addressed issues arising from the complexity of the Agricultural Resource Management Survey (ARMS): the complicated survey form, the problem of missing observations and nonresponse, the lack of comparability of measures across subsamples, and the difficulties in analysis imposed by the complex estimation and other procedures, among other topics.
At our second meeting, the panel focused on issues of statistical methodology and heard mainly from staff of NASS and ERS with responsibility for various aspects of survey design and data collection, estimation, and processing. In the discussion of questionnaire content, the panel learned from Doug Kleweno, Jim Johnson, Carol Jones, and Bill McBride about the determinants of the content of ARMS from the perspective of resource use, farm business performance, farm household well-being, and commodities, respectively. Two presentations were made on the process of turning concepts into questions, as well as the methods of testing and implementing the questionnaires by Danna Moore of the University of Washington, and Kathy Ott of NASS. Collection methodology and other statistical aspects of the survey were discussed by Bob Bass, Bill Iwig, Alix Riley, Richard Barton, and Chadd Crouse. In the final session, Phillip Kott, Bob Dubman, and Nigel Key introduced the challenging topic of methods of imputation and analysis of the ARMS data. The panel revisited the difficult issues involved in imputation with a complex survey design in our fourth meeting.
At the annual meeting of the American Agricultural Economics Association (AAEA) in August 2006, several panel members met informally with a large group of frequent ARMS data users to discuss accessibility of the data and data dissemination. The users also introduced issues regarding response rates, survey design, content, classification, editing and imputation, coding, cross-survey comparisons, and ongoing user forums. A summary of the discussion was reported to the entire panel at our third meeting.
The third meeting focused on the needs of major public- and private-sector data users and on data dissemination. Useful insights on data needs by major federal users were provided by Jim Langley, Joseph Cooper, and Joseph Glauber, representing the perspective of the Congressional Budget Office, the Council of Economic Advisers, and the Office of the Chief Economist, U.S. Department of Agriculture (USDA), respectively. Bob Young discussed the data needs of the general interest U.S. American Farm Bureau Federation, and Gary Adams laid out the requirements of the statistics program of the National Cotton Council, one of the major commodity-oriented interest groups. Discussing concepts and measures for understanding environmental analysis, commodities, and farm household well-being were Timothy Kiely and Arthur Grube, U.S. Environmental Protection Administration; Jeff Goebel, National Resources Conservation Service, USDA; and Roger Claasen, Bill McBride, Carol Jones, and Jim Johnson, ERS.