Modelling Investment Optimization on Smallholder Farms through Multiple Criteria Decision Making and Goal Programming: A Case Study from Ethiopia

Abstract

We use data from the Ethiopia Rural Household Survey and the Ethiopian Central Statistics Agency to demonstrate a set of techniques for estimating optimal investment allocation in smallholder farming. The approaches treat farming tasks, constraints, and investments as a portfolio problem, characterized by multiple competing objectives. We formulate several versions of the multi-objective problem and solve them in three alternative ways; 1) using a scalarized Markowitz portfolio optimization, 2) using a weighted goal programming model, and 3) a multi-horizon goal programming model, estimating all model parameters using real data. The main benefit of the goal programming formulation is the possibility to simplify in a single criterion problem complex situations in which the Decision Maker (DM) faces a trade-off between two or more objectives. We discuss the importance of portfolio allocations for smallholder farmers in minimizing risk and increasing return, and discuss how these approaches provide a framework that can be extended to practical applications in smallholder farming.

Acknowledgements

These data have been made available by the Economics Department, Addis Ababa University, the Centre for the Study of African Economies, University of Oxford and the International Food Policy Research Institute. Funding for data collection was provided by the Economic and Social Research Council (ESRC), the Swedish International Development Agency (SIDA) and the United States Agency for International Development (USAID); the preparation of the public release version of these data was supported, in part, by the World Bank. AAU, CSAE, IFPRI, ESRC, SIDA, USAID and the World Bank are not responsible for any errors in these data or for their use or interpretation.

Notes

1 In Ethiopia, most seeds are not enhanced varieties that must be purchased each season, so for this example we use prices for indigenous seed.

2 The Ethiopian currency; about 19.20 = 1 USD in January, 2014

3 Alternative measures include Guido Gryseels and Michael R. Goe, who calculate an average of 440 hr/ha/year, bounded at 600 hr/ha/year for the laborintensive Teff, and 300 hr/ha/year for pulses. These latter measures cover between two and four growing seasons in Ethiopia.

4 For the returns estimated in this analysis, labour and fertilizer costs are assumed to be stable, and to vary only with inflation.

5 The methodology employed here was formalized in Ballestero & Garcia-Bernabeu (2012)