How to Read USDA ERS Agricultural Productivity — Total Factor Productivity, Output and Inputs

As of 7 November 2026, 00:26 KST (retrieved by this page at that time); latest data on the USDA Economic Research Service (ERS) page: Agricultural Productivity in the United States, national accounts for 1948 to 2023. Dates differ by source: the main page shows Updated: 7/20/2026, the national tables show Last Updated 6/2/2026, and the revision history dates the release June 30, 2026. The next national update is planned for the end of 2027 and is expected to run through 2025. Primary source: https://www.ers.usda.gov/data-products/agricultural-productivity-in-the-united-states^[1][3][4]

Comparison basis: all values are ERS's average annual growth rates in percent, rounded to two decimals; the changes are this page's subtractions of ERS's rounded subperiod averages, not ERS figures. "Data reported for each subperiod are average annual growth rates, so the different lengths of the subperiods do not generally affect comparisons across subperiods." The subperiods run from one business-cycle peak to the next, as dated by the National Bureau of Economic Research, so they are of unequal length.^[2][3]

In three lines

1. ERS measures total factor productivity (TFP) as farm output growth minus growth of all inputs combined.
2. Over 1948–2023, U.S. farm output grew 1.52% a year and inputs 0.12% a year, so TFP grew 1.40% a year.
3. In the latest subperiod, 2019–2023, TFP growth averaged 3.19% a year, up from 0.51% in 2007–2019.^[2][3]

Definition. "Total Factor Productivity (TFP) growth is the difference between growth of total output and growth of total inputs used." "Output is defined as gross production leaving the farm, as opposed to real value added." TFP here is therefore built on gross output leaving the farm sector, not on value added as in GDP-type measures.^[2][5]

Built from other agencies' data. The accounts are compiled from multiple sources, including survey data; they are not a single survey of farms. Per ERS's methods page and revision history, ERS combines USDA data with other agencies' series: hours worked from the Bureau of Economic Analysis's National Income and Product Accounts; demographic detail from the Census of Population and the Census Bureau's American Community Survey; Bureau of Labor Statistics special tabulations derived from the Current Population Survey, used as farm-sector totals, with the Current Population Survey also serving as the national-accounts source for self-employed and unpaid workers; and fuel prices from the Energy Information Administration's Monthly Energy Review. Changes in those sources can revise ERS's history, and the revision history notes: "Some revisions in factor input measures were made to reflect changes in source data."^[3][5]

Limits ERS states. "First, because of data limitations, this dataset does not account for output quality changes." "In the short-term, however, agricultural productivity change can also be affected by random events like weather."^[5]

Geography and years. The national accounts cover the U.S. farm sector for 1948–2023. State accounts cover the 48 contiguous States for 1960–2015 only: "State estimates are not made for Alaska and Hawaii, as some major data are not available for the entire period." "An update schedule for the State-level accounts has not been established."^[1][2]

Nearly three times vs 210 percent more. "The level of U.S. farm output was nearly three times its 1948 level in 2023, growing at an average annual rate of 1.52 percent over the period." The same page also says: "The level of U.S. Farm output in 2023 was 210 percent more than in 1948, growing at an average annual rate of 1.52 percent." ERS also describes this approximately as 'nearly three times'; its stated 210 percent increase corresponds to about 3.1 times the 1948 level, as does 1.52 percent a year compounded over 75 years (this page's calculation).^[2]

Subperiod in the text vs subperiod in the table. The productivity growth page mentions a 2019–2021 subperiod in which inputs fell 3.35 percent a year, pointing to a chart this page has not read. The subperiod table on the same page has no 2019–2021 row: its latest subperiod is 2019–2023, with input growth of 0.14 percent a year. The two figures cover different periods and should not be compared directly.^[3]

TFP vs output minus inputs. "Looking at the latest subperiods, measured average annual TFP growth increased from 0.51 percent in 2007–2019 to 3.19 percent per year over the period 2019–2023." From the table, 3.33 minus 0.14 gives 3.19 for 2019–2023, matching ERS; for 2007–2019, 0.53 minus 0.03 gives 0.50 against ERS's 0.51 (this page's subtractions), a gap within the rounding of two-decimal figures. On the output side ERS writes: "On the other hand, average annual growth in agricultural output was 3.33 percent, led by a rebound in crop production after a decline in the 2007-2019 period." This subtraction alone does not establish the cause of the change in TFP growth.^[2][3]

Input mix vs total inputs. "Over the 1948 to 2023 period, labor and land inputs declined by 76 and 28 percent, respectively, while intermediate inputs grew by 109 percent." Total inputs barely grew over the period, but their composition changed: falling labor and land, rising intermediate inputs. In this page's reading, a small total input change does not mean that farming used the same inputs as in 1948.^[3]

Next update. "The next update to the U.S. aggregate accounts—planned for the end of 2027—is expected to contain estimates of agricultural outputs, inputs, and total factor productivity indices for the aggregate farm sector (national) through 2025." So the series currently ends in 2023, and the next national edition is expected to reach 2025.^[1]

State results are older. "According to the estimates, every State exhibited a positive average annual productivity growth rate between 1960 and 2015, ranging from 0.23 percent for New Mexico to 1.95 percent for Ohio." These State figures stop in 2015 and come from a separate set of accounts.^[2]