1. U.S. Environmental Protection Agency (EPA). Office of Water. 1997. “Drinking Water Infrastructure Needs Survey.” Report to Congress. EPA-812-R-97-001. Washington, D.C. January.
2. F-tests were completed by comparing the unrestricted model (Model I) with the three alternative specifications. In all cases the null hypothesis was rejected at a significance level of α = 0.01. For simplicity, only the restricted models with unitary elasticities of substitution with respect to the inputs for the hedonic and nonhedonic specifications are reported.
3. Spady and Friedlaender's (1978) favor hedonic representations of outputs and qualities as arguments, compared with conventional cost functions with exogenously specified quality adjusted outputs as arguments. The hedonic specification permits various quantity-technology combinations to reflect the same level of quality.
4. Depending on raw water quality, other filtration technologies may also pretreat water with CFS prior to filtration. Coagulants are added to the raw water and mixed to form larger particles. The larger particles (floc) separate out by gravity in a sedimentation tank, and the resulting water is filtered.
5. Boisvert R.N. , Tsao L. , and Schmit T.M. 1996. “The Implications of Economies of Scale and Size in Providing Additional Treatment for Small Community Water Systems.” Unpublished report to U.S. Environmental Protection Agency. Department of Agricultural, Resource, and Managerial Economics, Cornell University. February.