Through a series of laws, policies and initiatives, the government’s effectiveness in using data to benefit the Nation has created a strong foundation for progress.
Blog Co-Author: Nick Hart, CEO, Data Coalition
Government collects and manages a tremendous volume of data on all manner of programs, activities, and services. But that information is not always used to its full potential to serve the public good or the American people. Building on strategic initiatives and efforts over the past 30 years, federal agencies have slowly moved more ably and responsibly to use data collected for identifying and solving problems. What are the key aspects of these initiatives in recent decades and what do they suggest about planning for 2021 and beyond?
In the mid-1990s, Congress enacted a wave of data-related authorities that set the stage for efforts to improve government’s information management capabilities. Key reforms affecting information management included:
Government Performance and Results Act of 1993. GPRA was established to encourage agencies to undertake routine strategic planning processes while also implementing a new suite of performance metrics to support congressional and public monitoring of programs and activities across government. The law resulted in a proliferation of new output metrics across the federal government as well as a quadrennial strategic planning process for federal agencies. In 2010 Congress updated GPRA to require “priority goals,” or high-level metrics that could involve both individual agencies or cross-government goals. In addition, the GPRA Modernization Act created new expectations for leadership involvement, including quarterly reviews for core metrics by agency leadership and OMB. Many leaders have moved GPRA forward within government since its passage, including former IBM Center Director Jonathan Breul and current Senior Follow John Kamensky in the 1990s at OMB and the National Performance Review, respectively; and Dustin Brown and Mark Bussow in OMB’s Office of Performance ana Personnel Management today – this prior post has more on GPRA’s evolution.
Information Quality Act of 2000. The IQA provides OMB authority to manage data quality and requires federal agencies to comply with certain standards and processes for ensuring the data used in decision-making retains acceptable quality. Under this law, federal agencies all adhere to OMB guidance on quality and issue, then follow, agency-specific plans. OIRA has published an influential set of guidance documents to implement the statute, most prominently led by OIRA Administrator John Graham in the early 2000s.
E-Government Act of 2002. The E-Gov Act created the role of a federal e-gov and IT leader who has since been designated as the Federal Chief Information Officer, a position to focus on systems integration, computer security, and broader IT budget and management policy and oversight across government. The E-Government Act also included provisions under the Confidential Information Protection and Statistical Efficiency Act to establish a new privacy framework for statistical analysis, largely applicable to the federal statistical agencies. OMB worked with Senate, House, and GAO staff to draft this bill and carry it through the enactment (see more on the E-Gov Act story here).
Each of these new legal authorities was paired with extensive regulation or guidance, and was followed by a series of activities across government to achieve Congress’s intent and direction. Each law introduced a sea change for government agencies in improving IT infrastructure and data systems, applying data for performance management, and improved coordinating data collection activities across and within agencies. Despite these successes, the legal authorities also introduced new challenges in coordinating across agency leaders and a web of legal authorities related to information management.
In part, due to the limits of the existing infrastructure and the desire to ensure information was meaningfully used in decision-making processes, three core types of activities manifested from the existing legal framework and took root government-wide over the past decade that spanned these legal authorities.
Open Data. Improved information sharing capabilities can support efforts to build trust in government, ensure accountability for officials and programs, and promote equity and social mobility. The Obama Administration, starting with the Open Government Directive on Jan 20, 2009, and the introduction of data.gov in 2010, and continuing with efforts led by the White House Office of Science and Technology Policy Deputy Chief Technology Officers including Beth Noveck and Nick Sinai, put in place a strong foundation to make progress on a variety of data sharing issues and promoting the publication of open data (read here for more on the foundations of open data in the government). In 2014, enactment of the Digital Accountability and Transparency Act initiated a process for publishing more open and public data about government spending. The Obama Administration also launched an entire initiative aimed at promoting open data across government agencies, which included the launch of the data.gov website as a repository for open data assets ranging from environmental enforcement to administrative operations of federal agencies. Key examples of products produced as open data include:
- As part of the last administration’s My Brother’s Keeper Initiative, federal agencies produced and published key national statistics stratified by race, ethnicity, gender, and age to provide improved insights about potential disparities in health, education, and other social determinants
- The Department of Education launched the College Scorecard by connecting information from educational institutions with income and earnings data, ensuring college applicants had reliable information to compare graduation rates, expected post-graduation salaries, and college costs.
Data Standardization and Machine-Readability. With intentional efforts to promote open data underway, aspects of data management that are often considered more technical also expanded in federal agencies. A slow shift of paper-based reporting to digital reporting meant that more information provided to government could be machine-readable. With machine-readability, collected information also needs to be comparable. Many federal agencies like the Environmental Protection Agency have long maintained their own data standards for reporting and enforcement purposes, but increased digitization is also resulting in more cross-agency standards, including those that build on consensus standards from non-governmental entities. A few key examples:
- Passage of the Digital Accountability and Transparency Act of 2014 led to improvements in financial reporting data standards to ensure systems interoperability and reporting capabilities between agencies and the Department of Treasury. The result is the publication of insights about government spending on public websites like the USAspending.gov, overseen by a team of experts at Treasury who are led by Fiscal Service Assistant Secretary David Lebryk.
- Following the Great Recession, the financial regulatory agencies of the United States supported an international effort to develop common identifiers for businesses and organizations. While not yet fully adopted in federal agencies, many like the Securities Exchange Commission and the Bureau of Economic Analysis are beginning to shift data collection efforts to include these open identifiers, often in lieu of or in addition to identifiers like DUNS numbers.
- Congress passed the Geospatial Data Act of 2018, which sought to address a longstanding issue of different organizations coding and using spatial information in ways that were incompatible. The act created a standing committee to advise on consistent geospatial standards for use in federal agencies, building on informal practices that initiated during the previous Administration.
Evidence Movement. The publication of open data and use of standards is not enough to ensure decision-makers have both access to useful information and then, subsequently, actually use the information to support key policy decisions. This Evidence “Movement” involved efforts to expand program evaluation capabilities, data accessibility, privacy protections, and access to studies and findings, all with the goal of providing decision-makers a supply of information that could be efficiently used.
A milestone in the movement occurred in 2017 when 15 politically-appointed experts issued unanimous findings and recommendations from the U.S. Commission on Evidence-Based Policymaking about next steps in the movement; co-blogger Nick Hart served as Policy and Research Director for this Commission. The Evidence Commission recognized the need for government to improve access to existing administrative records for “statistical activities,” while also enhancing privacy protections and building capacity for evidence building. The Evidence Commission’s 22 recommendations covered these themes, many of which specifically encouraged the creation of a new National Secure Data Service to support data linking and sharing across federal agencies, state and local governments, and the research community.
Connecting Activities in New Law and Guidance
With the hope of connecting core strands of the data management and governance infrastructure, while also making progress on the Evidence Commission’s recommendations, in 2018 Congress passed the Foundations for Evidence-Based Policymaking Act, bipartisan and bicameral legislation sponsored by House Speaker Paul Ryan and Senator Patty Murray. The Evidence Act’s focus on capacity, open data, coordination, and privacy sought to better align the framework that in prior years became increasingly fragmented with legislative and regulatory accretion. Key aspects of the Evidence Act include:
Governance and Planning. The Evidence Act establishes a new leadership framework for data activities that includes chief data officers, evaluation officers, and statistical officials in agencies. OMB issued guidance on how to implement the leadership roles and most, though not all, agencies have now identified these officials. The Evidence Act also requires agencies to develop “learning agendas” in which they identify core policy questions to be addressed, as well as the data needed to answer those questions, which could reside in another federal agency. Interim learning agendas are to be submitted to OMB in September 2020. Co-Blogger Nick Hart is working with Kathryn Newcomer and Karol Olejniczak at George Washington University on a Center report about implementation of agency learning agendas, expected out in early 2021.
OPEN Government Data Act. The Evidence Act’s Title 2 specifically promotes open data in government, using the FOIA standard of openness to determine what non-sensitive information should be deemed open by default in a machine-readable format. Building on a 2013 memorandum, the law requires agencies to catalog datasets with appropriate metadata and publish the information publicly. The provisions of the OPEN Government Data Act apply to nearly every agency in the federal government. Implementation these provisions has not progressed far in most agencies, pending issuance of OMB guidance.
Confidential Information Protection and Statistical Efficiency Act (CIPSEA). The Evidence Act’s Title 3 applies primarily to the federal statistical system and includes new language that codifies Statistical Policy Directive No. 1 (practices for ensuring public trust in government data), establishment of a one-stop shop for researcher access to confidential data, enhanced risk assessment practices for managing de-identification and disclosure avoidance, and a new data sharing authority that shifts the presumption of accessibility for administrative records unless a statutory prohibition on use or sharing exists. OMB has yet to publish rules or guidance documents that enable implement for many of these new authorities.
The Executive Branch opted to go even further than the Evidence Act requires, for example by weaving together other data laws with emerging interests in artificial intelligence analytical applications. The Federal Data Strategy in turn outlines a series of principles and practices that federal agencies are expected to achieve over the next decade. To support implementation of that strategy, OMB published a one-year action plan with specific deliverables for agencies to support progress on the strategy.
Despite slow progress over the years, the pandemic and recession, as well as other policy challenges, highlight the need for rapid, government-wide solutions to data collection, use and reducing barriers to sharing. Current and prior efforts can be leveraged to recalibrate and enhance data use, rather than discarding or disregarding the progress to date.
The next stages in improving government’s data capacity will certainly include some aspect of a centrally-coordinated strategy as well as increased input and advice from those outside government. One promising endeavor is the Advisory Committee on Data for Evidence Building, established by the Evidence Act, which includes experts in evaluation, data governance, privacy, artificial intelligence, and other relevant issues. That committee is expected to offer further recommendations to OMB on how continued progress can be achieved.
The establishment and designation of new Chief Data Officers, Evaluation Officers, and Statistical Officials across agencies complements existing senior leaders for IT, privacy, security, and project management to promote coordination and collaboration across agency c-suites. This collaboration and leadership, alongside a cohesive legal framework and clear strategy for continuous improvement, represents key elements need to ensure that the progression of data initiatives continues in a way that meets the needs of government and the Nation.