Abstract:
The main objective of the study was to estimate non-market value of irrigation water as an
input in sugar cane production in the Lowveld of Swaziland. This study used two
independent approaches to non-market valuation, the residual value and production function
approaches, to calculate the value that sugar cane farmers in the Lowveld region of
Swaziland attach to irrigation water. The former estimated the average value of water, while
the latter estimated the marginal value. The study also estimated the output elasticity of
irrigation water, identified factors determining irrigation water values, and used stochastic
frontier analysis to estimate farmers’ technical efficiency (TE) scores, and investigated the
relationship hypothesised between irrigation water values and TE. Irrigated sugar cane
production was specifically selected for this study on account of its socio-economic
importance in Swaziland.
Using data obtained from 78 sugar cane farms, the mean estimated value of irrigation water,
measured in Emalangeni per metre cubed, was E1.60/m3 using the residual value approach,
and E1.51/m3 using the production function approach. A t-test showed that the observed
differences between the values estimated from the two independent approaches were not
statistically significant, suggesting that either method can be used to value irrigation water
employed in sugar cane production in Swaziland. The results from the t-test, in conjunction with the economic theory of duality, also allow us to conclude that the production technology
employed by irrigation sugar cane farmers exhibits constant returns to scale. Irrigation water
was output inelastic (0.711), lending additional credence to the constant returns technology
conclusion. The value calculated for irrigation water was negatively related to irrigation
water quantity, suggesting that price can be used as an instrument to directly regulate the
quantity of irrigation water the farmer employs. The value calculated for irrigation water was
negatively related to quantities of labour, quantity of irrigation water used, fertilizer and
chemicals employed, suggesting that price can be used as an instrument to indirectly regulate
the quantity of irrigation water the farmer employs. The value imputed for irrigation water
was positively related to farm size and total revenue, suggesting that the more resourceendowed
farmers can potentially pay higher for irrigation water. It is thus conceivable to
design irrigation water pricing policies with equity considerations. TE scores ranged from
0.397 to 0.955, with a mean of 0.840. Farmers with higher TE scores also had higher implicit
values for irrigation water, suggesting that irrigation water pricing can be used as a tool for
motivating resource use efficiency.
The key policy implication derived from this study is that price-based instruments have a
potential in the management of scarce irrigation water resources in Swaziland.
