The 2025 Annual Update of the National Economic Accounts

By Lisa S. Mataloni and Stephanie H. McCulla  |  November 19, 2025

The U.S. Bureau of Economic Analysis (BEA) released its annual update of the National Economic Accounts (NEAs), which include the National Income and Product Accounts (NIPAs) and the Industry Economic Accounts (IEAs), on September 25. The Regional Economic Accounts (REAs) are harmonized with the NEAs, so improvements incorporated as part of the NEA annual update also impact the regional statistics. The annual update of the REAs was released on September 26.

The update of the NEAs covered the first quarter of 2020 through the first quarter of 2025 and resulted in revisions to gross domestic product (GDP), GDP by industry, gross domestic income (GDI), and related components. The reference year for index numbers and chained-dollar estimates remains 2017.

The impacts of the annual update on the NIPA and IEA estimates are summarized in the tables and charts provided in this article. Refer to “Information on 2025 Annual Updates to the National, Industry, and State and Local Economic Accounts” for additional background materials.

The updated NIPA and IEA estimates reflect the incorporation of newly available and revised source data as well as the adoption of improved estimation methods and, for quarterly and monthly measures, the incorporation of updated seasonal adjustment factors.

Major source data incorporated

In addition to newly available and revised source data, BEA used alternative source data for several measures as part of this annual update. BEA typically incorporates source data from several U.S. Census Bureau (Census) annual surveys, including the Annual Retail Trade Survey (ARTS), the Annual Wholesale Trade Survey, the Service Annual Survey (SAS), and the Annual Survey of Manufactures (ASM). These surveys and others have been replaced by Census' newly designed and streamlined Annual Integrated Economic Survey (AIES). In July 2025, Census began releasing data from the 2023 AIES data collection cycle, but the detailed data required for the NEAs were not available in time to incorporate into the 2025 annual update. Therefore, BEA used alternative sources, including the monthly or quarterly counterparts of these annual surveys—such as the Quarterly Services Survey (QSS)—to produce estimates for 2023–2024, as shown in table 1.

Additionally, BEA typically incorporates Internal Revenue Service (IRS) Statistics of Income (SOI) tabulations of tax returns for corporations, for sole proprietorships, and for partnerships during annual updates. BEA was unable to purchase SOI data for this year's update; consequently, previously incorporated SOI data underlie the measures for 2020–2021, revised SOI data underlie the measures for 2022, and measures for 2023–2024 are based on revised data from the Census Quarterly Financial Report as well as from regulatory agencies and public financial reports.

Finally, BEA has changed the data sources for other measures, including the following:

• Beginning with the first quarter of 2020, measures of private fixed investment in manufactured homes are based on monthly data from the Census Manufactured Housing Survey, which are, in turn, based on Institute for Building Technology and Safety (IBTS) data on shipments of manufactured homes. Previously, these measures were based on monthly shipments data directly from IBTS.
• Beginning with the first quarter of 2020, quarterly measures of personal consumption expenditures for employment agency services, a component of professional and other services, are based on Census QSS revenue data rather than on U.S. Bureau of Labor Statistics (BLS) Current Employment Statistics (CES) data. This change aligns the monthly and quarterly measures with the annual measures, which are based on Census SAS data, the annual counterpart of the QSS data.
• Beginning with the first quarter of 2024, the price indexes underlying real measures of private fixed investment in “other electric power structures,” gas and petroleum pipelines, and water and sewage structures are based on U.S. Bureau of Reclamation data on quarterly construction cost trends. Previously, the estimates were based on price data from the Handy-Whitman Index.
• Beginning with 2024, BEA discontinued its use of Polk vehicles-in-operation data and National Auto Auction Association data for annual measures of transactions in used motor vehicles. Instead, annual estimates are based on retail sales by vehicle dealers data from the Census Monthly Retail Trade Survey (MRTS). This change further harmonizes the NIPA and IEA estimates of transactions in used motor vehicles. Additionally, beginning with January 2025, the prices underlying real measures of net purchases of used motor vehicles are based on the BLS Consumer Price Index (CPI) for used autos and trucks to replace discontinued data on used auction prices from the Auto Dealers Exchange Services of America.
• Beginning with the first quarter of 2025, a composite index of BLS Producer Price Indexes (PPIs) and CES wage data is used for quarterly measures of real private fixed investment in telecommunications structures. Previously, the measures were based on the Engineering News-Record Construction Cost Index and the AUS-Turner Telephone Plant Index.

The major source data incorporated into the NEAs as part of this year's update are summarized in table 1, and additional information on the NIPA components affected by the incorporation of newly available and revised source data is provided in the table “NIPA Revisions: Components Detail and Major Source Data and Conceptual Changes Incorporated, 2020–2024,” available on the “Updates” page on BEA's website.

Source data that affected the estimates include the following:

• QSS data for 2023–2024 impacted estimates of consumer spending for services and estimates of private fixed investment in intellectual property products and equipment and in gross output for private services-producing industries. Estimates for 2020–2022 are based on previously incorporated Census SAS data.
• Revised SOI data for 2022 and extrapolations of SOI data for 2023–2024 impacted estimates of corporate profits, proprietors' income, and net interest and of GDP by industry estimates. SOI data were extrapolated using various indicators, including data from regulatory agencies, public financial reports, Compustat, the Federal Deposit Insurance Corporation (FDIC), and Census. Measures for 2020–2021 were based on previously incorporated SOI data.
• Revised Census Monthly Survey of Manufacturers' Shipments, Inventories, and Orders data for 2020–2024 impacted estimates of private investment in equipment and inventories and in gross output for manufacturing industries.
• Revised Census ASM data, benchmarked to the 2022 Economic Census, impacted estimates of private investment in inventories for 2020–2021.
• Restated Census ARTS data for 2020–2022 and new and revised MRTS data for 2023–2024 impacted estimates of consumer spending for goods and private inventory investment and estimates of retail trade output in the IEAs.¹
• Revised BLS Quarterly Census of Employment and Wages (QCEW) data for 2020–2024 impacted estimates of private compensation, and revised QCEW data for 2020–2023 and new QCEW data for 2024 impacted estimates of government compensation.
• Revised U.S. Department of Agriculture (USDA) farm statistics for 2020–2024 impacted estimates of farm output, inventory investment, and proprietors' income.
• Revised BEA U.S. International Transactions Accounts (ITAs) data for 2020–2024 impacted estimates of exports and imports of goods and services and income flows with the rest of the world.²
• Revised Census Annual Survey of State and Local Government Finances (GF) for fiscal years 2020–2022 and newly available GF data for 2023 impacted estimates of state and local government spending.

Methodology improvements³

In addition to the source data changes described above, the annual update incorporated improvements to estimating methodologies and to the presentation of the estimates, including the following:

• Detail on data centers. In recent years, there has been a sharp increase in the construction of data centers (structures that house computing infrastructure, including servers, storage devices, networking equipment, and other systems necessary for managing and storing data). To improve the estimates of private fixed investment in office structures, BEA now provides separate detail on “data centers” and “general, financial, and other” offices in NIPA table family 5.4, beginning with the first quarter of 2020. Estimates of investment in data centers are based on Census construction spending data from its Value Put in Place (VPIP) Survey; measures are deflated using a new composite price index that equally weights BLS PPIs for industrial buildings and warehouses. General, financial, and other office buildings continue to be deflated using the BLS PPI for new office building construction.
• Improved measures of special dividends. In accordance with updated guidance in the System of National Accounts 2025 (SNA 2025) and the Balance of Payments and International Investment Positions Manual, Seventh Edition (BPM7), BEA has improved its treatment of “super” dividends. Beginning with 2020, when a company's dividends are disproportionately large relative to its distributable income, the excess dividends are treated as financial transactions, which are excluded from the NIPAs.⁴ Previously, the full value of large dividend payments were included in the NIPAs as “special dividends” unless they resulted from events that were not related to production, such as the sale of all or some portion of a company, in which case they were excluded.
• Improved measures of research and development. Estimates of investment by banks in research and development (R&D) for the most recent year are improved by using projections from the National Science Foundation Business Enterprise Research and Development Survey. Previously, estimates for the most recent year were based on a composite indicator of data from annual financial reports (known as 10–K reports) submitted by publicly traded companies to the Securities and Exchange Commission and Census QSS data. Evidence suggests these composite data underestimated R&D performed by banks, which have little representation in the 10–K reports.

GDP and expenditure components

In general, the picture of the economy shown by the updated estimates is similar to the picture previously published (table 2). Over the recent period from 2019 to 2024, the average annual rate of change in real GDP was 2.4 percent, the same as previously published (chart 1).

Details on the revisions to the percent change in real GDP are presented in table 3A, and details on the contributions to those revisions are presented in table 3B. Details on revisions and contributions to revisions for each expenditure component of real GDP are presented in appendix A (tables A1–A12). As noted above, the major source data incorporated for each component of GDP and GDI are presented in the annual update table “NIPA Revisions: Components Detail and Major Source Data and Conceptual Changes Incorporated, 2020–2024.”

The percent change in real GDP was revised up for 2020 and 2021 and was unrevised for 2022–2024 (chart 2).

• For 2020, the change in real GDP was revised up by 0.1 percentage point, from −2.2 percent to −2.1 percent, primarily reflecting upward revisions to exports of services, based on revised data from BEA's ITAs, and private inventory investment (led by chemical manufacturing), based on revised ASM data.
• For 2021, the change in real GDP was revised up by 0.1 percentage point, from 6.1 to 6.2 percent, primarily reflecting upward revisions to private inventory investment (mainly durable-goods manufacturing), based on revised ASM data, and state and local government spending (led by investment in equipment), based on revised Census GF data.
• For 2022, the change in real GDP was unrevised at 2.5 percent, primarily reflecting an upward revision to private inventory investment (led by durable-goods manufacturing) and downward revisions to nonresidential fixed investment (mainly equipment) and consumer spending on services (led by household utilities). The upward revisions to private inventory investment and equipment investment were primarily based on data from the 2022 Economic Census. The downward revision to consumer spending for utilities was based primarily on Census GF data on state and local government utility sales and Energy Information Administration data on electricity and natural gas usage.
• For 2023, the change in real GDP was unrevised at 2.9 percent, primarily reflecting an upward revision to nonresidential fixed investment (led by structures) and a downward revision to state and local government spending. The revision to structures investment was based on revised VPIP data. The downward revision to state and local government spending was based on newly available GF data.
• For 2024, the change in real GDP was unrevised at 2.8 percent, primarily reflecting a downward revision to nonresidential fixed investment, led by structures, based on Census VPIP data, and an upward revision to consumer spending, led by spending on goods, based primarily on revised Census MRTS data.

[View larger chart]

[View larger chart]

Prices

Revisions to BEA's various price measures, such as gross domestic purchases, GDP, and PCE, were small and reflect revised and newly available source data and, for the most recent year (2024), the regular incorporation of annual weights (tables 4A and 4B). Prices for gross domestic purchases—a measure of the prices paid by consumers, businesses, and governments—increased at an average annual growth rate of 3.6 percent over the recent period from 2019 to 2024, the same as previously published (chart 1). The revisions to the percent change in prices and price contributions are presented in tables 4A and 4B, respectively.

Income

Revisions to the components of national income and GDI primarily reflect revised IRS SOI data, new and revised QCEW data, revised ITA data, and data from the U.S. Department of the Treasury. Revisions are most notable for 2022–2024 (chart 3 and table 5).

• For 2022, the downward revision primarily reflected a downward revision to net interest, based on revised SOI tabulations of corporate tax return data, that was partly offset by an upward revision to corporate profits, also based on revised SOI data.
• For 2023, the upward revision primarily reflected an upward revision to corporate profits, based on SOI data for 2022 extrapolated using Census Quarterly Financial Report data, regulatory agency data, and public financial reports on profits. This upward revision was partly offset by downward revisions to net interest and taxes on production and imports. The downward revision to net interest was primarily based on extrapolations of 2022 SOI data using Compustat data on interest expenses and FDIC data on interest income, and the downward revision to taxes on production and imports was based on new and revised Census State Tax Collection (STC) Survey data and new and revised Census GF data.
• For 2024, the downward revision reflected downward revisions to net interest, corporate profits, and taxes on production and imports. These were partly offset by upward revisions to rental income, proprietors' income, and compensation. The revision to net interest is primarily based on extrapolations of 2022 SOI data using Compustat data on interest expenses and FDIC data on interest income. The revision to corporate profits primarily reflects revised QFR data, regulatory agency data, and profits data from public financial reports. The revision to taxes on production and imports reflects new STC data. The revision to rental income reflects newly available data for 2023 on housing rents and occupied units from the Census American Community Survey. The revision to proprietors' income primarily reflects new USDA data.

[View larger chart]

The update had a notable impact on the statistical discrepancy (chart 4 and table 6). In theory, GDI should equal GDP, but in practice, they differ because their components are estimated using largely independent source data. The most notable revisions to the discrepancy are for 2022–2024. For 2022, the discrepancy was revised up from −$75.6 billion, or −0.3 percent of GDP, to −$0.4 billion, or less than 0.1 percent of GDP. In 2023, the discrepancy was revised up from $244.6 billion, or 0.9 percent of GDP, to $325.9 billion, or 1.2 percent of GDP. In 2024, the discrepancy was revised up from $192.2 billion, or 0.7 percent of GDP, to $296.3 billion, or 1.0 percent of GDP.

[View larger chart]

Measures of personal income were also impacted by newly available and revised source data (table 7), most notably in 2023 and 2024:

• For 2023, the upward revision was led by an upward revision to personal income receipts on assets, notably, personal dividend income, based primarily on revised SOI tabulations of corporate tax return data on dividend payments for 2022 and revised indicator data.
• For 2024, the upward revision reflected upward revisions to all major components, notably, personal income receipts on assets, rental income, and proprietors' income. The upward revision to personal income receipts, primarily personal dividend income, was based on revised SOI tabulations of corporate tax return data on dividend payments for 2022 and revised indicator data. The revision to rental income was based on new Federal Housing Finance Agency and new Federal Reserve Board mortgage data. The revision to proprietors' income was led by farm proprietors, based on new USDA data.
• The personal saving rate (personal saving as a percentage of disposable personal income) was unrevised for 2020 and revised up for 2021–2024. For 2021–2024, personal income was revised up, and personal current taxes were revised down.

GDP and GDI

Chart 5 shows revised and previously published percent changes for real GDP for the first quarter of 2021 through the first quarter of 2025; the chart excludes 2020 because the magnitude of the changes in GDP in that year obscures the changes in other years. With the updated estimates, the overall pattern of economic growth over this period remains unchanged. In 2020, real GDP decreased 5.2 percent in the first quarter (revised up 0.3 percentage point), decreased 28.0 percent in the second quarter (revised up 0.1 percentage point), increased 34.9 percent in the third quarter (revised down 0.3 percentage point), and increased 4.6 percent in the fourth quarter (revised up 0.2 percentage point).

[View larger chart]

Revisions to quarterly estimates of GDP were generally led by updated estimates of consumer spending, based primarily on revised ARTS data for 2020–2022 and revised MRTS data for 2023–2024.⁵ Revisions to estimates of private fixed investment were led by residential investment, based primarily on revised VPIP data, and investment in equipment, based primarily on revised data from the Census Manufacturers' Shipments, Inventories, and Orders Survey.⁶ Revisions to exports and imports were based primarily on revised BEA ITA data, and revisions to federal and state and local government spending were based primarily on new and revised budget data and revised Census GF data.

From the fourth quarter of 2019 to the first quarter of 2025, real GDP was revised up by less than 0.1 percentage point (chart 6). Real GDI over the same period was revised down from 2.3 percent to 2.2 percent, and the average of GDP and GDI was revised up by less than 0.1 percentage point (chart 6).

Prices

From the fourth quarter of 2019 to the first quarter of 2025, the average annual rate of change in the price index for gross domestic purchases was 3.7 percent, revised up 0.1 percentage point from the previously published estimate. Over the same period, the price index for PCE increased 3.7 percent, revised up by 0.1 percentage point, and the core PCE price index, which excludes food and energy, increased 3.5 percent, the same as previously published.

Quarterly revisions to the core PCE price index primarily reflect updated BLS CPIs (chart 7).

[View larger chart]

Business cycles

The percent change in real GDP during the pandemic-related recession from the fourth quarter of 2019 to the second quarter of 2020 was revised up by 0.1 percentage point from a decrease of 17.5 percent to a decrease of 17.4 percent (chart 8 and table 8). The recovery from the second quarter of 2020 to the first quarter of 2025 was unrevised at 4.5 percent.

[View larger chart]

[View larger chart]

The picture of GDP by industry—or value added, which measures an industry's contribution to GDP—was largely unchanged with the annual update. Revisions to annual percent changes in real GDP by industry for 2020–2024 are discussed below and presented in chart 9 and table 9. Revisions to industry contributions to the percent change in real GDP are presented in table 10.

Revisions to GDP by industry

The updated estimates reflect upward revisions to goods-producing industries, services-producing industries, and government (chart 9 and table 9). As with the revisions to the NIPAs, the revisions to the IEAs were largely driven by newly available and revised source data.

• For 2020, the change in real GDP was revised up 0.1 percentage point, from a decrease of 2.2 percent to a decrease of 2.1 percent, reflecting upward revisions to private goods-producing industries (led by nondurable-goods manufacturing) and private services-producing industries (led by real estate and rental and leasing and by finance and insurance). The direction of change was the same as previously published for all 22 major industry groups.
• For 2021, the increase in real GDP was revised up 0.1 percentage point, from 6.1 percent to 6.2 percent, primarily reflecting an upward revision to private services-producing industries (led by finance and insurance and by transportation services). The direction of change was the same as previously published for all 22 major industry groups.
• For 2022, the increase in real GDP was 2.5 percent, the same as previously published. An upward revision to services-producing industries (led by finance and insurance and by information) was offset by downward revisions to private goods-producing industries (led by both durable-goods and nondurable-goods manufacturing and by construction) and to state and local government. The direction of change was the same as previously published for 20 of 22 major industry groups.
• For 2023, the increase in real GDP was 2.9 percent, the same as previously published. An upward revision to goods-producing industries (led by construction, mining, and durable-goods manufacturing) was offset by a downward revision to state and local government. The direction of change was the same as previously published for 21 of 22 major industry groups.
• For 2024, the increase in real GDP was 2.8 percent, the same as previously published. Upward revisions to federal government and goods-producing industries (led by mining and nondurable-goods manufacturing) were offset by a downward revision to private services-producing industries (led by finance and insurance and by wholesale trade). The direction of change was the same as previously published for 20 of 22 major industry groups.

[View larger chart]

From 2019 to 2024, the average annual change in real GDP was 2.4 percent, the same as previously estimated (chart 10). Over the same period, goods-producing industries increased 1.2 percent, an upward revision of 0.1 percentage point from the previously published estimate. Services-producing industries increased 3.0 percent, an upward revision of 0.1 percentage point. Government increased 0.7 percent, a downward revision of 0.3 percentage point. The direction of change over the period was unrevised for all major industry groups (chart 11). The largest revision over the period was to the mining industry (revised up from an increase of 1.7 percent to an increase of 3.0 percent).

[View larger chart]

[View larger chart]

Revisions to gross output

Gross output is principally a measure of an industry's sales or receipts, which includes sales to final users in the economy (GDP) and sales to other industries (intermediate inputs). Real gross output decreased 3.1 percent in 2020, the same as previously published. The percent change in real gross output was revised up 0.1 percentage point to an increase of 6.9 percent for 2021, revised down 0.1 percentage point to an increase of 2.9 percent for 2022, revised up 0.3 percentage point to an increase of 2.6 percent for 2023, and revised down 0.1 percentage point to an increase of 2.3 percent for 2024 (table 11).