Study of Statistical Methodology in farm cost and management Surveys

Abstract

Although the studies on the various aspects of farm management started in the western countries in the beginning of the present century and have since been getting considerable importance, in India no general efforts seem to have been made in this direction. Whatever data were collected contained information on some specific items only. No broad-based effort was made till very recently when the scheme ‘studies in Economics of Farm Management’ was started, with the objective of collecting primary data on various cost and production aspects of the farm management and for comparison of alternative methods of collecting the data. The present study, described in the previous sections, deals with four aspects of this, 1. In an effort to find out the functional relationship existing between the factors of input on the one hand and output on the other, various production functions have been discussed. Of these, the Dobb-Douglas production function has been fitted to all the crops although, sometimes, the quadratic function has also been tried. We do not succeed in arriving at a close relationship for the reasons that (i) The first year data (used here) may not be of the required standard and quality. (ii) The factors considered in the various functions may not be the ones actually contributing towards the production and that there are some ‘extraneous’ factors e.g. climate, soil-fertility etc. which need be studied. 2. In the case of comparison between the cost Accounting and Survey methods of collecting the primary data, the comparison was made in two ways. The first type of comparison was made on the basis of independent villages and the second one on the basis of common villages. The two types of comparison do not yield identical results, there being more results significant in the latter type which is to be expected there is no contradiction in the two types of comparison i.e. they give similar results. The methods were found to differ in almost all the factors but the difference is quite marked in the case of Bullock Labour/acre and Total Input/acre, the excess in favour of the Survey method. 3. In the case of estimation of sample size for the various strata individually as well jointly, sample sizes have been calculated for different standard errors viz. 5% and 1%. 4. In the case of estimation of gain in precision and usefulness of the stratification of holdings in a village as adopted in the survey, it has been found that there is practically no gain in the mode of stratification within the primary sampling units (as adopted). The possibility of the usefulness of stratification according to fixed size ranges needs to be explored.