Effects of Geographical Stratification in a Farm Accountancy Data Network on the Accuracy of the Estimates

This study analyses the effects on the accuracy of the estimates produced by an additional stratification of a sample, in this case the sample of the Farm Accountancy Data Network (FADN) of Navarra. This is carried out by studying the accuracy of the total standard gross margin estimator. For the analysis, the variances of the estimator were compared for two samples: one stratified according to type of farming and economic size, and the other with the addition of a geographical stratification. The conclusion of the analysis is that for samples whose size is similar to that of the FADN of Navarra (400 farms), accuracy is not improved by a geographical stratification. This result, a particular case in which stratification does not improve the accuracy of the estimates, may be clarified by the theoretical considerations expounded in this paper.