The Census of Agriculture accounts for all U.S. farms and ranches and the people who operate them. The Census, taken only once every five years, looks at land use and ownership, operator characteristics, production practices, income and expenditures.
The Census provides the only source of uniform, comprehensive and impartial agricultural data for every county in the nation. Through the Census, producers can show the nation the value and importance of agriculture, and they can help influence the decisions that will shape the future of American agriculture for years to come.
Census data are used by all those who serve farmers and rural communities – federal, state and local governments, agribusinesses, trade associations and many others.
—Farmers and ranchers can use Census data to help make informed decisions about the future of their own operations.
—Companies and cooperatives use the facts and figures to determine the locations of facilities that will serve agricultural producers.
—Community planners use the information to target needed services to rural residents.
—Legislators use the numbers from the Census when shaping farm policies and programs.
Report forms for the 2012 Census of Agriculture were mailed to farm and ranch operators in late December 2012 to collect data for the 2012 calendar year. Completed forms were due by February 4, 2013. Additional mailings were sent around February 14 and March 20 to farmers and ranchers who had not responded. There was also phone enumeration and personal interviews with producers across the country.
Yes. A law (Title 7, U.S. Code, and CIPSEA, Public Law 107-347) guarantees to all respondents that their individual information will remain confidential. NASS uses the information only for statistical purposes and publishes data only in tabulated totals. The report cannot be used for purposes of taxation, investigation or regulation. The privacy of individual Census records is also protected from disclosure through the Freedom of Information Act.
NASS will release final Census data at noon EST on May 2, 2014. The agency will present the results at a live webcast and publish them online at www.agcensus.usda.gov. To learn more about the webcast visit the Census Newsroom.
The Census of Agriculture required diligent analysis of the data to provide the most accurate, reliable and useful statistics about the agriculture industry. The work stoppage caused by the lapse in federal funding in October 2013 caused NASS to delay its data analysis, which in effect caused a delay in 2012 Census of Agriculture publication. NASS released a first look at the Census with preliminary data on February 20, 2014. The agency will release the full report on May 2, 2014 at noon EST.
Census data that were presented on February 20, 2014, were preliminary because the comprehensive Census review of all items to the county level was not complete. These data are therefore subject to limited change when NASS releases final Census results on May 2, 2014.
No, not all data will be available. Any tabulated item that could potentially identify an individual producer or operation is suppressed and coded with a ‘D’. For example, if there is only one producer of a particular commodity in a county, publishing the data would identify the operation and would not be published. A dash represents zero. For more information refer to the Introduction section of the Census publication.
Yes. In the full 2012 Census publication, NASS is looking forward to reporting many changes in agriculture over the past 5 years and releasing first-time information about bio-mass production, Internet access, agroforestry and more.
We provide several options for accessing the 2012 Census of Agriculture data. PDF version of the Census will be accessible online. Final data will also be available via a searchable database called QuickStats. Visit www.nass.usda.gov or www.agcensus.usda.gov to access information online. The data also are available through the NASS office in your state, many depository libraries, universities and other state government offices. For additional information on the Census of Agriculture or other NASS surveys, call the Agricultural Statistics Hotline 800-727-9540.
A farm is defined as any place that produced and sold, or normally would have sold, $1,000 or more of agricultural products during the Census year. If your farm meets this benchmark you can register online.
In statistics, methodology is the survey processes by which data are collected, analyzed and summarized.
Most of the methodology is the same as that used in 2007. However, enhancements to the methodology are always made from census to census. In 2012, we improved our outreach and awareness efforts to encourage producers to respond to the Census. Despite these and other efforts, over time we have seen declining response rates with the Census of Agriculture. This type of decline is being experienced across the research and survey community in all fields. In 2012, we used capture-recapture methodology, an accepted statistical methodology, to account for those who did not respond, under-coverage, and misclassification as to whether a property is a farm or not. We document our methodology thoroughly in Appendix A of the Census.
In an ideal world, response to a Census would be 100 percent, but that is not realistic today. This means that we need to account for those who did not respond. To make up for those farmers who did not take part in the Census, NASS used statistical methodology to correct for under-coverage (farms not on the original list), non-response (people not returning their Census questionnaires), and misclassification (whether an operation is correctly classified as a farm or not). The uncertainty these adjustments may introduce causes us to not know the exact numbers. However, we can quantify the uncertainty we have about the numbers. This measure of relative reliability is published at all geographic levels as the coefficient of variation.
As part of this process, we compare Census responses to data NASS already has and other known information that we call administrative data. For example, in June we conduct a survey that is a sample of farms the United States. By matching information from this survey with information from the Census, we can develop statistical models. These models take into account the size of a farm (in terms of both its agricultural sales and the land area it covers), the age of an operator, the type of farm, and a number of other features of an operation. Using these models, we can adjust for under-coverage, non-response, and misclassification to develop accurate and reliable estimates for U.S. agriculture.
The source of Census information is farmers, ranchers and producers who are closest to the information. They, and the organizations that serve them, use the information and understand its value. However, errors can occur on the response forms. NASS reviews all submitted forms and follows up on any entry that looks out of line. We may contact a producer to verify information and/or compare information to existing known data, all to ensure the most accurate information.
NASS then uses statistical methodology to correct for under-coverage (farms not on the original list), non-response (people not returning their census questionnaires), and misclassification (whether an operation is correctly classified as a farm or as a nonfarm). The uncertainty these adjustments introduce causes us to not know the exact numbers. However, we can quantify the uncertainty we have about the numbers. This measure of relative reliability, which is known as the coefficient of variation, is published at all geographic levels. With the 2012 Census, NASS is publishing a measure of uncertainty with all estimates at the national, state, and county level. This measure of relative reliability increases transparency and data usability.
Coefficient of variation (CV) provides a measure of uncertainty of an estimate. The lower the coefficient of variation, the higher the reliability of the estimate. For the Census, it means that those using the data can assess the comparable reliability of the Census estimates. By publishing the CV, NASS is increasing transparency and data usability down to the county level.
Yes, the numbers are comparable. The results of each Census of Agriculture represent U.S. agriculture at a point in time. Comparing the results from two or more Censuses can reveal trends as well as changes and new developments in the industry. Throughout each Census cycle, we evaluate our processes, apply new technologies, and make improvements in our methodology so that we can account for all agricultural operations in the nation.
As NASS continues to analyze data to the county level, changes in the initial estimates may occur from the preliminary report to the final report.
NASS takes great pains to produce the most accurate and useful data available. We have statistically valid procedures in place to collect, analyze, summarize and report the results of our surveys, including the Census of Agriculture. We check the reported information against other known administrative data and double check if it looks incorrect. Staff who have local knowledge and work in each state review the data as well. With all of that said, these are statistical estimates and therefore we publish the coefficient of variation as a measure of the uncertainty associated with each estimate. With that measure, people can better understand and use the data.
When we say a change is significant, we are using the statistical meaning of significant. Generally, this means that the change from 2007 to 2012 was two or more standard errors. However, there is a difference between statistical significance and practical significance. Some changes that are less than two standard errors may be of high practical importance; other changes that are more than two standard errors may be practically inconsequential. The data user must assess the practical significance of a change.
After the 2007 Census of Agriculture, NASS evaluated the performance of the Census of Agriculture form for operators on American Indian reservations in three states: Arizona, New Mexico, and Utah. We designed a special form to collect data for the unique nature of agriculture on reservations in these three states. An additional land tenure category; land used under a permit on an American Indian reservation, was added to the report form. In the 2012 Census, American Indian farm operators are classified as tenant farmers when the only land they operate is permit land on the American Indian reservation, whereas in the 2007 Census, we classified them as full owners. This updated classification has created several data anomalies in the land tenure summaries that are part of the 2012 Census of Agriculture released on May 2, 2014. NASS is currently reviewing these data and plans to address the issue in the American Indian Reservation Report in August 2014.
Appendix A of the 2012 Census of Agriculture explains the methodology we used to collect, analyze and summarize the information in depth. You can learn more about the statistical method we used to correct for under-coverage (farms not on the original list), non-response (people not returning their census questionnaires), and misclassification (whether an operation is correctly classified as a farm or not) by reading Capture-Recapture in the 2012 Census of Agriculture: A Beginner’s Guide.
If you have a concern about state or county-level data, please contact the appropriate USDA NASS regional office (www.nass.usda.gov/statistics_by_state/rfo/index.php). If your concern is about national-level data, please contact email@example.com or 800-727-9540. We can direct you to the appropriate subject matter expert. Finally, if you want to make an official request to correct data, please follow the instructions posted online at www.nass.usda.gov/About_NASS/Information_Quality_Guidelines/
Last Modified: 10/05/2015