Data Collection:

Opportunities to Improve USDA's Farm Costs and Returns Survey

RCED-92-175: Published: Jul 30, 1992. Publicly Released: Aug 11, 1992.

Additional Materials:

Contact:

John W. Harman
(202) 512-5138
contact@gao.gov

 

Office of Public Affairs
(202) 512-4800
youngc1@gao.gov

Pursuant to a congressional request, GAO reviewed the Department of Agriculture's (USDA) Farm Costs and Returns Survey (FCRS), focusing on: (1) FCRS scope and design, as they affect the quality and reliability of the data generated; (2) USDA procedures for granting access to unpublished data; and (3) USDA communication with users of FCRS data.

GAO found that: (1) USDA methods for conducting FCRS may limit the quality and reliability of data for estimating farmers' expenditures and incomes; (2) the budget for producing and administering FCRS was $6.24 million; (3) FCRS response rates are overstated by 6 percent because USDA includes individuals who do not qualify as farmers; (4) failure to determine whether nonrespondents differ from respondents possibly generates biased results; (5) the policy of imputing answers to unanswered questions may lead to inaccurate data; (6) the USDA definition of farms results in an inefficient use of resources; (7) the USDA policy on access to unpublished data causes confusion among data users; and (8) USDA communication with land grant university researchers and other users needs improvement.

Recommendations for Executive Action

  1. Status: Closed - Implemented

    Comments: USDA plans to develop written criteria on how decisions will be made for granting or denying access to specific commodity data. Data related to the structure and operation of the farm business will be made available independently from data on enterprise costs and returns.

    Recommendation: To reduce the concerns that currently exist regarding the granting of and limitations on access to unpublished FCRS data, the Secretary of Agriculture should establish written criteria that clearly state how decisions for granting or denying access to data on specific commodities are made.

    Agency Affected: Department of Agriculture

  2. Status: Closed - Implemented

    Comments: USDA plans to work more closely with data users by alerting invitees when economic statistics are the focus of data user meetings and by presenting FCRS information at several economics groups' meetings in 1993 and 1994.

    Recommendation: To close the communication gap that currently exists between USDA and non-USDA users of FCRS data, the Secretary of Agriculture should establish with data users additional lines of communication devoted specifically to discussing concerns about FCRS.

    Agency Affected: Department of Agriculture

  3. Status: Closed - Implemented

    Comments: Beginning with the 1993 survey report and all future releases, USDA: (1) will publish a response rate that excludes those who do not qualify as farmers; (2) will publish more detail on the procedures used to edit or impute missing data and explain the effects of editing and imputation on survey data; and (3) has taken steps to reduce the number of nonrespondents and to reduce the bias in the survey data caused by nonrespondents.

    Recommendation: To better inform all users of FCRS about the data's integrity, the Secretary of Agriculture should, after taking the steps GAO recommended for improvements, publish information on the methodology USDA used to calculate response rates, the level of data editing/imputation that exists, and the similarity between respondents and nonrespondents, together with information on any bias in FCRS estimates.

    Agency Affected: Department of Agriculture

  4. Status: Closed - Not Implemented

    Comments: USDA stated it cannot change the definition of a farm because the Secretary of Commerce sets the definition. USDA further stated that information on small farms is required for farm income and costs of production estimates. USDA restratified its sampling procedures to more efficiently sample smaller and middle-size farms and is testing a shorter questionnaire to reduce costs.

    Recommendation: So USDA can improve the overall quality and reliability of FCRS and the estimates it generates, the Secretary of Agriculture should change the definition of a farm to eliminate from FCRS those small farms that have little impact on the overall financial conditions of U.S. farms or develop less costly methods for obtaining data on small farms less frequently, such as using a shortened mailed version of FCRS or a telephone survey once every 5 years.

    Agency Affected: Department of Agriculture

  5. Status: Closed - Implemented

    Comments: Beginning with the 1993 survey report, NASS will publish more detail on the procedures used to edit or impute missing data and explain the effects of editing and imputation on survey data. ERS will make its procedures and measurements of imputation more readily available in all 1993 releases that are based on FCRS data.

    Recommendation: So USDA can improve the overall quality and reliability of FCRS and the estimates it generates, the Secretary of Agriculture should evaluate the level of data editing/imputation that occurs in the survey.

    Agency Affected: Department of Agriculture

  6. Status: Closed - Implemented

    Comments: USDA has adopted nonresponse adjustment procedures, restratified the sample, and developed a shorter version of the questionnaire to improve response rates. These actions should reduce potential bias in the estimates.

    Recommendation: So USDA can improve the overall quality and reliability of FCRS and the estimates it generates, the Secretary of Agriculture should minimize potential bias in the estimates by determining if nonrespondents to FCRS are significantly different from respondents.

    Agency Affected: Department of Agriculture

  7. Status: Closed - Implemented

    Comments: USDA will calculate and publish a response rate that excludes individuals who do not qualify as farmers starting with publication of the 1993 report in September 1993.

    Recommendation: So USDA can improve the overall quality and reliability of FCRS and the estimates it generates, the Secretary of Agriculture should adopt a method of calculating response rates that excludes those individuals who do not qualify as farm operators.

    Agency Affected: Department of Agriculture

  8. Status: Closed - Implemented

    Comments: USDA plans to make FCRS survey data available at two regional sites on a pilot basis in 1993. After the pilot project, USDA will consider expanding the number of sites on the basis of data use and program resources.

    Recommendation: To reduce the concerns that currently exist regarding the granting of and limitations on access to unpublished FCRS data, the Secretary of Agriculture should work with non-USDA researchers to lessen the burden of using the data. In doing so, the Secretary may wish to conduct a pilot program that provides access to unpublished FCRS data at a location outside Washington, D.C., while still maintaining the same rules regarding confidentially that apply to accessing FCRS data at Economic Research Service headquarters.

    Agency Affected: Department of Agriculture

 

Explore the full database of GAO's Open Recommendations »

Jun 9, 2016

May 26, 2016

May 19, 2016

Apr 14, 2016

Jan 14, 2016

Jan 8, 2016

Jul 9, 2015

Jun 18, 2015

May 26, 2015

May 20, 2015

Looking for more? Browse all our products here