying on loan-level data purchased on the private data market, data that covers only part of the mortgage market and only some of the pertinent loan structure and underwriting details). 

I remember back in 2003 when the Office of the Comptroller of the Currency suggested that the subprime mortgage loan market was price competitive and therefore not "predatory." The OCC based this conclusion in part on loan pricing and loss rate data from a single subprime lender. That lender had provided researchers with its claimed loss rate data from 2002 and with the interest rate sheets for 30-year fixed rate loans offered by that lender during a single week in 2002 for two subprime loan programs the lender ran in Colorado and Utah only. There was no evidence that any of this data was representative of subprime lending generally, or even of this particular lender's national portfolio or outside of the single week in which the rates were in effect.

How convenient for industry to be able to feed regulators the “data” it selects and restrict the use of that data as it sees fit. And if that data turns out less favorable than industry might have expected, it can then argue that any criticism of industry is based on “incomplete data.”

Each and every time the government issues, changes, or removes a regulation, industry should be required to report back with follow-up data, anonymized to protect individual consumer privacy, about which consumers have been affected by the change and how they have been affected. To the extent that trade secrets can be removed from the data, that data should be made available to the research community as well. 

No one can ever foresee all consequences of any regulatory change, or all the ways in which potentially affected parties might avoid the intended consequences. Particularly when a host of regulatory changes are made all at once, as in the Dodd-Frank regulations, vigilant monitoring of the effects is necessary to address the inevitable problems that will arise. 

Data alone will not yield all the answers; value-laden judgments are required because costs and benefits are rarely in commensurate currency. But exchanging data is no longer an expensive and time consuming proposition. Financial firms compile reams of data for their internal purposes, and the additional cost of sharing it with the government is almost certainly negligible. There is no need to deny the government the resources industry has. There is no need to regulate in complete dark when some light is available.