Explanatory Model for the Estimation of the Minimum Monoculture Planted Area for Coffee that would allow reaching a Living Income

Abstract

This study's objective was the construction of an explicative model to determine the minimum area necessary to be planted in coffee, to afford a living income, on monoculture coffee farms. In the sustainable coffee production scenario, the annual net income of families is evaluated in function of the achievement of a decent standard of living for all of its members. The metric which makes this possible is called living income. Unfortunately, among small coffee farmers, there is a large gap between real net income and living income, which is a multidimensional phenomenon, affected by exogenous, difficult-to-control factors, such as area cultivated in coffee and its sale price, in addition to variables derived from productive models employed, such as productivity, costs, and farm food production. Due to it being a multifactorial phenomenon, data mining algorithms were applied (decision tree and random forest), in order to classify those farms that would close the gap between net and living income. Model iterations were performed on a dataset that gathered historical production cost, income, and cultivated area information from farms located in four Colombian departments. It was concluded that the variables which explain the gap in a monoculture coffee system are as follows: Area planted in coffee, productivity, and sale price.