Principles and Practices for a Federal Statistical Agency: Eighth Edition (2025)
Chapter: 2 The Value of National Statistics
2
The Value of National Statistics
Highlights
- Similar to the physical infrastructure of a nation, such as its interstate highways, national defense assets, and interstate utility grids, a national data infrastructure connects and serves society.
- The statistical information that makes up our national data infrastructure powers decision making. It informs political representation and business decisions. Statistical information allows policymakers to monitor the economic and social health of the nation, states, and localities. It provides evidence for developing and evaluating public- and private-sector programs. It informs federal, state, and local government action. Statistical information also provides input to research that contributes to public understanding of our world.
- Federal statistics are a sound national investment. The cost of federal statistical programs is a tiny fraction of overall U.S. federal spending. The annual benefit to our nation has been valued at $778.0 billion in 2022.a
__________________
a See https://www.commerce.gov/news/blog/2024/12/revenue-industries-heavilyreliant-us-government-data. This paragraph and footnote were modified after release of the report to provide a more precise estimate of value and its context based on publication released on December 3, 2024.
STATISTICS SUPPORT OUR NATIONAL INFRASTRUCTURE
People in the United States and other countries rely on data and statistics to live their lives, often without realizing it. They may check weather, traffic, or air quality reports and other readily available data to guide how they go about their day. They may use data to inform key family and personal decisions, such as where to live, based on information about housing, crime, schools, and jobs. In their own jobs, people may use data to guide policies and programs, make investment decisions, plan for the future, and develop knowledge.
People in a democracy also rely on accurate and trustworthy information to carry out their civic duties and maintain a government that protects and serves their interests. An informed citizenry must judge the merits of government actions through periodic votes for elected officials, and to become informed citizens depend on widely distributed information available both directly and through the media. Lawmakers require accurate population statistics for drawing up legislative districts that give all voters an equal voice. Fiscal and monetary policy actions rely heavily on economic releases. Identifying problems to address and opportunities to pursue requires that decision makers in both the governmental and private sectors have objective and timely information on the society and economy. Adding to knowledge about the society and the economy, in turn, requires detailed information for researchers to analyze in a wide range of fields.
The cornucopia of statistical information that people use in all these ways and often take for granted comes from a wide range of sources—censuses, surveys, sensors, commercial transactions, and records of all kinds. The information is made available not only by governmental entities, but also by businesses, the media, and other organizations, in the form of tables, graphs, maps, datasets, and other formats available today through the internet and other modes of access.
To meet the many and essential uses described above, information on a nation’s society and economy must be credible and trustworthy. The consumers of the information must believe that the information is objective and not affected by any political or ideological perspective concerning the phenomena being measured. They must trust that the technical expertise of the producers of the information is sufficient to produce statistics that will meet their needs, which include consistency in definitions and methods so that one can judge whether things are getting better or worse over time and can compare different parts of the country.
Who produces such information to inform societal and economic planning, decision making, and knowledge generation and to power the myriad needs and requirements of modern democracies and advanced economies? Many actors provide useful information, but across the world statistical
agencies in central governments have the role of producing key national statistics in ways that maximize their credibility and utility to inform policymakers and the public. Furthermore, by extension, statistical agencies also help citizens assess the reliability of the information released by other sources.
The United Nations General Assembly in 2014 formally endorsed Fundamental Principles of Official Statistics (see Appendix C).1 The first of these principles accords worldwide recognition to the indispensable role of official statistics:
Official statistics provide an indispensable element in the information system of a democratic society, serving the Government, the economy and the public with data about the economic, demographic, social and environmental situation. To this end, official statistics that meet the test of practical utility are to be compiled and made available on an impartial basis by official statistical agencies to honor citizens’ entitlement to public information (United Nations Statistical Commission, 2014, pp. 1–2).
In this regard, national statistical information forms a data infrastructure that resembles the role of physical infrastructure for a nation, like interstate highways, national defense assets, interstate utility grids, and basic scientific research (see also American Statistical Association, 2024). All of these national investments serve the common good. Their benefits are sometimes relatively small for an individual and often unrecognized, but they are essential to the welfare of the whole country. In some sense, these infrastructures are among the threads of the fabric of society.
In their day-to-day lives, most people do not think about the benefits of highways and bridges—until they exhibit a problem, perhaps being closed for repairs, or worse, when they fail and collapse. So too, when statistical information is disrupted or compromised, its value is vividly illustrated by decisions that, in retrospect, appear misguided. As just one example, inadequate information that results in underestimating the depth of a looming recession or, conversely, an economic boom, can lead to less-than-optimal policies to rekindle or rein in growth (Reamer, 2014). As another example, national data on the societal impact of pandemics or environmental change, such as COVID-19 and wildfires, are essential for policymaking. Delays or incomplete national data miss opportunities to provide guidance to the public at a critical time. Issues such as these require ongoing monitoring of their impact across many facets of our society.
___________________
1 These were originally adopted by the United Nations Statistical Commission in 1994; see Appendix C.
STATISTICAL INFORMATION POWERS POLICYMAKING
Following are a few examples of the many ways that statistical information provided by federal agencies serves the nation.
Informs Political Representation
The U.S. Constitution mandates a decennial census of the population every 10 years (the first census was taken in 1790) for determining the allocation of seats in the U.S. House of Representatives among the states. Reapportionment in turn triggers the redistricting process, by which states, using census data, redraw the boundaries of congressional districts to accommodate changes in the number of seats and in the geographic distribution of the population. States and many local governments also use census and other data to reapportion and redistrict their legislative bodies.
Informs Economic Decision Making
Federal statistics drive important decisions. The federal government currently labels 36 statistics—such as gross domestic product (GDP), the employment situation, monthly wholesale trade, weekly natural gas storage, crop production, consumer credit, and others—as “principal federal economic indicators.”2 The release of these statistics moves financial markets. The Office of Management and Budget’s (OMB’s) Statistical Policy Directive No. 3 requires these indicators to be published by the cognizant statistical agency on specified release dates under procedures designed to protect the integrity and credibility of the estimates and ensure that they are not subject to manipulation and do not give any user an unfair advantage, so that businesses and the public can be confident the statistics are objective (OMB, 2024d). (See Appendix A, and Practice 2 in Chapter 3.)
The indicators often lead the headlines upon their release, contribute significantly to public- and private-sector decision making, and help inform the public as to where the nation has been and where it is going. They and many other federal social and economic statistics have real consequences: the Consumer Price Index (CPI) determines annual cost-of-living adjustments to Social Security and Supplemental Security Income monthly benefits, which in May 2024 amounted to $120.6 billion provided to 67.8 million people.3 Annual changes in the CPI also affect commercial and residential rents, public- and private-sector wages, and components of the federal income tax code. Annual changes in prices for geographic areas
___________________
2 See Statistical Policy Directive No. 3 in Appendix A.
enter into local decisions, and monthly changes in prices are a major input into Federal Reserve Board decisions on short-term interest rates. Annual poverty estimates from the Small Area Income and Poverty Estimates Program of the Census Bureau are used to allocate funding to school districts under Title I of the Elementary and Secondary Education Act.4
Informs Business Decisions
The U.S. and global economies are powered by data.5 Whether starting or expanding a business, exploring prospects for different occupations, anticipating demand for products, projecting the labor force, evaluating effects of trade patterns, targeting investments, forecasting energy prices, planning for hurricanes, funding pension plans, devising better ways to serve customers with disabilities, or finding suppliers, business owners and community members rely every day on data produced by the federal government. Beginning in the 1960s, data provided by the U.S. government electronically spurred the development of a new sector: firms6 that provide government-data-related products to households, businesses, and organizations. Over the past 10 years, the revenue produced by industries that rely on government data has increased. Between 2012 and 2022, the revenues of government data-intensive sector (GDIS) grew from $407.9 billion to $778.0 billion. In this timeframe, the GDIS grew faster than the rest of the economy, increasing its share of U.S. Gross Domestic Product (GDP) by 50%, from 1.9% to 2.9%. In FY2023, the 13 principal statistical agencies—which is just part of the federal data collection—had a combined budget of roughly $3.5 billion.7 There are numerous other uses by businesses and governments (see, e.g., Department of Commerce, 2014; Hughes-Cromwick & Coronado, 2019).
Assists Federal, State, and Local Government Action
Federal statistics provide high-quality, comparable information across the country. The American Community Survey (ACS), for example, provides key information that states and local governments use for disaster preparedness, economic development and workforce planning, public
___________________
4 https://www.census.gov/topics/income-poverty/poverty/guidance/data-sources.html
5 https://www.weforum.org/agenda/2016/01/a-global-economy-powered-by-data
6 More recent examples of such firms include Google, KPMG, and Zilllow, to name just a few.
7 See https://www.commerce.gov/news/blog/2024/12/revenue-industries-heavily-reliant-us-government-data. This paragraph and footnote were modified after release of the report to provide a more precise estimate of value and its context based on publication released on December 3, 2024.
health surveillance, and regional transportation planning (see National Academies of Sciences, Engineering, and Medicine [NASEM], 2019b; National Research Council [NRC], 2007a, 2013c). Data from the decennial census and the ACS are used to distribute hundreds of billions of dollars to states and localities for Medicaid, for education, and for housing, food assistance, veterans, and transportation programs.8 Additional examples of federal data products that inform state-level analyses and decisions include: state- and substate-level monthly employment and unemployment statistics produced by the Bureau of Labor Statistics;9 small-area estimates of poverty and population produced by the Census Bureau; state-level counts of deaths due to COVID-1910 and drug overdose;11 and state-level estimates of student performance based on data collected through the National Assessment of Educational Progress by the National Center for Education Statistics.
Monitors the Social and Economic Health of the Nation, States, and Localities
Regularly published social and economic indicators from statistical agencies are widely cited in the media and consulted by the public to identify trends and, when estimates are available for state and local areas, to compare across areas. Some examples include America’s Children: Key National Indicators of Well-Being from the Interagency Forum on Child and Family Statistics; the Condition of Education from the National Center for Education Statistics; Criminal Victimization from the Bureau of Justice Statistics; Statistics of Income from the IRS; Income and Poverty in the United States from the Census Bureau; Health United States from the National Center for Health Statistics; and Science and Engineering Indicators from the National Science Board and National Center for Science and Engineering Statistics.12
___________________
8 https://gwipp.gwu.edu/counting-dollars-2020-role-decennial-census-geographic-distribution-federal-funds#Briefs
9 BLS produces these statistics using Current Population Survey and Unemployment Insurance data collected by the Census Bureau and individual states, respectively. In particular, the Local Area Unemployment Statistics program produces monthly and annual employment, unemployment, and labor force data for census regions and divisions, states, counties, metropolitan areas, and many cities, by place of residence. See https://www.bls.gov/lau/.
10 https://www.cdc.gov/nchs/nvss/vsrr/COVID19/index.htm
11 https://www.cdc.gov/nchs/nvss/vsrr/drug-overdose-data.htm
12 https://www.childstats.gov/americaschildren/; https://nces.ed.gov/programs/coe/; https://www.bjs.gov/index.cfm?ty=tp&tid=9; https://www.irs.gov/statistics; https://www.census.gov/library/publications/2019/demo/p60-266.html; https://www.cdc.gov/nchs/hus/index.htm; and https://ncses.nsf.gov/pubs/nsb20201
Provides Evidence for Developing and Evaluating Public- and Private-Sector Programs
Data on the condition of housing and housing finance to inform housing policy come from the ongoing American Housing Survey (see NRC, 2008a). Statistics on the various types of energy used for heating, cooling, information technology, and other uses are provided by energy consumption surveys for commercial buildings and for residences (see NRC, 2012b). The Commission on Evidence-Based Policymaking highlighted a number of examples of how administrative data have been used to generate evidence to inform government policies, including substance abuse education and workforce investment (Commission on Evidence-Based Policymaking, 2017, pp. 9–10).
Provides Input to Social Science Research That Informs the Public
Many policy-relevant insights have resulted from statistical analysis of long-running federally funded surveys, including longitudinal surveys that follow individuals over time (see NRC, 2005). Some examples: the National Center for Education Statistics runs a number of longitudinal surveys following children through their K–12 education and postsecondary education and beyond to look at the transitions from high school and college to the labor force; the Bureau of Labor Statistics’ National Longitudinal Surveys follow young adults through their working lives to look at career paths; and the Health and Retirement Study of the University of Michigan, supported by the National Institute on Aging and the Social Security Administration, follows older adults through retirement to look at health and well-being.13
THE COSTS AND BENEFITS OF FEDERAL STATISTICS
The cost of federal statistical programs is a tiny fraction of overall U.S. federal spending. In fiscal year (FY) 2022, the combined budget request for all the major statistical agencies and other statistical programs in federal agencies (including the decennial census, economic, and agricultural censuses) totaled $7.1 billion.14 This amounted to about 0.03 percent of GDP15
___________________
13 https://nces.ed.gov/surveys/; https://www.bls.gov/nls/; http://hrsonline.isr.umich.edu/
14 See https://www.whitehouse.gov/wp-content/uploads/2024/02/statistical-programs-20212022.pdf. OMB includes all statistical programs with at least $3 million in estimated or direct funding in FY 2019, FY 2020, FY 2021, or FY 2022.
15 https://www.bea.gov/news/2023/gross-domestic-product-fourth-quarter-and-year-2022-third-estimate-gdp-industry-and#:~:text=Imports%20increased%20(table%202).,(tables%201%20and%203)
or about 0.1 percent of the total appropriated budget of about $6.3 trillion for the federal government,16 equivalent to about $21 per U.S. resident per year.17 This tiny fraction of funds allocated for federal statistical programs has been declining in recent years,18 despite ever-growing demands for more accessible and timely data (American Statistical Association, 2024).
The benefits from this investment in federal data collection and statistics permeate every corner of the United States. A recent publication from the Department of Commerce estimates the value of government data-intensive sector as $407.9 billion in 2012 rapidly growing to $778.0 billion in 2022.19
It is impossible to capture the full economic and societal value of having reliable data on economic, social, health, agricultural, industrial, and environmental characteristics of the country,20 although some experts have estimated the annual value of making federal data from statistical and program agencies “open,” that is, freely available to the public.21 The United Nations (2018) report, Recommendations for Promoting, Measuring, and Communicating the Value of Official Statistics,22 argued that official statistics have value far beyond their dollar worth. If the national government did not collect or acquire such data, the private sector might fill the breach—but likely at a greater cost to obtain data of comparable or less quality. Response rates would be lower compared with federal surveys, with no guarantee of continuance or continuity. The report went further to suggest the absence of government data collection and acquisition presented the possibility of a two-tier system whereby only those who could pay
___________________
16 https://www.cbo.gov/publication/58888
17 Based on a total U.S. resident population of 332,403,650. https://www.commerce.gov/news/blog/2022/01/us-population-estimated-332403650-jan-1-2022
18 As an example, this decline in response rates can be seen in the National Crime Victimization Survey (declines in the person-level response rate from 80% in 2011 to 50% in 2021) and the National Health Interview Survey (declines in the adult response rate from about 70% in 2011 to 50% in 2021). See (American Statistical Association, 2024).
19 See https://www.commerce.gov/news/blog/2024/12/revenue-industries-heavily-reliant-usgovernment-data. This paragraph and footnote were modified after release of the report to provide a more precise estimate of value and its context based on publication released on December 3, 2024.
20 https://www.commerce.gov/sites/default/files/migrated/reports/the-value-of-the-acs.pdf tells how the ACS is used by federal, state, and local governments and by businesses, school districts, and academic researchers. A panel from the American Enterprise Institute discussed the immense value of government data for commerce, as well as the private companies that, essentially, repackage, build upon, and sell government data. See https://www.aei.org/wpcontent/uploads/2017/02/170302-AEI-Vital-Role-of-Government-Statistics.pdf
21 Making data collected by the federal government available to the public at no cost in a machine-readable format without restrictions on its use is referred to as “open data.” See http://reports.opendataenterprise.org/2017OpenDataRT1-EconomicGrowth.pdf
22 https://www.unece.org/fileadmin/DAM/stats/publications/2018/ECECESSTAT20182.pdf
would have access to the data they need. An additional cost is the potential loss of trust and credibility (and therefore, value) where private sector companies producing the data could have a stake in the results (for example, grain companies reporting on crop production; energy companies on gas production; and tobacco companies on the health effects of tobacco use).
The fundamental characteristic of federal statistics as a public good (see Box 2-1) and the demonstrated policy, planning, research, and informational value of today’s portfolio of statistical programs justify adequate budgets for federal statistics. Such funding needs to provide for research and development for continuous improvement in relevance, accuracy, timeliness, and accessibility (see Practice 5). In turn, it is incumbent on federal statistical agencies to communicate the value of their programs to policymakers and others and to analyze the cost-effectiveness and value of their programs on a continuing basis so that they can ensure the best return possible on the tax dollars invested in them.
BOX 2-1
Federal Statistics as a Public Good
Familiar public goods, which benefit everyone in a country but cannot readily be provided through the marketplace, include the judiciary and the national defense. Formally, a public good has two components (see Gravelle & Rees, 2004). First, a public good must be nonrivalrous—that is, when one individual consumes it, there is no actual or potential reduction in the amount available for another individual to consume. A public good must also be nonexcludable, that is, it must be difficult for a business or other private entity to try to establish a market for such a good that is open only to those willing to pay the price. Federal statistics are essentially a public good—they satisfy the first component and practically speaking satisfy the second (NRC, 1999, Chapter 2).
In regard to the second component, a chief virtue of official federal statistics (e.g., the unemployment rate) is that they are designed to be uniquely representative of the whole country and/or the segments they cover. A private entity would face high costs and an inadequate return from providing such statistics on an ongoing basis. Consider the many examples of statistical information provided by businesses, academic institutions, and other nongovernmental organizations. The information may be based on a survey or another data source, and it may add value to an underlying federal data series. Yet nongovernmental organizations often do not attempt to gather information and provide timely, ongoing statistics on the scale necessary to represent the entire population of persons or organizations. Even the largest “Big Data” series often exclude the most vulnerable members
of society who may not have cell phones or do not subscribe to or purchase a company’s services.
Major federal household and business surveys and censuses provide vitally important information for a broad range of data users. Other federal statistical programs are important for scientific research and program evaluation (e.g., longitudinal surveys) and may have a specialized base of users. Both are costly to carry out. And, in either case, nongovernmental organizations are not likely to view them as viable business propositions. Business reasons may also lead a nongovernmental data holder to modify or abandon a useful statistical series.
Moreover, private-sector series often depend on federal statistics in various ways. For example, private-sector price indexes based on web-scraping (e.g., the MIT Billion Prices Indexa) have used federal CPI information for market-basket weights and benchmarking, and public opinion polls and marketing surveys use federal statistics on gender, age, ethnicity, and other characteristics to adjust the raw data to represent population groups. Thus, federal statistics not only are useful and often indispensable themselves, but they also are a necessary ingredient to many other data products and services that fulfill information needs.
__________________
