Decision Taking Under Uncertainty: A Linear Programming Model of Peasant Farmer Behaviour*

Abstract

This article presents a method for taking uncertainty into account when representing the production decisions of near‐subsistence farmers. The main assumption is that farmers maximise expected income subject to meeting yearly subsistence requirements in adverse conditions. A game theoretic criterion is employed to minimise the cost of securing subsistence requirements under uncertainty. The results of incorporating the formulation into a linear programming model of peasant farming in South East Ghana indicate that the assumption made as to farmers' objectives are valid for the situation studied and it is suggested that farmer behaviour observed elsewhere in Africa may be explained in similar terms. Résumé PRISE EN COMPTE DE L'INCERTITUDE DANS LES PRISES DE DÉCISION: UN MODÈLE DE PROGRAMMATION LINEAIRE DU COMPORTEMENT DU PAYSAN AGRICULTEUR Cet article présente une méthode pour prendre en compte les risques d'incertitude dans la représentation des décisions de production des agriculteurs de quasi subsistance; le postulat principal est que les agriculteurs maximisent les recettes attendues pourvu que soient satisfaits les besoins annuels de subsistance dans des conditions adverses. Un critère de théorie des jeux est employé pour minimiser le coût de la satisfaction assurée des besoins de subsistance dans des conditions d'incertitude. Les résultats de l'insertion de la formulation dans un modéle de programmation linéaire de l'exploitation agricole paysanne dans le sud‐est du Ghana indiquent que les hypothèses faites quant aux objectifs des agriculteurs sont valables pour la situation étudiée, et les conclusions de l'étude laissent penser que le comportement des agriculteurs observé ailleurs en Afrique peut s'expliquer dans des termes analogues. Zusammenfassung DAS ENTSCHEIDUNGSVERFAHREN BEI UNGEWISSHEIT: EIN MODELL MIT LINEARER PROGRAMMIERUNG ÜBER DAS VERHALTEN DER LANDBEVÖLKERUNG Dieser Artikel stellt eine Methode dar, Ungewissheiten zu berücksichtigen, wenn Produktionsentscheidungen durch Bauern getroffen werden, die an der Existenzgrenze leben. Die Hauptannahme ist, dass Bauern ein grösstmögliches Einkommen erwarten, vorausgesetzt, dass bei ungünstigen Verhältnissen die jährlichen Bedürfnisse erfüllt werden. Ein Spieltheoriekriterium wird eingesetzt, um die Anschaffungskosten von Lebensnotwendigkeiten bei Ungewissheit auf ein Mindestmass zu verringern. Die Ergebnisse, die bei der Aufnahme der Formulierung in ein Modell mit linearer Programmierung der durch Bauern in Südostghana betriebenen Landwirtschaft entstehen, weisen darauf hin, dass die Annahme bezüglich Ziele der Bauern in der beobachteten Situation wohlbegründet ist. Es wird vorgeschlagen, dass das Verhalten der Landbevölkerung, das anderswo in Afrika beobachtet wurde, in ähnlicher Weise beschrieben werden könnte.

Direct payments for conservation — the importance of environmental measures in farming systems for bird populations in a fragmented landscape. A case study in Guatemala

“…Furthermore, forest and vegetation conservation also makes »sense from the viewpoint of minimising high environmental risk under considerable uncertainty« (Nasi et al 2002). This coincides with the safety-first criterion of Low (1974) ensuring minimum conditions to meet farmers' necessities by conserving environmental measures within farming systems. In the opposite situation, when trees are harvested and vegetation eradicated, farmers are in a way ›destroying‹ their own minimum needs and staying empty handed.…”

Uncertainty, Diversification, and Mental Maps In Spatial Choice Problems

“…There is now considerable empirical evidence (see, for example, [16, 18,21,351) that a portfolio approach provides a better description of such activity than the simple hypothesis that a farmer maximizes the expected returns to a given investment. Using the above ideas, it is possible to give a radical reinterpretation of Wolpert's study [36] of Swedish farmers, in which Wolpert attempted to prove that farmers were satisficers and/or of bounded rationality by showing that they did not maximize expected returns.…”

“…If one assumes that a positive amount of information is collected from each location, it is possible to obtain the optimum amounts collected from each location (since the dual variables of the nonnegativity constraints are zero), by setting the derivatives of (4) to be equal to zero, Using the methods of comparative statics (see [21]), it is possible to obtain the dependencies of the optimal quantities (nl, n,) on the various parameters. First, if the marginal cost of information pi increases, so the optimal quantity n, decreases, while the optimal quantity collected from the other location actually increases [( ani/ api) > 0, i # j ] .…”

Incorporating economic risk aversion in agroforestry planning

“…Portfolio theory, as operationalized through quadratic programming, offers an alternative to deterministic modeling that allows for the explicit recognition of risk and may lead to modeling decisions that more closely resemble those made by rural agriculturalists [13]. Blandon [5] discussed the role of portfolio theory in reducing the financial risk of agroforestry systems with a cash crop component.…”