OMB’s Office of Information and Regulatory Affairs is seeking to establish priorities and coordinate research efforts across the Federal Statistical System to focus on improving federal statistics. In particular, a priority has been placed on using new techniques and methodologies based on combining data from multiple sources. To support this effort, OMB/IRA has published a RFI in the January 12 FEDERAL REGISTER requesting information on
(1) Current and emerging techniques for linking and analyzing combined data;
(2) on-going research on methods to describe the quality of statistical products that result from these techniques;
(3) computational frameworks and systems for conducting such work;
(4) privacy or confidentiality issues that may arise from combining such data; and
(5) suggestions for additional research in those or related areas.
While there are regulatory and statutory constraints on combining data within the federal government, the information sought concerns how best to combine data once they are accessed appropriately and successfully. The intent is for the research to inform the adoption of revised statistical standards regarding the use of such combined data for federal purposes, including but not limited to the production of principal key economic indicators and demographic statistical products.
Comments are due within sixty days (March 12). All responses must be submitted electronically to the following email address: FN-OMB-Combined-Data-RFI@omb.eop.gov.
This RFI is for information and planning purposes only. It should not be construed as a solicitation or as an obligation on the part of the Federal Government, the Office of Management and Budget, the Chief Statistician of the United States or SSP. OMB does not intend to make any awards based on responses to this RFI or to otherwise pay for the preparation of any information submitted or for the Government's use of such information.
- The federal government produces a wide array of statistical data that are a critical national resource. The use of these data are central to our democracy and include: Supporting constitutional duties, such as reapportionment of the House of Representatives; allocating resources to states, localities, tribes, businesses and individuals; supporting good planning at all levels of federal, state, local and tribal governments; describing our economic wellbeing; providing evidence to address critical problems facing our nation, such as opioid addiction; supporting informed public and private decision making that will create jobs and improve our infrastructure; and creating opportunities for local communities. These statistics use well-tested and documented processes that rely on censuses, sample surveys or administrative records.
- However, the federal government is facing a number of challenges for these traditional methods supporting informational needs of the future. It is well documented that survey response rates are declining, and costs are rising. At the same time, data users increasingly demand much more timely and granular information, such as local rather than national data. To meet the needs of the many stakeholders and policy-makers who depend on high quality, reliable federal statistical data, the statistical agencies must take advantage of new technologies and data sources to both reduce costs and make improvements. We believe there are many opportunities to increase the efficiency of the statistical system and reduce the response burden on people and businesses.
- The Chief Statistician of the United States and SSP are well aware of these issues and are seeking to change the paradigm underlying the production of these statistics. The Federal Statistical System must adopt new methods and standards to provide statistics that continue to meet the data needs of our nation for the 21st century. Given the existing environment, an important component of this transformation will be based on combining data from multiple sources to produce statistical products and information.
- Important work in the area of combining data from multiple sources has been conducted; see, e.g. National Research Council (2017; www.nap.edu/catalog/24652/innovations-in-federal-statistics-combining-data-sources-while-protecting-privacy); the related information provided through: http://sites.nationalacademies.org/DBASSE/CNSTAT/DBASSE_170268; and references cited therein. However, much more research must be carried out before the Federal Statistical System can adopt these techniques for the production of its key statistics. The Chief Statistician is therefore seeking to set priorities and coordinate Federal Statistical System resources to focus on such a program of continued research and therefore is requesting relevant information.
- This RFI seeks to identify published works, current and planned research, and descriptions of best practices taking place in private sector firms and academic institutions related to combining data from multiple sources to produce statistical data and products. The RFI is also seeking suggestions for new areas of research that the federal government should pursue in order to adopt new methods for combining data from multiple sources to produce statistics. These include but are not limited to:
- computational environments for accessing and processing multiple data sources
- measurement and documentation of the quality of statistical data derived from combining multiple data sources
- Ø new techniques for harmonizing and linking multiple data sources; issues regarding privacy and disclosure avoidance
- standards for describing the fitness for use of key statistics based on combined data sources; and principles for curating and disseminating these new data and associated products
- In addition, descriptions of and citations to papers or projects where data have been combined to do analyses that highlight sources of data that may be useful for government data integration, or how new data sources can be helpful in assessing how federal statistics can be better structured and presented to increase their value to the nation, are welcome.
- Finally, the RFI is seeking information on tested best practices related to securing partnerships across data holders and providing access to secondary users.
- The Chief Statistician of the United States and SSP plan to consider this input in focusing Federal Statistical System research efforts, including the Federal Committee on Statistical Methodology, on a program that informs policy and provides guidance on the Federal use of data combined from multiple sources.