Grantee Research Project Results
2019 Progress Report: Estimation of Spatially Explicit Water Quality Benefits throughout River Systems: Development of Next Generation Stated Preference Methods Using National Probability Samples and Online Labor Pools
EPA Grant Number: R836167Title: Estimation of Spatially Explicit Water Quality Benefits throughout River Systems: Development of Next Generation Stated Preference Methods Using National Probability Samples and Online Labor Pools
Investigators: Johnston, Robert J , Wollheim, Wil , Moeltner, Klaus
Institution: Clark University , University of New Hampshire , Virginia Tech
EPA Project Officer: Packard, Benjamin H
Project Period: April 1, 2016 through March 31, 2019 (Extended to March 31, 2022)
Project Period Covered by this Report: April 1, 2019 through March 31,2020
Project Amount: $799,919
RFA: Water Quality Benefits (2015) RFA Text | Recipients Lists
Research Category: Water
Objective:
This project will develop novel modular approaches to stated preference valuation designed to address the challenges of use/nonuse value estimation for complex temporally/spatially explicit aquatic ecosystem change with heterogeneous benefits for different user and nonuser groups. The approach, denoted Free-form Choice Experiments (FCEs), will restructure the way that willingness to pay (WTP) is elicited and estimated in survey-based valuation by hybridizing methods from contingent valuation choice experiments, revealed preference modeling, online labor pool sampling, and Bayesian econometrics. The approach will be developed to estimate WTP for water quality and ecosystem service improvements throughout river networks. FCEs are designed to be transformative across multiple dimensions, including: (a) novel ways to elicit stated preferences and design surveys, with scenarios coupled directly to ecological models, (b) flexible value elicitation and modeling that generates benefit functions linked to temporally/spatially explicit effects, (c) an ability to evaluate the relevance of many ecological indicators to different user/nonuser groups, (d) novel modeling of data using Bayesian econometrics, and (e) new approaches to stated preference sampling using online labor pools. Biogeochemical forecasts for valuation scenarios will be projected using FrAMES (Framework for Aquatic Modeling of the Earth System), a process-based spatially explicit water quality model.
Progress Summary:
We have completed Tasks I - IV in the project proposal. Task V is nearly complete (only pilot testing remains). Task VII is in progress. Other tasks remain to be completed. Survey development is complete, including development of water quality scenarios and indicators, choice questions, and other questionnaire materials. Questionnaires have been pretested and revised with input from focus groups and pretest interviews, with feedback from external experts in stated preference methodology. We have jointly developed this scenario design with Bayesian econometric models suitable for model estimation and have evaluated model performance using simulated response data. We have completed GIS map zooming and tracking functionality that will enable respondents to view different water quality changes at any desired scale and resolution within the study area with the resulting "zooming and panning" data available to populate the econometric model. We have also completed the surrounding code and shell that will enable the questionnaire to be pilot tested and implemented on the AWS Mechanical Turk service (abbreviated as "MTurk") platform.
Development of Water Quality Indicators and Scenarios
The Framework for Aquatic Modeling in the Earth System (FrAMES) was applied to the Northeast US region to simulate specific conductivity, fecal coliform, and dissolved inorganic nitrogen, which were used to characterize three water quality indicators of direct relevance to the public. An aquatic life indicator was developed based on a reported range of chloride concentrations that affect aquatic organism survival. A water safety indicator for drinking and recreation was developed based on fecal coliform count guidelines. Finally, an indicator for overall water quality was developed as a combined metric using all three modeled solute concentrations, compared to reference conditions for the case study watersheds. Grounded in these indicators and focus group input we developed policy scenarios that included combinations of region-wide management actions to generate future scenarios of water quality change for the valuation survey. Forty-one scenarios were defined using a full factorial (36) of region-wide management actions, including riparian buffer expansion (0%, 90% of agricultural and developed land), waste water treatment plant upgrades (50%, 90% from secondary to tertiary treatment); storm water retention (0%, 65%, 90% urban rain infiltration), and road salt application (700, 4000, 7600 g/mm/m2 snowfall). Five additional scenarios are added representing minimum and maximum levels of storm water retention and road salt application applied only in northern (ME, NH, VT) and southern (CT, MA, RI) New England states, with riparian buffers and WWTP upgrades at maximum levels in the target region only.
Final Development and Testing of Questionnaires: Based on these scenarios complete versions of questionnaires have been developed, pretested and finalized, and are hosted on Qualtrics. Questionnaires included the value elicitation binary choice question along with survey instructions, supporting informational materials, ancillary questions (including warm-up, debriefing and validation questions), and elements to enhance incentive compatibility. Valuation scenarios were constructed around indicators described above. Spatially explicit levels for these indicators were mapped across the study area. Valuation scenarios compare outcomes for these indicators with and without proposed management alternatives. These outcomes, paired with the hypothetical household cost of each management alternative, comprise each valuation question. Each respondent considers a single, hypothetically binding, binary referendum question designed to meet conditions for incentive compatibility. All scenarios are illustrated for a case study area of approximately 95,800 miles of rivers and streams in Connecticut, Massachusetts, Rhode Island, Vermont, New Hampshire and Maine, over 71,992 square miles of land (Fig. 1). An illustrative map of the indicator showing the safety of waters for human uses is shown by Figure 2. All maps have GIS zoom capability, enabling respondents to view water quality baselines and changes at any desired magnification and local area. When clicking on the map link in the Qualtrics survey, the respondent is taken to the corresponding water quality scenario map hosted on hosted in ArcGIS Online, allowing them to pan and zoom as desired.
Development of Bayesian Econometric Models
We have completed (i) a simulation exercise to inform the optimal number and placement of bid values to be offered to respondents, (ii) a simulation exercise to determine the diagnostic and predictive abilities of the Bayesian model search algorithm, (iii) a simulation exercise to test the algorithm under realistic field conditions, i.e. using the same water quality scenarios as will be given to respondents in the actual survey, and (iv) a latent class simulation exercise to evaluate the capacity of the model to identify different preference patterns and choice heuristics. These simulations were used to inform survey scenarios and evaluate the performance of the Bayesian estimators and search models. We are now in an excellent position to process the field data when they become available.
Future Activities:
Work during year five will focus complete all tasks described in the proposal. Tasks include completion pilot testing over a sample of approximately 100 – 150 respondents (Task V). These pilot tests will be used to make final adjustments to the questionnaire, MTurk interface and bid vector, prior to final survey implementation on MTurk (Task VI). We will also implement a subset of survey scenarios using a standard online panel for comparison (Task VI). We will proceed with data verification and Bayesian econometric analysis (Task VII), including evaluation of differences between MTurk and online panel results (Task VIII).
Journal Articles:
No journal articles submitted with this report: View all 6 publications for this projectSupplemental Keywords:
Media: water, watersheds; Ecosystem Protection: ecosystem, indicators, aquatic, habitat; Public Policy: decision making, cost-benefit, non-market valuation, contingent valuation, survey, preferences, public good, Bayesian, willingness-to-pay; Disciplines: social science, economics, ecology, hydrology; Methods/Techniques: modeling, analytical, surveys; Geographic Areas: Northeast, EPA Region 1; Other: ecosystem service, choice modeling, choice experiment, nonuse value, welfare analysis, water quality.Progress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.
Project Research Results
- Final Report
- 2020 Progress Report
- 2018 Progress Report
- 2017 Progress Report
- 2016 Progress Report
- Original Abstract
1 journal articles for this project