POGO Joins Comments Calling for Public Access to Consumer Complaint Narratives

Director Richard Cordray

Bureau of Consumer Financial Protection

1700 G Street N.W. Washington, D.C. 20552

CFPB Consumer Complaint Narrative Comments

Docket Number: CFPB-2014-0016

Sept. 19, 2014

Dear Director Cordray,

The undersigned consumer, privacy, civil rights, labor and open government groups unequivocally support the Consumer Financial Protection Bureau (CFPB) proposal to publish consumer narratives in the public complaint database. Director Cordray has called the complaint narrative the “heart and soul” of a complaint. We agree. Complaint details will provide a much more complete portrait of consumers’ experiences with financial companies and equip consumers to protect themselves from financial trouble. Complaint narratives will help the public understand problems and trends in the industry, and inspire companies to compete based on customer service excellence.

Offering consumers access to complaint details in the database is a powerful, practical tool that will allow people to learn more about the financial companies they may do business with, as well as the pitfalls and solutions devised for products and services they are considering. It will help consumers make more informed decisions about how to choose the best products and services for their needs, and how to protect themselves from problems.

Disclosing complaint details in the CFPB public database will:

• Empower consumers with timely, valuable information pre-purchase to prevent problems and reward companies that respect their customers
• Empower consumers post-purchase to report unreasonable, unfair or deceptive practices and alert others in advance of problems
• Allow others to assist the Bureau in detecting destructive patterns before they do extensive damage, as was the case in the lead up to the foreclosure crisis
• Encourage more people to use the database, as it becomes a more useful tool, creating a cycle of increased information about consumer experiences in the financial services marketplace.

We support the Bureau’s recognition that dispute details permit readers to more fully understand complaints, their resolutions, and to put them in context. We’ve long believed that the addition of narratives gives individuals a much richer understanding of the problems reported, and allows the public to assess their validity and draw its own conclusions. The CFPB database currently shows general categories of complaints. For example, banking complaints, may be listed as a “deposit or withdrawal” complaint. The difficulty is that this category provides so very little information about what the problem really was. Was it a late deposit dispute? Lack of notification? Repeated withdrawals without authorization? All of these are very different issues requiring different responses from a company.

Consumers also value the ability to access specific information about complaint resolutions. Substantive detail on complaint resolutions should be part of the publicly available complaint record as well. Information as to how a complaint was - or wasn’t - resolved can help shape consumers’ perceptions of a company, as well as inform them on how to protect themselves. Companies that resolve matters swiftly and successfully may come to view the public database as a useful tool to help a business improve its public image and reputation. Details on complaint resolutions can reveal a company’s willingness to work with consumers when problems arise – leading some to conclude that this is a company worth doing business with.

In fact, the database may spur competition between businesses for better complaint resolution records. It may encourage companies to resolve more complaints more successfully. The public database has already helped improve customer care at many companies. According to an analysis of the CFPB database by Deloitte “financial institutions appear to have improved their effectiveness at addressing complaints”. [1] Other industry analysts recommend, “… before problems escalate, companies should take advantage of CFPB data.” [2] Additional complaint information is likely to enhance companies’ attention to customer service and encourage competition for excellence in customer care which stands to benefit consumers, high quality corporations and regulators.

Consumer Choice

We endorse the Bureau’s proposal to leave the decision as to whether to disclose complaint details added to the database up to each individual consumer through an opt-in policy. Consumers should have the opportunity to make a personal choice based on the particulars of each dispute. The consent option should be made available before consumers submit their complaints, as this is the time when the details are freshest in their minds and they are most engaged in the complaint process. For efficiency, consumers should be able to note (possibly in a separate disclosure box) if certain details of a complaint should remain confidential in the database, even if the problem is made publicly accessible. We concur that the consumer should be given the opportunity to change her mind and return the complaint to a private setting.

Furthermore, the complaint intake form should include a box that allows consumers to provide permission for a representative (housing counselor, legal aid attorney, consumer advocate) to confer with the CFPB. Disclosure should be clear to all parties acting on behalf of consumers, that if a consumer opts in to including complaint details in the public database that all parties’ comments relevant to the complaint (minus personal information) would be accessible in the database.

Updating Narratives

At the CFPB field hearing in El Paso, TX. a panelist raised the prospect of updating narratives as the situation unfolds. Complaints could be resolved, left unaddressed, or new issues may develop from when a dispute was first submitted. We would support allowing consumers to update the narrative to ascribe new developments to the situation. Updates could be posted in the database alongside the original complaint with the original ID number and a date for the new submission.

Consistent with the value that the Bureau places on an open data policy, allowing consumers to add further chapters to the original complaint story may also help expose breakdowns in a company’s processes, may disclose harmful discriminatory patterns, or may reveal a consumer’s misunderstanding or misinformation about a dispute. Updates could also offer a solution to address proposal opponents’ outsized fears that information in the database might be inaccurate. For practical purposes, the Bureau may choose to limit the length of updates.

Companies’ Response

We support the Bureau’s proposal to permit a business to include its response to a complaint, generally published side by side with the original dispute. This approach allows the public to review the dispute in its entirety. Complaints should be posted after the 15-day window; company responses can be added as they are submitted. This allows others to draw their own conclusions as to the validity of a complaint, handling of the problem, validity of the resolution, or reasons why a dispute was not resolved. As a matter of fairness, it also allows a company to air its perspective in a dispute and allows for it to address any reputational concerns. Since the various perspectives in a dispute will be displayed simultaneously in the database, the company will have a reasonable opportunity to address points that it deems inappropriate or incorrect. If necessary, it can set the record straight in a timely manner and appropriate location.

The Bureau has asked for feedback on whether a company’s public response should be separate from its response to an aggrieved consumer.

In order to create the most robust and meaningful database possible, it is our recommendation that a company’s detailed response be included in the database with as much specificity as the situation warrants. Given the need to protect individuals’ personally identifiable information (PII), we recommend that the Bureau use the same scrubbing techniques on a company’s response as it would to the original complaint, with allowances for Bureau discretion to limit certain complaint details for consumers’ privacy protection.

In exceptional circumstances where a company is not be able to release certain complaint details to the public, the Bureau should decide on a case by case basis if its own redaction techniques would not be sufficient to protect a consumer’s identity. If in rare cases, the Bureau allows a company to supply a separate, more detailed response to the CFPB than for public disclosure, both the detailed private response and the separate public response should be shared with the Bureau and the consumer.

The proposal to publish narratives in the database lives up to the Bureau’s commitment to transparency while incorporating strict precautions to protect consumers’ personal information.

Protecting Personal Information

We appreciate the level of caution the Bureau is proposing to protect consumers’ personal information and prevent re-identification. While releasing complaint details is not without risk, we believe the precautions that the Bureau has proposed will reasonably protect people’s personal information. We support the CFPB’s plan to require informed consumer consent to disclose individual complaint narratives in the public database. We are pleased that the CFPB plans to use search algorithms to comb complaint data for personally identifiable information (PII) to diminish the possibility of accidental disclosure. Another layer of manual inspection of the data will follow this effort to further scrub each complaint of details that might inadvertently identify individuals. We support the Bureau’s plans to exercise discretion to minimize risk of personal identification.

During the complaint intake process, we recommend that frequent reminders be posted to prompt consumers not to include PII.

Additional steps to protect privacy

The consumer complaint database should include narrative information to allow analysis by company, complaint type, geographic region/zip code, and other fields of interest to the public, researchers and regulators.

The Bureau plans to start with the HIPAA Safe harbor method to de-identify data which removes 18 specific identifiers.[3] We endorse that proposal with one key modification.

In response to the Bureau’s request as to whether to continue publishing five-digit zip code information in the database, we recommend reporting five-digit zip codes for most communities, while modifying zip code disclosures where populations are small.

To prevent possible re-identification of people who live in zip codes with very few residents, we suggest the CFPB omit zip code reporting for zip codes with fewer than 10,000 residents. (Seventeen percent of Americans live in zip codes with less than 10,000 residents.)[4] A geographic identifier for both city or town and county should be used for complaint records from consumers in zip codes with fewer than 10,000 residents. In sparsely populated cities and towns, reporting at the county level only would be appropriate. In counties or cities that have only one zip code with fewer than 10,000 residents, residents’ privacy can be protected by also omitting zip code reporting for the zip code with the second smallest number of residents. This way, a complaint filer’s zip code cannot be reverse engineered from the data.

Nearly 83% of U. S. residents live in zip codes with at least 10,000 people. Requiring zip code identification at the 10,000-person level would provide a sufficient amount of information to identify patterns of service or abuse for most citizens. If the Bureau were to omit zip code identification for all zip codes with fewer than 20,000 residents it would exclude almost 35% of the population, and obscure critical information about the distribution of complaints needed to identify potential abuse patterns.

Moreover, in certain “down-sized” cities around the nation, setting a zip code cut-off as high as 20,000 residents would have a significantly disproportionate effect on neighborhoods with large numbers of persons of color. For example, in Rochester, New York that higher cut-off would leave out several zip codes in the heart of the city that account for 46% of the city’s total African American population (zip codes 14605, 14607, 14608, 14611 and 14619), thus preventing Rochester advocates and New York State agencies from identifying complaint patterns possibly relevant to the racial and ethnic composition of local neighborhoods. [5]

While using a population cut-off that is too large is undesirable because it is likely to exclude persons of color, the consequences of using a “three-digit” zip code protocol would be even worse. In areas where persons of color are concentrated in particular neighborhoods, a three-digit protocol would err in the other direction by being too inclusive. That is, using an area that is geographically too large, would deny analysts and policy makers the ability to identify significant local anomalies within the broader area. For example, Rochester’s ”146” three-digit zip code prefix encompasses a total area more than twice the size of the city itself. Such a large geography impedes the ability to ascertain patterns associated with neighborhoods having larger concentrations of persons of color or high poverty.

In contrast, reporting zip codes for complaint filers who live in areas with 10,000 or more residents would both ensure comprehensive reporting in many small and medium-sized American cities and facilitate the identification of harmful patterns of unfair or discriminatory behavior.

Detecting discrimination via protected class data

In addition to making public the narrative data, we urge the CFPB to encourage complainants to voluntarily provide protected class data to assist regulators, researchers and the public in detecting and rooting out discriminatory practices. At intake, the CFPB should request complaint filers’ information on ethnicity, race, marital status, age, source of income and gender. (This data is already requested in every mortgage loan application). While providing this data should remain optional for consumers, we suggest that the CFPB encourage reporting by including a compelling explanation as to why this information is so valuable.

Since individuals are often not in a position to determine discrimination, we recommend that collected protected class information be included in the consumer complaints database so that regulators, researchers and consumers have the opportunity to evaluate whether there are patterns of discrimination and abuse.

We share your deep commitment to safeguarding personal consumer information, and as you consider the challenges of data disclosure we ask that you remember that so much of the data that we seek to protect is already available to companies that can afford to purchase it. Consumers are already identifiable to those businesses that can pay to access an enormous amount of detailed individualized personal information, from people’s personal web browsing habits and shopping history, to information about their credit scores, mortgages, and more.

Given the CFPB’s pledge to educate and protect consumers using transparency and useful information, the release of complaint narratives aligns with the agency’s mission to provide the most meaningful information possible for individuals to make financially sound decisions.

Improving the process

In addition to making complaint narratives public, we offer a few additional suggestions for improvements to the overall complaint process.

Built into the process should be a periodic review of the complaint system to ensure it is working as expected. This would include interviewing or surveying consumers who’ve used the system to better understand why people did or did not achieve complaint resolution.

The CFPB should provide more direct communication with consumers when complaints are not resolved or not responded to by a company. The Bureau should explain what options the consumer currently has, what if anything the Bureau will do on the consumer’s behalf (including having CFPB Supervision and Enforcement divisions review the complaint) and alert consumers to the choice to dispute a company’s response via the CFPB’s complaint dispute process. Where possible, consumers who’ve complained should be notified if their problem is being investigated for possible legal violations and informed of any Bureau findings or actions related to the complaint. Additionally, if the Bureau does investigate and close a complaint the consumer should be notified and an explanation for the closing should be provided.

Building on others’ success

Consumers are accustomed these days to seeking input from others’ experiences, informally, on websites and through various databases prior to making financial decisions.

Government databases such as the Consumer Product Safety Commission’s saferproducts.gov and the National Highway Traffic Safety Administration’s safercars.gov already serve to help consumers make more informed buying decisions with the use of complaint narratives. Increased access to complaint and ratings information has led to better products and programs. For example, NHTSA has reported that its New Car Assessment Program has encouraged manufacturers to voluntarily achieve high ratings and thereby increase vehicle safety[6] The Bureau’s proposal to include complaint details in its database would also prepare consumers to make wiser financial evaluations based on others’ real-life experiences. The CFPB clearly states that it verifies a consumer’s account relationship with a company prior to posting, not the disputes. No public database verifies disputes; that is not its purpose. The mere existence of a government complaint database that includes narratives encourages companies to operate with a set of best practices that can be advantageous to all customers.

We are confident that the Bureau’s proposal to allow the public to learn from others’ problems (with informed consumer consent) will benefit us all by providing a tool to better identify and avoid unfair, deceptive or dangerous financial products and practices, incent clear information, quick resolution of problems, and excellent customer service.

Thank you for the opportunity to comment on this important proposal.

Sincerely,

Americans for Financial Reform

Alliance for a Just Society

Arkansas Community Organization

California Reinvestment Coalition

CCAG (Connecticut Citizen Action Group)

Center for Digital Democracy

Center for Responsible Lending

Community Legal Services, Philadelphia

Connecticut Fair Housing Center

Consumer Action

Consumer Federation of America

Consumers for Auto Reliability and Safety

Consumer Watchdog

Demos

Electronic Privacy Information Center

Empire Justice Center

Florida Alliance for Consumer Protection

Home Defenders League

International Union, United Automobile, Aerospace & Agricultural

Implement Workers of America (UAW)

Keystone Progress

Leadership Conference on Civil and Human Rights

Massachusetts Consumers' Coalition

MASSPIRG

Miami Valley Fair Housing Center, Dayton Ohio

Missourians Organizing for Reform and Empowerment

NAACP

National Association of Consumer Advocates

National Consumer Law Center (on behalf of its low income clients)

National Council of La Raza

National Fair Housing Alliance

National People’s Action

New Economy Project

New Jersey Citizen Action

New Jersey Communities United

Oregon Consumer League

Privacy Rights Clearinghouse

Privacy Times

Project On Government Oversight

Public Citizen

Public Justice Center

South Carolina Appleseed Legal Justice Center

Southwest Center for Economic Integrity

Texas Legal Services Center

The Institute For College Access and Success

U.S.PIRG

Virginia Citizens Consumer Council

Woodstock Institute

World Privacy Forum

[1] http://www.deloitte.com/assets/Dcom-UnitedStates/Local%20Assets/Documents/FSI/US_FSI_CFPBConsumerComplaintDatabaseFINAL2_091913.pdf pg 9

[2] http://data-informed.com/leveraging-regulatory-data-improve-customer-experience/#sthash.7uyJINcZ.dpuf

[3] http://cphs.berkeley.edu/hipaa/hipaa18.html

[4] Tabulations run from U.S. Census Bureau, 2008-2012 American Community Survey, 5-year data, Table DP05, Demographic and Housing Estimates.

[5] Ibid.

[6] National Highway Traffic Safety Administration, The New Car Assessment Program Suggested Approaches for Future Program Enhancements Jan. 2007,