Data Envolopment Analysis (DEA) Methodology

An Application to Rice Production in Uruguay

Abstract

The objective of this work is to understand which practices lead a rice-field to be on the production frontier, and to identify the sources of productive inefficiency. A production frontier was defined using the Data Envelopment Analysis (DEA) methodolo-gy. Four different types of efficiency were estimated: technical, scale, allocative, and economic. In a second stage, the effect of management and environmental variables on the results of efficiency was estimated.The estimation was performed using a truncated regression, due to the nature of the data generating process. Data set comprises 573 rice fields from 26 growers over the period 2004/05 to 2008/09. Ten fields defined the production frontier considering constant returns to scale (efficient) with an average efficiency of 0.6. Under variable returns to scale the number of efficient fields increased to 25 and the average technical efficiency increased to 0.74. When considering the prices of inputs and output the number of economically efficient farms is three, and the average efficiency is 0.53. The results of the truncated regression on various efficiency indicators show that the most important determinants are the choice of varieties and soil type. Seeding characteristics, type of soil tillage, and time of tilling also have significant effects on some types of efficiency.