Update and Revision History

Agricultural Resource Management Survey (ARMS) data are usually released and/or revised twice a year. This schedule governs updates to the data in the tailored reports, data available in USDA, National Agricultural Statistics Service (NASS)/Economic Research Service (ERS) data enclaves, data used by Memorandum of Understanding holders, and data used by ERS researchers. The releases are scheduled for approximately May 31 and November 30 each year. Generally, the farm structure and financial data are updated/revised in the fall; any newly available crop surveys are added in the spring and/or fall.

This page compiles updates and revisions to the data and interface as of December 14, 2023.

December 14, 2023

This update includes the release of new farm and household finances data for 2022.

The ARMS webtool provides a reference data series for users to convert data on farm household income from nominal into real dollars. The Personal Consumption Expenditure (PCE) price index, available from the Bureau of Economic Analysis, measures price change from the perspective of the consumer. It includes both urban and rural consumers and is therefore appropriate for making farm household well-being comparisons over time.

How to use the PCE price index with USDA/ARMS household income historical data:

The price series provided is equal to 100 in the base year of 2017, so one can convert historical nominal dollar values from any year into 2017 values. Two hundred dollars in 1996 can be converted to 2017 dollars by multiplying the 1996 value by 100 and dividing by the 1996 PCE index value of 69.2, or $289 (289 = 200*100/69.2). If you want to change the base year to 2022, thereby expressing all dollar values in terms of 2022 inflation-adjusted dollars, divide each index value by the 2022 index value of 116 and then multiply by 100. The 1996 index in 2022 dollars is $59.7 (59.7 = 69.2/116*100), and $200 for 1996 in 2022 dollars is $335 (335 = 200*100/59.7).

December 15, 2022

This update includes the release of new farm and household finances data for 2021.

Data for 2019 and 2020 now include producer costs for participating in the Dairy Margin Coverage Program, a change that affected about eleven percent of the records. This change resulted in a reduction in the level of net cash farm income and net farm income of $40 dollars, or 0.1 percent,  on average. The correction to previously-reported data on related financial indicators for 2019–20, including expenses, household income, adjusted gross income, and income-based financial ratios, are likewise affected, although these changes are minor.

September 30, 2022

Errata: Some data within the Tailored Reports: Farm Structure and Finance in the ARMS Farm Financial and Crop Production Practices data product were misreported as "NA" following issues with the underlying data server, starting on or about September 20 through September 27. The issue was resolved on September 27, 2022, and the relevant data were restored.

December 16, 2021

This update includes the release of new farm and household finances data for 2020.

Previous data handling methods were improved with this release, correcting rounding errors that occurred in about five percent of the records previously posted for 1996-2019. The value was reduced by 1 in the case of variables expressed in dollars and was reduced by 0.1 in the case of share and percentage variables.

October 29, 2021

Errata: Following issues with the underlying data server, some data within the Tailored Reports: Farm Structure and Finance in the ARMS Farm Financial and Crop Production Practices data product were misreported as "NA,” starting on or about October 20, 2021, through October 28, 2021. The issue was resolved on October 28, 2021, and the relevant data restored.

May 27, 2021

Errata: Due to a server failure, some data within the Tailored Reports: Farm Structure and Finance in the ARMS Farm Financial and Crop Production Practices data product were misreported as "NA" starting on or about April 21, 2021, through May 26, 2021. The issue was resolved on May 26, 2021, and the relevant data restored.

December 18, 2020

This update includes the release of new farm and household finances data for 2019 and includes changes to improve internal consistency in data-handling methods.

• Previously posted estimates of Adjusted Gross Income (AGI) for 1996-2018 were updated to reflect a change in methods for how limitations to deduct retirement contributions as well other deductions are modeled across time. The new limitations result in an increase in estimated AGI of $4,400 at the national level (5.3 percent) compared to the previous method.

• Estimated value of production for farm operations in 2018 was revised downward by $3.83 billion (1.1 percent) due to corrections in processing the underlying survey data, especially potato production, and due to a restatement of a marketing contract record. Value of production is directly reported in the structural characteristics tailored report. Also, because value of production is used to classify the commodity specialization of operations restating value of production resulted in changes to the number of farms for several of the commodity specialization categories. The largest increase in operations within a single category was in operations classified as primarily specialty crop (fruit, vegetable, and nut) operations, raising the total of specialty crop operations by 1029, or 0.7 percent, to 148,931 operations, with smaller increases in the number of farms classified as "other field crops" and "general cash grains" operations, and also resulted in a corresponding decrease in the number of farms otherwise classified as "wheat," "corn," "soybeans," or "tobacco, cotton, or peanuts" operations.

• For 2018 data, estimates for four financial ratios included in the financial indicators report are updated with new information on agricultural wage rates released in the National Agricultural Statistics Service (NASS) Farm Labor survey. The agricultural wage rate is used as an implicit hourly wage for reported operator labor hours. The implicit labor cost is used in calculating the rate of return to assets and equity, operating profit margin, and economic cost-to-output ratio. The wage rate increase reduced the return to assets and equity by 0.1 and 0.2 percentage points, to 0.4 and 0.0 percent respectively, and reduced the operating profit margin by 1.2 percentage points to 3.1 percent. It increased the economic cost-to-output ratio by 2.3 percentage points to 111.5.

• For 2011 data, a procedure was implemented to impute reported government payment amounts by category for certain operations, resulting in a small (less than 0.1 percent) change in the number of operations receiving direct payments and countercyclical-type payments, as well as the average amounts of these types of payments, and the average gross cash and net cash farm income for recipients of these types of program payments. The total amount of government payments recorded in 2011 is unchanged.

• The standard method for calculating land tenure status for farm operations, which classified operations as either a full ownership operation, part owner operation, or full tenant operation, was extended to data from 1997 to 2003, changing the share of full owner operations downward by 0.4 percentage points, on average, and part owner operations up an equal amount. In determining the categorical status, the standard method accounts for acres rented in by operations as well as acres rented out.

December 10, 2019

Errata: On December 10, 2019, the data found in Tailored Reports: Farm Structure and Finance was revised to correct an error in the calculation of average and median net farm income estimates for 2012-17. Revised net farm income estimates impact all "subject" and "filter" selections for average and median net farm income, 2012-17. Estimates of U.S. average net farm income, for example, are revised downward by -4.5 to -8.8 percent across the 6 years, lowering these estimates by $1,858 to $4,409. The error was restricted to the underlying calculation of the net farm income variable alone; no other data values were affected within the income statement. However, net farm income is a component in the calculation of a subset of financial ratios also found in the tailored reports. Thus, the reported rate of return on assets, rate of return on equity, operating profit margin, and term debt coverage ratio estimates for 2012-17 are revised as well.

December 10, 2019

This update includes the release of new farm and household finances data for 2018, contains new variables that are defined across the 1996-2018 period, and includes changes to improve internal consistency in data-handling methods.

New data in this release include:

• In the Government Payments tailored report, a revised classification of government payments from 1996 to the present into five separate program types (direct payments, countercyclical-type payments, marketing loan benefit payments, conservation payments, and other program payments). These categories are described more fully in The Evolving Distribution of Payments From Commodity, Conservation, and Federal Crop Insurance Programs (EIB-184, November 2017). The change will improve comparisons across years because the previous method for categorizing government programs referred to individual farm bill programs that did not persist across the entire time period.
• In the Farm Household Income tailored report, a new variable from 1996 onward showing a household's Adjusted Gross Income (AGI). Adjusted Gross Income is calculated using formulas and specifications found in IRS Form 1040 of each corresponding tax year. AGI as reported in the table includes deductions for self-employed health insurance and self-employment taxes. Adjusted Gross Income as calculated from ARMS data is described more fully in Estimated Effects of the Tax Cuts and Jobs Act on Farms and Farm Households (ERR-252, June 2018).
• Within the Farm Business Income Statement tailored report, a new variable is shown from 2012 onward for Adjusted Breeding Livestock Income. The variable is one of five variables that account for the difference between net cash farm income and net farm income. For operations with income from the sale of breeding livestock, net cash farm income includes the cash receipts from the sale of breeding livestock, while net farm income includes only the realized gain or loss from the sale of breeding livestock, calculated as cash receipts minus acquisition costs. All five variables accounting for the difference between net cash farm income and net farm income in the Farm Business income statement (nonmoney income, value of inventory change, depreciation, labor non-cash benefits, and adjusted breeding livestock income) are show in the farm business income statement beginning with this release.

Additionally, this update improves data-handling methods for improved consistency across time.

• ERS' standard imputation method for handling missing survey data on farm debt for 2012 to the present was extended back to 2009 data. The change in methods resulted in an increase in total debt by 7 percent, on average, across 2009-11, and also affect debt-based financial ratios including the current ratio, the working capital-to-expense ratio, the debt-to-asset ratio, and the term debt coverage ratio.
• Prevailing interest rate data from Chicago Federal Reserve's AgLetter was used to replace multiple sources of interest rate data across 1996-2018. The difference in interest rates used for formula calculations averaged 65 basis points (0.65 percentage points) per year. This data is used in calculating Financial Ratios and Debt Repayment Capacity tailored reports from 1996 to 2008, and market interest rates are also used to impute debt appearing in the Farm Business Balance Sheet and the Operator Household Balance Sheet from 1996 to 2018. Debt-based financial ratios including the current ratio, the working capital-to-expense ratio, the debt-to-asset ratio, and the term debt coverage ratio are similarly impacted.
• In the Structural Characteristics tailored report, the standard method for calculating the classification of farms by their land tenure categories "including a full owner, part owner, and full tenant category" was consistently applied to all farm operations across time, changing previously-reported values for 1997 and 2004-06. The standard method takes into account acres rented in by operators as well as acres rented out. From 3 to 5 percent of farms switched from their previous categories as a result of changes in 1997 and 2004-06.

August 29, 2019

The 2017 farm and household finances data were revised. Revisions reflect adjusted survey weights based on the 2017 Census of Agriculture farm numbers and revised acreage and production metrics from NASS. The total number of farms shown in the Tailored Reports listing has changed from 2,040,262 originally released for 2017, to 2,033,724 farms after revision.

Additionally, four variables in the "Farm Business Income Statement" were removed because they collectively showed an incomplete measure of the difference between net cash farm income and net farm income. Although these variables (farm depreciation, non-cash labor benefits, value of inventory change, and nonmoney income) were listed as if they accounted for the entire difference between net cash farm income and net farm income, they in fact only account for a portion of the difference between the two measures. A more complete accounting is in development.

July 10, 2019

A new configuration of the farm structure and finance tailored reports resulted in a few cosmetic changes to the web-based interface. Some users may note changes to color and icon displays. The underlying ARMS farm financial data are unchanged from the December 12, 2018 release.

December 12, 2018

Added 2017 data for farm and household finances. The 2017 ARMS data were added to the existing years of data (1996-2016) for the Farm and Household Finance Tailored Reports in a new data dissemination tool developed in response to stakeholder feedback. This tool allows for more interaction with the ARMS data, includes charting capability, and has an application program interface (API) for direct queries. If users have any questions regarding the content or capabilities of the tool please contact the ERS ARMS Team.

Revised 1996-2016 farm and household finances data. This release implements new procedures, resulting in changes to some historical values back to 1996. First, revised respondent confidentiality protection procedures implemented by NASS and ERS reduced the number of publishable estimates. Second, the procedure now tests for statistical dominance (an issue for small sample sizes) in both the numerator and denominator of ratios, while previous versions tested only for dominance in the numerator. This rule change limited the release of some ratio estimates. Third, the relative standard error (RSE) is reported for means and ratios but is no longer reported for number of farms, based on the statistical properties of the delete-a-group jackknife estimator implemented by ERS.

Finally, the tool includes median values alongside mean estimate values, wherever available, back to 1996.

December 14, 2017

Added 2016 data for farm and household finances. The 2016 ARMS data were added to the existing years of data (1996-2015) for the Farm and Household Finance Tailored Reports. 

July 24, 2017

Four reports were removed from Tailored Reports: Crop Production Practices because of a problem in the computer program that processed the underlying data in surveys for 2011-13. The reports are:

1. Herbicide family application
2. Insecticide family application
3. Herbicide use by method
4. Insecticide use by method

The primary programming error has been identified—acres for sample farms that did not use a particular herbicide or insecticide family were not included in the denominator used to compute percent of acre-treatments. As a result, the sum of percent of herbicide/insecticide acre-treatments exceeded 100.

ERS will restore the reports when the problem has been resolved.

December 8, 2016

Added 2015 data for farm and household finances; revised selected historical data. The 2015 ARMS data were added to the existing years of data (1996-2014) for the Farm and Household Finance Tailored Reports. This release also incorporates revisions made by the U.S. Census Bureau to the series used for average U.S. household Income. The data for this series (which are used only in the calculation of "Farm operator household income: Percent of U.S. average household income" in Tailored Reports: Farm Structure and Finance) are drawn from the latest Current Population Reports, P60-252, Income and Poverty in the United States: 2015. In addition, selected Farm Structure and Finance reports for 2014 were revised to reflect quality improvements made this year to methods for imputing debt in instances where the underlying data were missing.

May 2, 2016

Revised 2012-14 farm and household finances data. On May 2, 2016, the ARMS data tool was reposted to correct two programming errors. The first inadvertently excluded cash sales of hay and sugarcane/sugar beets from ARMS crop cash income in 2013 and 2014. Correction of this error affected all financial measures based on crop cash sales and income, including financial ratios. For example, the correction increased the reported average per-farm net cash income by $3,200-$3,500 in each of those 2 years (about 7 percent per year). For just farm businesses, the increase in average per-farm net cash income was $6,300-$6,700 (5-6 percent). The average income from farming for households operating family farms increased between $2,300 and $2,700 per household (8-10 percent) in 2013 and 2014. The second error inadvertently double-counted cash sales for aquaculture, equine, bees, and sheep/goats in 2012; the correction reduces 2012 livestock income, gross cash income, net cash farm income, and net farm income by $2.9 billion for all farms and 2012 total family farm household income by $1.9 billion (0.8 percent). The 2012-14 ARMS data were also updated to reflect quality improvements that ERS analysts made in the underlying ARMS dataset since last year's release.

December 1, 2015

Added 2014 data for farm and household finances. The 2014 ARMS data were added to the existing years of data (1996-2013) for the Farm and Household Finance Tailored Reports. The 2014 ARMS survey was conducted as part of the Tenure, Ownership, and Transition of Agricultural Land (TOTAL) survey, which collected information from both farm operators and nonoperator landlords. The survey of farm operators had a significantly larger sample than previous ARMS surveys. Several format changes that make report contents more straightforward were also implemented.

Revised 2013 farm and household finances data. The 2013 ARMS data were updated to reflect quality improvements that ERS analysts have made in the underlying ARMS dataset since last year’s initial release.

September 30, 2015

Added a direct link to multiple State selection for Farm Finance Tailored Reports. Multiple-State selection for the Farm Finance Tailored Reports is now available from the report landing page. This feature allows side-by-side comparison of multiple survey States for a selected year in a single report. The original filter (by U.S. or single State) for selected years is still available. No updates were made to the underlying data. Look for the "Filter by multiple States" icon located beside the Tailored Report titles.

April 23, 2015

Added 2013 data for rice and peanuts. The 2013 ARMS Phase II Crop Production and Practices summary was released for rice and peanuts.

December 2, 2014

Added 2013 data for farm and household finances. The 2013 ARMS Phase III Farm Financial summary was released. All questionnaire versions are included. Multivariate imputation methods were adopted and applied to 2012-13 debt variables in the ARMS III questionnaire and other selected ERS imputed variables in the dataset. These changes are mainly a part of the Farm Business Balance Sheet.

Revised 2012 data reflects new ARMS weights. 2012 farm and household finances data were revised. Revisions reflect adjusted survey weights based on the 2012 Census of Agriculture farm numbers and revised acreage and production metrics from NASS. The total number of farms shown in the Tailored Reports listing has changed from 2,161,844 originally released for 2012, to 2,102,074 farms after revision. Multivariate imputation methods were adopted and applied to 2012-13 debt variables in the ARMS III questionnaire and other selected ERS imputed variables in the dataset. These changes are mainly a part of the Farm Business Balance Sheet.

Added 2012 data for soybeans. The 2012 ARMS Phase II Crop Production and Practices summary was released. Only high-level aggregated data is available for 2012 soybeans at this time. Categorical data for 2012 soybeans will be added in a future update.

Reloaded 2011 data for barley and sorghum. The 2011 ARMS Phase II summary was reloaded to ensure proper functionality of the Tailored Reports for 2011 sorghum and 2011 barley data.

November 26, 2013

Added 2012 data for farm and household finances. The 2012 all-version ARMS Phase III dataset was released. Note that differences between 2012 estimates and estimates from prior years reflect improved sample coverage (particularly among very small farms) and changes in survey methodology and implementation associated with the 2012 Agricultural Resource Management Survey, in addition to changes in the economic situation of farms. Estimates from the 2012 ARMS are preliminary and are expected to change when new weights, based on the 2012 Census of Agriculture, are released by NASS.

Began using the ERS revised farm typology for 2011 forward. The original typology is applied to 1996-2010. The ERS farm typology was revised in 2013. The revised typology (and corresponding collapsed typology, and definition of farm businesses) is applied to data in the tailored reports for 2011 forward. Data for 1996 through 2010 use the original typology. See the ERS report for more information:

• Updating the ERS Farm Typology (EIB-110, April 2013)

Re-summarized all data for 1996-2011. Many small updates were researched and implemented in the ARMS III datasets since the last release. Changes were made to the underlying program procedures, record data accuracy, etc. Production specification categories for California and Florida were adjusted to reflect the unique mix of farm types in those States. To refresh the complete summary data and ensure that all updates are available to the public, 1996-2011 data were re-summarized and loaded for this release of the tailored reports. All users of the historic data should consider reloading data from this release so the most up-to-date numbers are used for analysis.

Correction of Government Payments table 1996-2011. Twenty-six variables in the Government Payments table were updated and recalculated due to a variable truncation problem within the summary programs used to produce the tailored reports. Anyone using these data should refresh all years in their work files.

November 27, 2012

Added 2011 data for farm and household finances. The 2011 all-version ARMS Phase III dataset was released.

June 13, 2012

Corrections to fertilizer timing data in the Crop Production Practices surveys for 2006-10 surveyed crops. Removal, until further notice, of 1996-2005 fertilizer timing data. Revised data describe the timing of nitrogen, phosphate, and potash applications from the 2006 soybean, 2007 cotton, 2009 wheat, and 2010 corn surveys. Revisions correct a coding error that resulted in some applications of fertilizers being double-counted, thereby overestimating fertilizer applications by season. Timing of fertilizer applications data for all other surveys/years has been removed, and will be revised and loaded back into the tailored reports once verified.

November 30, 2011

Added 2010 data for farm and household finances. The 2010 all version ARMS Phase III dataset was released. Notable additions to the 2010 Version 1 questionnaire include detailed questions on rented land in section A and practices for major crops in section B. Section I has major changes to questions on the Internet, input purchases, direct sales, the Average Crop Revenue Election (ACRE) program, and base acres (dropped).

Added 2010 corn production practices data. The 2010 ARMS Phase II dataset on corn producers was released.

Added comparability between conventional and organic production systems. For the 2006 soybean, 2009 wheat, and 2010 corn production practices surveys, organic operations were oversampled. In this ARMS data addition, most production practice tables can be sorted into conventional and organic systems, which allows easy comparison of statistics between the two.