Understanding the risks to economic statistics

At a cost of less than one-tenth of one percent of the federal budget, government statistical agencies provide invaluable information that informs millions of decisions made by Americans every day. Social Security payments are tied to changes in the measured inflation. Retail stores decisions about what items to put on which shelves rely on government data describing local area demographics. The Federal Reserve and private-sector investors use government data to understand current economic conditions and the economic outlook, shaping global capital flows.

The federal statistical system (FSS)—the thousands of workers across over a dozen government agencies that produce this vital data—faces serious challenges, from rapidly evolving technology to distrust in the data. FSS products represent an enormous return on the taxpayer dollar, and yet the system is chronically underfunded and understaffed. Strengthening the FSS should be one of the most important tasks facing government, but instead, many in the government have been openly adversarial.

In response to these challenges, the Brookings Institution recently launched the Economic Indicators Initiative (EII), a collaborative effort with experts from several other research institutions with the mission of highlighting risks to the FSS and strengthening the production of economic statistics. The EII project is part of a larger response to administrative actions regarding how the United States produces key economic statistics.

The EII foundational principle is that the integrity—and, crucially, public perception of the integrity—of economic statistics is of vital importance to U.S. households, businesses, investors, and policymakers. Core to this effort is the bedrock principle that official statistics not be a subject of political interference and controversy. As Brookings president Cecilia Rouse said at a recent event, “Politicizing federal statistics and questioning the integrity of those who produce them harms decision-making in every sector. It compromises the ability of policymakers in the executive branch, Congress, the Federal Reserve, and leaders throughout the business community to properly analyze the state of the economy and develop the best practices to ensure prosperity.”

The U.S. has been a world leader in economic measurement since the first statistical agencies were established over a century ago. Over time, the development of modern statistical sampling and estimation techniques and the introduction of modern data collection technologies has fundamentally changed how we measure economic activity. Despite this progress, much more needs to be done to introduce new and innovative tools that hold the promise of producing more accurate statistics at lower costs. Statistical agencies have been hampered significantly toward reaching these goals owing to a lack of funding and more recently, a loss of experienced personnel. These observations motivate the work of EII.

Introducing the EII

The broad mission of the EII is to build up insights, develop consensus, and create proposals to strengthen the production of key economic indicators. While our primary focus is on measuring the nation’s inflation, employment, income, and wealth, we are also concerned with sub-national aggregates and distributional statistics.

The first collaborative activity of the EII was commissioning four academic papers that together frame key issues confronting the statistical system: Threats; Use cases; Trust; and Innovation. The concepts and issues explored in these papers are interrelated. Threats to the statistical agencies make it difficult to maintain trust in the statistics, produce relevant statistics for users, and have resources for innovations. In turn, the papers demonstrate that innovations can aid in dealing with threats, such as low response rates, and improve accuracy to increase trust and address user needs.  Each paper highlights possible improvements to reduce threats and improve trust and usability, such as use of alternative data (both administrative and commercial data). The paper on use cases provides specific data needs along four dimensions, along with possible improvements. The “trust” paper shows that building trust is more difficult today because of the new threats and challenges facing the statistical agencies. All of the papers demonstrate the importance of the statistics that the agencies produce, and provide a framework for us, as data users, to help the statistical agencies minimize threats, produce relevant statistics, build trust, and continuously improve.

Threats

Statistical agencies’ ability to produce accurate, timely, and trusted data depends on expert staff working in the weeds with stable resources. But the dedicated staff of these agencies have been working in a challenging environment for over a decade, with falling budgets and response rates and minimal funding for improvements. The last year has brought even larger budget cuts and staff reductions, all while trust in the statistics has fallen and new challenges in modernizing systems and statistics have emerged. Moreover, the Trump administration’s decision to disband advisory panels of outside experts has added to the challenges facing the agencies.

The “threats” paper by David Johnson provides an overview of the internal and external threats to the FSS and identifies a key strategy for how agencies can survive in this challenging climate: Keeping the staff motivated. After reviewing the structure of the FSS and headline economic indicators (or Principal Federal Economic Indicators), the paper focuses on the internal threats of falling response rates and difficulties in innovation and the external threats of decreasing budgets, staff, trust and increasing political interference.

The “threats” paper shows that falling response rates are a concern if they bias the statistics, demonstrates the challenges faced by agencies in modernizing their statistics, highlights the recent reductions in staff, and the implications of decreasing financial resources over the past 15 years. The paper concludes with discussing attempts to limit the independence of the agencies and discusses the implications of decreasing trust.

Use Cases

As uncertainty around the quality and continuity of federal data systems increases, stakeholders across government, business, research, and civil society are recognizing both the value of current statistics and the need for improvements to better serve diverse user needs. Federal economic statistics are critical in supporting economic policy, business decision-making, and societal understanding, while identifying key gaps and opportunities for innovation.

The “use cases” paper by Rekha Balu and William J. Congdon reviews four key federal economic concepts—employment, prices, income, and wealth—by mapping key user groups (federal/state/local government, Federal Reserve, businesses, researchers, civil society) and assesses how existing statistics meet user needs and where innovations are needed. It highlights directions for statistical or measurement innovation to better serve policy, research, and public understanding. Throughout these domains, the paper identifies technical gaps in data collection methods, conceptual challenges in defining key measures, and limitations in current statistical products.

The paper contributes to current conversations about the future of federal statistics by highlighting potential data collection and usage responses to the changing nature of work, changing nature of assets, emerging data providers, and the evolving role of AI. The paper connects to other current conversations on privacy‑protecting methods for making more data available, including potentially from private sector providers, and how to make data more usable.

Trust

Public confidence in federal economic statistics has been steadily eroding. Drawing on long-running surveys from Pew Research Center and recent findings from the American Statistical Association, the “trust” paper by Michael Horrigan shows that declining trust in government has increasingly spilled over into skepticism about official economic data. While some Americans have complete faith in federal statistics and others reject them outright, most fall somewhere in between, and evidence suggests that, on average, public trust has been moving downward over time. President Trump’s baseless assertion that Bureau of Labor Statistics data were falsified for political purposes, along with his decision to fire the BLS commissioner, seem designed to continue the erosion of trust in the data.

The paper argues that trust depends on several core principles: political independence, rigorous and transparent statistical methods, professional standards, and clear communication. It explains how users judge data quality across familiar dimensions: accuracy, timeliness, reliability, relevance, comparability and coherence, confidentiality, and accessibility. But the paper also emphasizes that perceptions of trust vary widely by audience, such as policymakers, journalists, businesses, and the general public.

The “trust” paper explores the threats and challenges facing statistical agencies today, including declining survey response rates, shrinking budgets, politically charged interpretations of routine data revisions, and misunderstandings about how official statistics are produced. High-profile events such as large revisions to payroll employment estimates can fuel suspicion when agencies do not proactively explain what the numbers mean and why they change. The paper points to promising innovations such as the use of administrative records, private-sector data, and machine learning that could reduce reporting burdens and improve measurement, but only with sufficient resources and institutional support.

Ultimately, the paper argues that restoring trust will require more than technical fixes. It calls for renewed investment, stronger safeguards for agency independence, clearer public communication, and sustained engagement with researchers, businesses, and policymakers to ensure that federal statistics remain credible, relevant, and worthy of public confidence.

Innovation

Despite the administrative, technological, and societal challenges the FSS faces, staff in the statistical agencies and outside collaborators are constantly working on innovative methods to produce data that is more accurate, more timely, and less costly.

To understand the specific challenges that innovative FSS staff are trying to overcome, the “innovation” paper by Claire McKay Bowen examines how federal data are produced across the full data lifecycle—from collection to termination or archival—through the lens of four data‑governance values: accuracy, accessibility, privacy, and usability. The paper highlights how federal economic data are fundamental but often invisible to the public and policymakers who depend on them.

Across the lifecycle, the paper identifies fifteen persistent challenges in producing high‑quality economic data. These include integrating administrative and private‑sector datasets, ensuring representation of hard‑to‑reach populations, balancing privacy protection with data utility, and modernizing metadata. To illustrate potential solutions, the paper highlights four recent innovations projects where outside researchers are actively working with the statistical agencies. These projects include the Re-Engineering Statistics using Economic Transactions (RESET) project using transaction-level data to improve prices and quantity measurement; the Census Bureau’s National Experimental Well-Being Statistics (NEWS) project blending survey and administrative data; the Comprehensive Income Dataset (CID) linking survey and administrative data; and the Statistics of Income Division at the IRS and Urban Institute Safe Data Technologies project applying privacy-enhancing technologies for secure access to confidential administrative tax data.

Together, these examples show how technical and policy innovations can strengthen the FSS’s ability to deliver cost‑efficient, high‑quality economic statistics that addresses threats, meets user needs, and builds trust. The paper concludes by calling for sustained modernization, cross‑disciplinary training, and adaptive policies to ensure federal economic statistics remain robust, trusted, and fit for purpose in a rapidly changing technological and economic landscape.