The Possibilities and Limitations of Objective Sampling in Strengthening Agricultural Statistics

The ever-increasing demand for more and better statistics relating to agriculture in its many phases can be met by the two governmental agencies responsible for these statistics, (The Agricultural Marketing Service of the U. S. Department of Agriculture and the Bureau of the Census of the Department of Commerce) by the development and practical use of objective methods of sampling and, estimation. Three specific suggestions for using objective sampling methods in strengthening agricultural statistics were made by the writer in a previous paper, as follows: (1) Use a split schedule in Census enumerations to broaden the scope of basic agricultural statistics by from 50 to 200 percent, (2) Take a rotating partial census in intercensal years in commercial fruit and truck areas to obtain more reliable and more useful basic statistics regarding these specialty crops than is possible in connection with a general enumeration, (3) Take a nation-wide annual enumerative sample of individual farms covering crop acreages and production, livestock numbers and production, etc. The need for these improved methods has been developed more fully in the present paper, in which a critical description has been given of some of the more important methods of sampling and estimating now used by the U. S. Department of Agriculture and of some of the results obtained in developing more objective sampling methods.