Risk preferences and soil conservation decisions of South African commercial sugarcane farmers.

Abstract

In this study commercial sugarcane farmers' risk preferences are measured and tested, amongst other factors, as determinants of their soil conservation decisions. Motivation for this research stems from ongoing erosion on South African commercial farms, including sugarcane farms; recognition that agricultural economists have a poor understanding of factors affecting soil conservation; and because little positive economic research has been conducted linking risk preferences to environmental choice. This research also affords the opportunity to investigate both theoretically and empirically the effect of the range of the data on the Arrow-Pratt absolute risk aversion coefficient (AP); and to investigate factors affecting farmers' risk preference. Both of these are important topics for research. The proposed conceptual model views farmers' soil conservation decisions in a system analogous to a demand system from which a probabilistic choice model of the adoption of soil conservation practices is developed. This model relates adoption of soil conservation technologies to the demand for soil conservation using derived demand theory. Consequently, soil conservation decisions are analysed in terms of both achieved soil conservation efficiency and farmers' choices of soil conservation measures. This conceptual model further advances previous research in that it provides a sound economic base for the analysis and it accounts for intra-farm variations in soil conservation decisions. It is demonstrated that AP's must be adjusted for the range as well as the scale of the data for comparison of risk preferences across different risk situations. Although Raskin and Cochran's (1986) theorems adjust AP's for the scale of the data and are widely used in agricultural economic research, they only adjust correctly for the range of the data under special circumstances. It is demonstrated that in previous agricultural economic research AP's have been incorrectly adjusted for the range of the data and that the effect of the range on AP's has been often incorrectly interpreted as a wealth affect. It appears that informing agricultural economists of the sensitivity of AP's to the range of the data is important. An approach is proposed for adjusting AP's for both the scale and the range of the data. Fifty three commercial sugarcane farmers from KwaZulu-Natal were surveyed by personal interview during May and June 1996 to elicit farmers' risk preferences and information on their soil conservation decisions. Arrow-Pratt absolute risk aversion coefficients are elicited using a direct elicitation of utility approach. Two farmers refused to participate in lottery games for religious or moral reasons. Of the remainder 57.2 percent were risk averse, 29.6 percent risk neutral and 13.2 percent risk preferring. On average they were risk averse although risk preferences vary significantly amongst individuals. Regression analysis indicates that on average sugarcane farmers are averse to a possible loss in wealth relative to initial wealth and they exhibit increasing absolute risk aversion although at a decreasing rate with increasing gamble range. Technological information is used to assess KwaZulu-Natal commercial sugarcane farmers' intra-plot (panel) soil conservation relative to requirements of the Conservation of Natural Resources Act of 1983. Findings indicate that farmers consider enforcement of the Act to be unlikely and that 28 percent of farmers surveyed do not meet adequate intra-panel soil conservation standards. This partially reflects that a large proportion of farmers have not yet completed implementing their soil conservation plans. Nonetheless, policy makers should examine factors impeding adequate enforcement of the 1983 Act. Multiple regression and principal component analysis techniques are used to analyse farmers' soil conservation decisions. Decisions are analysed in terms of achieved intra-field soil conservation adoption and soil conservation effort, and farmers' choices of soil conservation measures. Measurements of soil conservation efficiency are defined to be objective, be suitable for relative analysis, and reflect intensiveness of adoption. Extensiveness of adoption is captured by eliciting information for more than one field per farm. Achieved goodness of fit statistics are good compared to previous similar studies and indicate that intra-farm variations in soil conservation contain important information explaining soil conservation decisions. Findings generally support the conceptual exposition that the demand for soil conservation technologies is derived primarily from the demand for soil conservation. However, the demand for fire insurance does affect adoption of strip cropping, and hence soil conservation decisions. Results indicate that although risk and risk preferences are not significant determinants of the quantity of soil conservation demanded by sugarcane farmers, they are significant determinants of their soil conservation decisions. Firstly, the rate at which more risk averse farmers tend to undertake extensive restructuring of sugarcane field layouts tends to be relatively slow. This verifies the hypothesis that risk averse farmers value an option to postpone capital intensive, irreversible investments. Secondly, relatively risk averse farmers are more likely to adopt strip cropping programmes. This finding is attributed to their aversion to income risk and not risk of excessive soil loss. The hypothesis that risk averse farmers adopt soil conservation practices to reduce risk of serious erosion is not supported. It is concluded that farmers' abilities to assess soil conservation efficiency is poor, especially with respect to the partial contribution of conservation structures to achieved soil conservation efficiency; while farmers tend to implement soil conservation on steeper slopes first; and perceive minimum tillage and strip cropping programmes to be imperfectly compatible. Adoption of minimum tillage, trash mulching and strip cropping are primarily constrained by physical climatic, field and soil characteristics and to a lesser extent by farmers' management time and technical abilities. Implementation of improved soil conservation structures is constrained by financial constraints. Amongst other factors, education and use of extension information sources and adoption of land use plans are positively related to both conservation adoption and effort. Soil conservation policies must make farmers more aware of soil erosion. This is best achieved through provision of specific technical information, especially land use plans.