Reliability of agricultural statistics in developing countries: reflections from a comprehensive village survey on crop area statistics in India

Abstract

Despite the importance of agriculture in developing countries, and the general recognition of the need for strengthening data quality, very few studies examine the quality of available data and the data generation methods in agriculture. In this paper, we use data from an extensive deployment of geospatial technology, administrated concurrently alongside the conventional method in the Indian state of Karnataka, to assess the discrepancy between methods in terms of the magnitude of difference in the crop area, type and number of crops grown. The crop area estimates based on alternative method, utilising the geospatial technology, exceeded that from the estimates based on conventional method. Conventional method is unable to respond quickly to changes in the cropping pattern and therefore, do not record accurately the area under high value cash crops. This has wider implications for commercializing agriculture and delivery of farm credit and insurance services in the developing countries. Some research and policy implications are discussed.