Publication

Abstract : This paper presents empirical results on progressive incidence of water charges that contradict conventional wisdom in developing countries and many developed countries (Whittington, 1992, 2003; Boland and Whittington, 2000). Water pricing policy is shown to possess the potential of redistribution. This finding is supported by traditional incidence exercise incorporating family size and metering, linear expenditure systems approach and error correction methodology. Family size among high income categories relative to low-income groups is larger in Mauritius. Because of this unusual demographic dimension, increasing block tariffs coupled with lifeline rates that are below long run marginal cost applicable to low-volume users produce this different result. It is important to document that extensive metering both in urban and rural areas allows low income consumers to benefit from increasing block tariffs. Regional disparities among residential water users further contribute to redistributive strategy. More precisely rural low-income households are subsidised by higher income consumers who reside predominantly in urban regions. A comparative analysis of price and income elasticities for different income categories reveals that low income households face more than average price elasticity of demand for water and less than average income elasticity. In the light of this result, it is established that increase in demand for water among richer households has the potential of improving equity in water charges. The computed marginal budget shares derived from linear expenditure system shows that the amount of supernumerary income would be higher for richer people so that a percentage increase in income may result in a higher level of spending on water as compared to lower-income households. In this case, water charges are found to be re-distributive.