March 4, 2002

Ms. Kim T. Nelson
Assistant Administrator
Office of Environmental Information
U.S. Environmental Protection Agency
1200 Pennsylvania Avenue, NW
MC 2810A
Washington, DC 20460

Dear Ms. Nelson:

     The purpose of this letter is to encourage the Agency to take steps to improve the utility of the data available on the environmental performance of industrial facilities. As you know, environmental information plays an essential role in advancing EPA's objectives of protecting public health and the environment.

     In light of the importance of environmental information, we believe that EPA should explore several steps to enhance the practical utility of the information available to the public by establishing a single facility identification number, setting up an integrated system for reporting and access of data across multiple programs, and improving the timeliness of the availability of Toxic Release Inventory (TRI) data. In undertaking such steps, EPA will be advancing the goal -- identified by Congress in the 2001 Appropriations Act (Public Law 106-554; H.R. 5658, section 515(a)), and detailed in OMB's recently issued Information Quality Guidelines -- of ensuring and maximizing the quality, objectivity, utility, and integrity of information disseminated by Federal agencies.

     The adoption of a single facility identification number for reporting facilities has long been recognized as an important step in making data readily available to regulators and the public. This effort will augment the practical utility of the data by making it easier for the Agency, other governmental entities and the public to link data reported by a facility in different contexts (e.g., TRI data and water discharge data). In 1995, EPA began work on a rulemaking to establish common facility identification information across all its information collection requirements; more recently, however, EPA has channeled its efforts toward the development of the Facility Registry System. We encourage EPA to place a priority on the completion of the effort to create a unique facility identification number.

     Second, the adoption of a single entry data system will improve data quality, provide a comprehensive data set for each facility for use by EPA, State, and the public, and eliminate duplicative reporting requirements and the associated burden on regulated facilities of making multiple submissions of essentially similar data across EPA's programs. We understand that the development of the central data exchange (CDX) will provide for the gradual integration of existing databases into a single data repository and will eventually realize the goal of providing a single data entry system. We recognize that creating single data entry represents a significant undertaking that will require a coordinated agency-wide effort. The creation of a unified Office of Environmental Information to coordinate data collection, management, and dissemination throughout the Agency was an important first step. We encourage the Agency to develop a strategy to integrate databases as they are brought into the CDX and to move to a single data entry system as expeditiously as possible.

     Finally, EPA should explore ways to expedite the release of TRI data. Under EPCRA, industry respondents report their releases for the reporting year on July 1 of the following year. However, there has been a considerable lag in recent years in the public release of this data. For example, we understand EPA plans to release the TRI data for 2000 this Spring -- almost a year after it was received and more than a year after the end of the reporting year.

     Under our regulations implementing the PRA [§1320.3(e)], one of the measures of the practical utility of a data collection is the timeliness with which data is available to the collecting agency, to other government entities, and to the public. In order to improve the timeliness of the release of TRI data, we encourage the agency to explore ways to expedite the processing of the TRI data in order to achieve an earlier release. One way of doing so is to encourage greater use by respondents of electronic reporting. The increased use of electronic reporting reduces the transcription and quality control burden on the Agency and should allow quicker processing of the data.

     In processing the TRI data, the Agency also conducts a careful quality analysis of the data to identify potential errors in facility reports of releases or transcription of facility reports. For example, EPA compares the reported data with release data from earlier years to identify "anomalous" or "outlier" reports and follows up with the respondent in an effort to identify errors in reporting. In conducting this analysis, EPA is assuming some of the quality assurance burden that properly rests with the respondent. Respondents should be responsible for the quality of the reported data. Of course, EPA remains responsible for errors that occur during its own processing and analysis of the data and should continue to ensure that these are minimized. We encourage EPA to review its quality assurance role in order to reduce the processing time for TRI data and allow an earlier release of the data to the public.

In conclusion, OIRA recognizes that EPA has limited staff and budgetary resources to implement a large number of statutory programs. We look forward to discussing this issue with you soon, and would appreciate a response to this letter within 60 days. As always, my staff stands ready to assist you in this effort.

Sincerely,

    /s/

John D. Graham
Administrator
Office of Information and Regulatory Affairs