Congress Resources: Papers, posters and presentations

< Return to abstract list

Residential Water Demand Estimation In A Developing Country Under Contrasting Scenarios Of Climate Change

Congress: 2015
Author(s): Sergio Orrego, Dari­o Jiménez, Felipe Vásquez
Universidad Nacional de Colombia1, Universidad del Desarrollo, Chile2

Keyword(s): Sub-theme 14: Valuing water: monetary and non-monetary dimensions,


Expected future population growth in developing countries and emerging economies will increase the water demand for residential use. Moreover, projected changes in temperature and precipitacion as a result of climate change will affect the water available to alternative uses. An increase in the demand along with a potential reduction in the supply of water constitute an important and critical issue. Issue that will have a substantial influence in the efficient management of water resources and the public policies adopted to deal with water scarcity and adaptation to climate change in developing countries. However, efficient management and water-related public policies require rigorous economic analysis of the potential effects of the climate change on water demand. This research explored two main aspects: i) how sensitive households are to changes in water prices, measured by the price elasticity, ii) how sensitive expected mean water consumption is to climate change.


A structural discrete/continuous choice (DCC) model consistent with utility theory (Olmstead et al., 2007) was used to estimate residential water demand for households in Manizales, Colombia. Households faced nonlinear prices as a result of a increasing block price (IBP) structure set administratively by the water supplier and based on regulations of a public utility comission. Although the DCC model seems to be more complicated to estimate compared to alternative econometric approaches (e.g., ordinary least squares or instrumental variables), it allowed to understand the consumer behavior under nonlinear prices and to estimate consistent price and income coefficients in a model of water demand.. We estimated a demand function using monthly water consumption (m3), period 1997-2013; weather variables corresponding to monthly average temperature and precipitation, period 1981-2013; household characteristics such as household size, number of bathrooms, type of residence, and possession of washing machine; marginal prices; and virtual income, or income adjusted by the subsidies derived of purchasing not all water units at the same marginal price. While climate data for the 1981-2013 reference scenario were obtained from instrumental records, climate variables for the period 2014-2030 and four contrasting greenhouse-gas concentration scenarios were projected using appropriate downscaling methods. The climate change scenarios were related to Representative Concentration Pathways (RCP) and denoted different radiative forcing expressed in watts per square meter: RCP2.6, RCP4.5, RCP6.0 y RCP8.5. A survey and a two stage sampling with unequal probabilities were used to collect socio-economic and demografic information from 486 households. The estimated demand function included two error terms to account for heterogeneous household preferences for water consumption and both optimization and measurement error. A log-log demand function for residential water was estimated which facilitates comparison of results with previous empirical studies. The estimated model was used to calculate price elasticities and predict expected mean water consumption under contrasting climate change scenarios.

Results and discussion

Results indicate that household characteristics affect positively the residential water demand. Findings that are clearly consistent with anticipated expectations. The coefficient estimates for the temperature and precipitation variables were both negative and statistically significant. The price coefficient of -0.32 was also statistically significant and reflected the price elasticity conditional on the observed block of water consumption. This result indicates that a 10% increase in the water price will lead to a 3.2% decrease in the consumprion of water, all other determinants of water demand being equal. The calculated unconditional price elasticity was -0.308 and reflected the probability that a household may change their consumption block in response to a price change. The unconditional price elasticity was smaller in absolute value compared to the corresponding conditional elasticity. This suggests a potential attenuating effect derived from the fact that many sample households decided to use a quantity of residential water equal or very close to the boundary quantity between blocks or kink point, where price elasticity is zero. The income coefficient was statistically significant but of low magnitude, which seems to be consistent with the low proportion of household expenditure on water. Estimates for the two errors were statistically significant with heterogeneous preferences being more responsible for the unexplained variation in water demand. Using the parameter estimates of the DCC model including as explanatory variables household characteristics, price, virtual income, temperature and precipitation for the period 1981-2013, it allowed to calculate an expected mean water consumption of 16.77 m3 per month. This was considered as our reference mean value. The expected mean water consumption for the four climate change scenarios (RCP2.6, RCP4.5, RCP6.0 y RCP8.5), using the projected series of temperature and precipitation for the period 2014-2030, were 16.72, 16.72, 16.72 y 16.68 m3 per month, respectively. A 0,5 % decrease in the expected mean water consumption was obtained for the climate change scenario RCP8.5 compared to the reference mean value. Although it is a slight decrease, the results show the direction of the potential effect of climate change on residential water demand in a developing country.


The estimated DCC model adequately explained the discrete and continuous choices when a household is faced to a decision related to water consumption. The calculated unconditional price elasticity was -0.308, indicating that the demand for residential water is inelastic, as found in previous empirical studies (Espey et al., 1997; Dalhuisen et al., 2003; Olmstead et al., 2007). Economic information that is of interest to public institutions responsible for the water pricing. Moreover, the expected mean water consumption decreased slightly under the climate change scenario RCP8.5 compared to the reference mean value. This research provides information that may be valuable to identify feasible and cost-effective strategies and policies for the adaptation of the water sector to climate change, mainly in developing countries where water scarcity will likely be critical over the next decad

1. Dalhuisen, J.M., Florax, R.J.G.M., de Groot, H.L.F. and Nijkamp, P. (2003) Price and income elasticities of residential water demand: a meta-analysis. Land Economics 79,292-308.

2. Espey, M., Espey, J. and Shaw, W.D. (1997) Price elasticity of residential demand for water: a meta-analysys. Water Resources Research 33,1369-1374.

3. Olmstead, S., Hanemann, W.M. and Stavins, R.N. (2007). Water demand under alternative price structures. Journal of Environmental Economics and Management 54,181-198.

© 2011 IWRA - International Water Resources Association - - Admin