Investigation of the economics of water as used by smallholder irrigation farmers in South Africa

Abstract

This study investigates the economics of water as used by smallholder irrigation farmers in South Africa. The productivity and value of water were estimated with data from two smallholder irrigation schemes: Zanyokwe and Thabina. Production parameters such as fertilisers, seeds, pesticides, equipment, transport, labour, and water were treated as inputs. Various methodologies used to estimate water value, including cost-based approaches, were thoroughly reviewed. The aim was to select the ones with more justification for use in smallholder irrigation sector and also to compare a number of approaches. On a case study basis, three methods were applied: residual valuation method, willingness to pay and cost-based approaches (i.e. accounting costs of O&M). Water productivity and values were then evaluated as per crop, farmer, and scheme. Also, cross-section regression analysis was used to investigate the effect of some key socio-economic factors of production on gross margin and willingness to pay. The results indicated that on average, the value of water varies according to methods, crops, farmers and schemes. In the Zanyokwe scheme, water value estimated by the residual method in cabbage is higher (R1.64 per m3) than the one in dry maize (R0.35 per m3), if intensive and high yield irrigated crops are grown per year. Also, in Thabina, water value for cabbage (R1.14 per m3) outperforms water value for dry maize (R0.02 per m3. This means that there is greater potential in vegetable crops than food grain crops, although the two schemes have different irrigation systems, and such analysis is based on one-year data, which may lack accuracy. Low water valuation is ascribed to low yield and extensive cropping systems, because gross margin per hectare is very low. This signifies the need for expansion in high value crops rather than low value crops. At farm and scheme level, the results were derived by using the Smile database and simulated platform. The Smile platform is a data capturing and a calculation tool. It calculates a number of indicators, economic figures, at scheme and individual farm level, allowing for evaluation of the current situation. The results suggested that at present, the Zanyokwe scheme requires about 1 739 255 m3 of irrigation water per year. The total operational costs (accounting costs of O&M) are about R146 097.42. In other words, supplying 1 m3at farm level will cost R0.084. This implies that if irrigation charges are levied so as to cover O&M costs of the Zanyokwe scheme, the current costs (R0.084) will form only 23% of the average gross margin of R0.37 per m3 used at scheme level. Furthermore, in the Zanyokwe irrigation scheme, the results revealed that the most active and efficient farmers (specialized farmers) can make an average gross margin of R4 105 per ha per year, also achieving the highest water productivity R0.69 in gross margin per m3 consumed. However, in the Thabina scheme, the results indicated that, to supply 1 m3will cost R0.062. Thus, the current water supply costs cover about 56 % of what is earned (i.e. R0.11 per m3 used) at scheme level. Again, the most active farmers (commercially oriented pensioners) are more efficient, with average gross margin of R3 092 per ha per year, also achieving the highest water productivity (R0.53 in gross margin per m3 used). These results suggested that certain smallholder farming systems seem capable of paying for irrigation charges of their respective schemes if they are obliged to do so. As far as willingness to pay (WTP) and cost-based approaches (CBA) are concerned, the results clearly show that the active farmers in the Zanyokwe scheme have lower WTP per m3 (R0.03) than the GM of output (R0.69) per m3 of water used. Also, the accounting cost (R0.084 per m3 of water used) is lower than the GM gained. However, in the Thabina scheme, the situation is quite different. The active farmers are willing to pay R0.19 per m3 of water used. This implies that, if farmers are to pay for the charges in order to cover O&M costs, the farmers in Thabina are ready to pay as much as three times the proposed costs of O&M (R0.062 per m3 of water used), although both the WTP and the accounting cost are lower than GM gained. In these results, it is significant that both the accounting cost and the willingness to pay are lower than the GM per m3 of water used at least in the Zanyokwe scheme. Even though the data were drawn from a sample for one year only, this finding on the perception of farmers has implications for extension and training to improve future productive use. Regarding the findings from cross-section regression analysis, the results for GM (as dependent variable) in general indicated that in the Zanyokwe scheme, only credit affects output positively and significantly. Production costs have significant effect on output, but with a negative correlation. It is striking that all the other factors of production including hired labour show negative and insignificant effects on output. In the Thabina scheme, the most important factors of production in the model are land size (hectare) and production costs (Rand per ha). These two variables influence output positively and significantly. As far as WTP (as dependent variable) is concerned, the results indicated that in the Zanyokwe scheme, it is striking that in all the investigated factors, only credit affects WTP positively and significantly. All other factors are insignificant. Also, gross margin of output per m3, unexpectedly, has displayed a negative and insignificant effect on output. In the Thabina scheme, the results show that the land size per hectare, and gross margin of output per m3 affect WTP positively and significantly. Such results are consistent with the assumptions made in the conceptual framework that a farmer with high gross margin gained at farm level is more likely to pay for water costs than those with poor gross margin. On the basis of these findings, the following recommendations were formulated. Specific policies should include promotion of high value crops and improved varieties of seed for food grain crops (e.g. maize) and vegetables (e.g. cabbage). While improved agronomic practices remain important, there is also potential to increase productivity and profitability of the crops by improving water management practices at the canal-system level, such as better timing of water delivery and increased overall canal-water supplies at the farm level. Finally, from a cost recovery point of view, government should develop a program of cost sharing for capital costs of irrigation development. With regard to inter-sectoral competition, these results highlight that, if inter-sectoral competition is left to uncontrolled market forces may result in smallholder farmers' selling their water rights to sectors which value water at higher levels. To avoid a “liberal trap” such as in the example of Chile (where smallholder farmers "en masse" sold their water rights, resulting ultimately in deeper rural poverty), some form of control/management of water rights transaction involving smallholder farmers is necessary. Finally, the findings of this study can be used in various ways. Since, these values determine the farmer’s ability to pay for water now or in the future, the incentive to use water judiciously will be governed by these values. Secondly, the results can be used to evaluate whether the costs estimated and gross margin per m3 gained at farm level, are in line with the farmers’ willingness to pay. Further work is recommended to clarify these conclusions and provide more policy clarification on the better use of water by smallholder irrigation farmers in South Africa.