Statistical and Spatial Analysis of Census Data for the Study of Family and Industrial Farming in Colombia

This article presents a study of the Third National Agricultural and Livestock Census data to examine the characteristics of family and industrial farming in Colombia using together, statistical and spatial analysis techniques. From the available anonymized census microdata, which was added up to a level of rural relevance called vereda, a selection of variables was made according to properties found out in the state of art as related to both agriculture types and to the established theoretical framework. From there on, a Principal Component Analysis allowed distinguishing association between the attributes of family and industrial agriculture, while an Exploratory Spatial Data Analysis lets identifying polarized geographical distributions of these two phenomena in the country. In this way, this study seeks to become a starting point and a benchmark for analyzing the Colombian agricultural sector, as well as demonstrating the importance of census information, data reduction and statistical and spatial exploration methods.