Lesson Learned

Learning for Future MCC Economic Models: Primary Substitution Effect Takes Longer, and Secondary Substitution Effect Did Not Occur: The primary substitution effect (substitution of treated wastewater instead of freshwater for irrigation) did occur.

Learning for Future MCC Economic Models: Primary Substitution Effect Takes Longer, and Secondary Substitution Effect Did Not Occur: The primary substitution effect (substitution of treated wastewater instead of freshwater for irrigation) did occur. However, it occurred about two years later than what was expected in the MCC economic model. The secondary substitution effect (substitution of utility water for bottled water) did not occur. In the case of primary substitution, the outcome was primarily contingent upon a transition to new agricultural products, which involved not only changes in cultivation—and the private investments to do so—but the development of new marketing relationships. At the time that the compact was being designed, there was some evidence of changing agricultural production and business models in the Jordan Valley, but the factors driving the speed of these changes were not explored in any particular depth. Future similar interventions might benefit from more systematic treatment of linkages to strategic business decision making—at least to ground expectations regarding behavioral change involving enterprises and to identify possibly salient exogenous risks. Regarding the secondary substitution, as noted earlier, perceptions are not always an explicit, highlighted feature in project logics but can a be latent, dominant determinant of investment performance that is contingent on the level of use of a public service—particularly when that use is sensitive to users’ trust in its quality. In some settings, perceptions may change very slowly: only after a sufficiently long series of observations, and the information from those observations might also only slowly diffuse among potential users. In the future, the use of social networking might be a necessary element to weave into design of similar projects. Where the success of interventions hinges on desired shifts in consumer preferences and the adoption of particular technologies by private enterprises, future MCC economic assessments may profit from probing the diffusion processes that are hypothesized for each. Incorporating representations of the incentive and agent-based decision environments explaining behavioral change into economic modeling may allow some testing of the sensitivity of results to particular features governing behavior and identify topics at an early stage of project design or due diligence that might be investigated through suitable survey approaches.