Abstract
Issues of water scarcity can be ameliorated through household adoption of water-saving technologies and by adaptation of consumption behaviour. In this paper the determinants of the adoption of water-efficient devices and of water-saving habits in Spain are analyzed using data from 27,000 households. This includes information on choices about self-reported conservation habits and decisions about the adoption of water-saving equipment. The findings show that educational campaigns have a strong positive effect on both decisions to undertake investments and decisions to adapt habits. These results also allow campaigns to be aimed at certain socio-economic groups identified in the econometric analysis.
Acknowledgements
We would like to thank the financial support of the Ministry of Science and Innovation (through the project with reference MICINN-09-ECO2009-08824) and the comments and suggestions we received after presenting a previous draft at the May 2011 Workshop on Water Pricing and Roles of Public and Private Sectors in Efficient Urban Water Management, organized by the Third World Centre for Water Management, the International Water Resources Association, Global Water Intelligence and the University of Granada.
Notes
1. We built a score of pro-environmental opinion based on several issues asked on the survey. Individuals were asked if they agreed with the following proposals: (1) to oblige people to separate waste (being fined if they do not do it); (2) to restrict water overuse; (3) to set a green tax on the most polluting combustibles; (4) to constrain private transportation use; (5) to establish a green tax on tourist activities; (6) to install renewable energy plants in their city/village; (7) to pay more for alternative energies; and (8) to reduce noise on the main roads. All of these are dummy variables which take the value 1 if the answer is yes, and 0 if the answer is no. To build the pro-environmental opinion score, we sum up all these variables. The higher the score, the more ‘pro-environmental’ the individual is, according to this criterion.
2. A likelihood-ratio test of independence of equations yields a statistic distributed χ2(1) equal to 422.55, which clearly leads us to reject the null hypothesis of no correlation.