Date of Award

1982

Degree Type

Thesis

Degree Name

Doctor of Philosophy (PhD)

Department

Economics / Economic Policy

Supervisor

Professor D.W. Butterfield

Abstract

Knowledge of farmers' decision making behaviour constitutes an essential element in the formulation of appropriate policies for agricultural development in low-income countries. Since agricultural production is inherently a risky process in which farmers are exposed to numerous sources of uncertainty, any study of farmers' decision making behaviour should explicitly consider the role of risk. This study analyses farmer behaviour in Bangladesh, focussing on the effects that risk has on resource allocation decisions. The key issues explored in this context are: (a) analysis of peasant behaviour towards risk in terms of subsistence requirements and income earning potential of the farm family; and (b) investigation of allocative efficiency in the presence of risk for the small-holding farmers in Bangladesh.

In peasant agriculture where the farms are extremely small and cultivation is dependent upon highly variable rainfall, the farm family may be preoccupied not with maximising income but with maximising their chances of survival. Safety-first models of decision making are based upon this notion of disaster-avoidance and this study specifically employs Roy's Safety Principle under which the decision maker attempts to minimise the probability that income falls below a certain level. Based on the allocative efficiency conditions derived from our safety-first model, alternative testable hypotheses are developed to ascertain whether the small-holding farmers in Bangladesh efficiently allocate their resources to various crops in the presence of risk. A comparison of resource allocation behaviour under two rival criterion for analysing decision making under uncertainty shows that it is possible to analyse the risk preferences of the farm household in a safety first model much in the same way as in an expected utility model. The interpretation of the risk coefficient, however, will be somewhat different for the safety-first model. Emphasizing the farmer's behaviour towards risk, the safety-first approach attempts to explain, in terms of subsistence needs and income generating capacity of the farm households, whether the household is 'forced to gamble' or 'allowed to accept less risk' in their choice of crop portfolio.

The model was estimated using field survey data collected in four different regions in Bangladesh. In particular, data on minimum consumption needs were directly elicited from each farm household in the sample. Our estimates of the risk coefficients in each region show that while most of the farm households in Sylhet should display 'gambling type' behaviour, which occurs when the subsistence needs of the farm family exceeds their income earning potential from farming activities, the situation for the farm households in the other three regions is not so desperate. For most of these households, the minimum consumption needs were less than their income earning capacity, thereby allowing them to undertake less risky activities. Attempts were then made to establish a systematic relationship between our coefficients capturing peasant behaviour towards risk and a number of socio-economics variables characterising the peasant households and their access to income-earning opportunities. Our regression results show that although the explanatory power of the risk-predictive equations are not very high, major explanatory variables such as family size, farm size and off-farm income have the expected signs in most case and are statistically significant. We are thus able in our study to identify some specific determinants of peasant behaviour towards risk and also to quantify their impact on decision making.

Our empirical results indicate that in general the safety-first model outperforms the expected profit maximisation model in our sample. This tends to validate the hypothesis that in subsistence agriculture like that in Bangladesh, peasant households are preoccupied with their security and survival, and therefore, are very likely to follow some sort of safety-first behaviour in their resource allocation decisions. There were, however, considerable regional variations in resource allocation behaviour. Such observed differences in the behaviour pattern across the four regions examined in Bangladesh are explained in terms of a disaster-avoidance motive which is ultimately traced to the capacity of the farm households to generate sufficient income to meet their subsistence needs.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Included in

Economics Commons

Share

COinS