Brief Summary of Minutes of Annual Meeting

Summary of Minutes of Annual Meeting

 

Current President Steve Dundas called the meeting to order at 5:00 pm on 04/26/2022. There were four major elements discussed.

 

1. The first order of business was to discuss the successful re-charter of the group, W5133 "Economic Valuation and Management of Natural Resources on Public and Private Lands" from October 2022 to September 2027. Steve discussed the difficulty putting together the annual report because of the format of universities' individual reports. Each state's annual report should be succinct and follow a template matching the NIMSS format

 

2. The second order of business was discussion to respond to the request for additional outreach in the group. Several points included:

• Better showcasing our outreach (presenting work to agencies, resource managers, non-economists, etc.) and the impact (changed policy, behavior, etc.)
• A potential website to showcase past meetings and links to members' webpages.

 

3. The third order of business was to elect a new officer, an incoming Secretary. Corey Lang (Univ. Rhode Island) nominated himself. There were no other nominations and there was unanimous support for his nomination.

 

4. The fourth order of business was discussion of next year’s meeting. Several points of feedback were provided:

• Overall enthusiasm to be back in person and towards staying in-person even if it meant changing dates.
• If possible, a desire to return to meeting in February/early March on Wednesday-Friday schedule
• Steven Dundas will share the budgets with Jerrod Penn
• Jerrod Penn proposed San Antonio, TX or New Orleans, LA as potential sites for the 2023 annual meeting. Meeting participants were supportive of both locations.

 

 

Steven Dundas ends his three-year officer rotation. The group is especially thankful for his work on the recharter of W5133.  Jerrod Penn begins his term as President, Jesse Burkhardt transitions from Secretary to Vice-President, and Corey Lang begins his officer rotation as Secretary.

 

The meeting was adjourned at 5:45 pm.

 

The Accomplishments, Outreach, and Publications of W4133 in 2021 is based on reports from the following 15 institutions:

• Colorado State University
• Kansas State University
• Louisiana State University
• Michigan State University
• Mississippi State University
• Montana State University
• Northern Arizona University
• Oregon State University
• University of Georgia
• University of Illinois Urbana-Champaign
• University of Maine
• University of Rhode Island
• University of Tennessee
• University of Wyoming
• Virginia Tech

Accomplishments

 

Objective 1: Resource Management

Colorado State University:

• Manning and co-authors demonstrated that while natural disasters can affect beliefs about climate change and support for policies, impacts are small and inconsistent across disasters.
• Manning and co-authors completed research illustrating variation in the value of groundwater in the Midwest as well as how irrigators respond to electricity prices and nearby irrigated land retirement.

 

Kansas State University:

• (Task 1-1 and 2-1): We examined farmers' willingness to grow oilseeds under contract for use in biojet fuel production, which can help to mitigate climate change and improve conservation on the land scape through more advantageous crop rotations. We examined different oilseed crop (harvest windows, pesticide resistance, shatter resistance, and winter hardiness) and contract attributes (returns to production, contract length, crop insurance, cost-share, and risk) that farmers would desire. In addition, we estimated farmers' willingness-to-pay for these attributes and the likelihood that farmers would grow oilseeds under different contract scenarios. We find that net returns from oilseed production plays a very significant, impacting the likelihood of farmers growing and estimated supply of oilseed production that may occur for biojet fuel in the western U.S. Farmers prefer shorter more flexible contracts with cost share and crop insurance options. In addition, they are willing-to-pay for desirable crop attributes, such as pesticide resistance. Oilseed production for biojet fuel will compete with other oilseed production, as well. This research helps to meet project objectives by helping farmers, policymakers and agribusiness better understand the needs for establishing  a supply chain for biojet fuel at the farm level, the level of supply that may initially be provided for such an industry, how farmers can benefit in such a supply chain, conservation and land-use impacts from adoption, and how such operations can help farmers diversity income and lower risk from adoption of such enterprises in the face of a changing climate. Research was disseminated during extension presentation; research, findings and methodologies were presented to students in both graduate and undergraduate classes; and research findings were published in peer-reviewed journals and presented at professional research conferences. Students (undergraduate and graduate) received training on statistical methods, interdisciplinary research, and bioenergy production.
• (Task 1-1) We analyzed data about adoption of conservation practices, specifically no tillage, cover crops, and use of manure, by farmers in Kansas. Using statistical techniques to analyze the data, we found geography and soil characteristics and management influence adoption of conservation practices. In addition, farm characteristics, such as farm size and integration of livestock influence conservation practice adoption, particularly for use of manure. There does exist some complementarity, or association, between adoption of the practices examined. That is, the likelihood of adopting a second conservation practice is higher, if you have already adopted of other practices examined. The research helped to meet the goal of better understanding factors that influence conservation of agricultural land and the potential to increase conservation across the agricultural landscape. In addition, it helped to better understand how we can better tailor outreach efforts to agricultural producers to help increase conservation efforts on-farm. Research was disseminated during extension presentation; research, findings and methodologies were presented to students in both graduate and undergraduate classes; and research findings were published in peer-reviewed journals and presented at professional research conferences. Students (undergraduate and graduate) received training on statistical methods, interdisciplinary research, and conservation on agricultural lands.

 

Louisiana State University:

• Penn finished advising an MS student (Ana Gutierrez Castillo) on research related to forest easement landowners’ willingness to accept for conservation thinning to improve forest health in the Lower Mississippi Alluvial Valley in late 2020, with the publication occurring in early 2022.
• Penn advised data collection for a stated preference survey to understand forest landowner willingness to consult with forest consultants in order to pursue sustainable forest management practices (e.g., stream management, certification) in Spring 2021, which led to the completion of an MS student (Erika Largacha) in Fall 2021.

 

Michigan State University:

• Completed a stated preference study on the effects of fire on recreation visitation within a National Forest and completed another stated preference study of the effects of beach hazards on beach use on the Great Lakes.

 

Mississippi State University:

• Investigator Yun with Dr. Ayoung Kim has co-supervised an MS student (Jessica Browne) working on beach closure caused by Harmful Algal Blooms (HABs) and E. Coli impacts on the local economy. Investigator Petrolia is on the thesis committee, and Browne will defend her thesis proposal on June 15^th.
• Investigator Yun led the project building historical ecosystem services of water quality, habitat quality, carbon balance, crop yields, and bird diversity from various data sources as an output of USDA AFRI grant of the project “Sustainable Bioenergy Production and Integrated Valuation of Ecosystem Services. The data will be released through a website built in a subdomain of Mississippi State University. With investigator Interis, the paper studying crop diversity impact on water quality is under process. Investigators Yun and Petrolia have begun discussions and paperwork about the conceptually consistent use of the modern portfolio analysis on conservation.
• Investigator Petrolia supervised an MS student (Kelvin Amon) who completed thesis work focused on the impacts of FEMA’s Community Rating System. Investigator Yun served on the thesis committee, and student defended his thesis on June 10, 2021.
• Investigators Petrolia and Yun completed a project focused on the (hedonic) valuation of wind mitigation techniques for coastal properties, funded by the Mississippi-Alabama Sea Grant Consortium.
• Investigator Petrolia completed a project focused on an economic comparison of offshore versus nearshore sediment resources for coastal restoration, funded by the Bureau of Ocean Energy Management, in collaboration with researchers at Louisiana State University, the University of New Orleans, and the Water Institute of the Gulf. The project analyzed how sediment quality (grain size) impacts project performance and the economic tradeoffs between quality/performance and dredging/transport costs.

 

Montana State University:

• Academic publication: Bigelow, D.P., D.J. Lewis, and C. Mihiar. 2022. "A major shift in U.S. land development avoids significant losses in forest and agricultural land." Environmental Research Letters, 17 (2).
• Contribution to stakeholder report: USDA-NRCS, Montana Association of Land Trusts, and Heart of the Rockies Initiative. 2022. Working for Montana Agriculture: Economic Benefits of Conservation Easements for Montana’s Farms, Ranches, and Communities. Web link: Working for Montana Agriculture (montanalandtrusts.org).
  • I wrote the two subsections appearing on page 13 of the report, as well as a more detailed technical summary appearing on the Montana Association of Land Trust website.
  • The report was prepared in advance of Montana’s next state legislative session in order to inform policy discussions surrounding the tax treatment of land with conservation easements.
• Co-advised M.S. thesis committee for Hannah Wing, who wrote a thesis on the linkages between Conservation Reserve program contract expiration and organic practice adoption. Hannah had her work accepted for a presentation at the 2022 Western Agricultural Economics Association Summer Conference in Santa Fe. We are currently in the process of formatting her thesis for a journal submission.
• Advised M.S. thesis committee for Conner McCollum, who wrote a thesis on the relationship between conservation easement adoption and on-farm investment.
• Completed working paper on farmland tax policy and submitted to academic journal.
  • Bigelow, D.P. and T. Kuethe. 2022. "The impact of preferential farmland taxation on local public finances." (revise and resubmit at Regional Science and Urban Economics).

 

Oregon State University:

• Alix-Garcia and collaborators published a literature review about alleviating poverty in forests and a summary of published article prepared in advance of the COP26 meeting about how deforestation alerts can reduce emissions from land use change.

 

University of Georgia:

• Designed survey instrument to assess WTA coastal buyouts and WTP for “rent-backs” along the coast (Dare County, North Carolina and Worcester County, Maryland): Perry and Landry. “Willingness to Accept Buyout and Willingness to Pay for Rentbacks in the Context of Coastal Adaptation”
• Conducted analysis to assess the impact of hurricanes on NFIP market penetration in Georgia (which is generally considered a coastal “safe haven”): Grist, Landry, and “Flood Insurance Market Penetration in Georgia Before and After Hurricane Irma: A Difference-in-Differences Analysis”
• Directed research into identification of price gouging in the wake of hurricanes: Mensah, Landry, and Smith. “Efficient and Inefficient Pricing in the Wake of a Catastrophe”
• Directed research into barriers to trading in carbon markets: Yoon, Filipski, Landry, and Yoo. “Endowment Effects, Expectations, and Trading Behavior in Carbon Cap-and-Trade”
• Analyzed novel choice experiment data to assess WTP for rebuilding and WTA buyout post-storm: Landry, Keeler, and Scyphers. “Household Preferences for Post-storm Coastal Adaptation: An Application of Choice Experiments”
• Directed research into measurement of risk perceptions and assessing their role in structural models of decision making under risk:
• Turner and Landry. “The Nature of Coastal Hazard Risk Perceptions” Working Paper University of Georgia: Athens, GA
• Turner and Landry. “Accounting for Uncertainty in Decision Weights for Experimental Elicitation of Risk Preferences” SSRN: https://ssrn.com/abstract=3882694
• Turner and Landry. “Structural Estimation of Decision Making under Natural Hazard Risk” SSRN: com/abstract=3574835

 

University of Illinois Urbana-Champaign:

• Ando worked with Corey Lang to draft a paper using data on conservation and house values in Massachusetts to estimate how the capitalization benefits of land conservation accrue across racial groups.
• Ando worked with a team of coauthors from a wide range of states on a paper submitted to the Review of Environmental Economics and Policy about how management and valuation research can improve outcomes for endangered migratory species in the face of climate change
• Ando finalized a paper with Liqing Li on the impacts of bison restoration on local economies and submitted it to Agricultural and Resource Economics Review. It has been invited for revision and resubmission.
• Ando wrote and published a paper in Science Advances with a team of ecologists and economists to show the benefits that investments in nature could have through reducing the likelihood of future pandemics.

 

University of Maine:

• Over the reporting period, Maine researchers achieved progress on multiple active studies, and shared findings and updates with key stakeholders. We continued to study resource management behaviors and intentions of landowners, examining, among other factors, their participation in estate planning programs. We compared patterns in the use of wills and trusts using data collected in regions of Maine, Massachusetts, New York, and Vermont to learn more about differences in estate planning across the Northern Forest Region. Observations and findings from our work offer guidance to stakeholders and researchers engaging with landowners to achieve long-term resource management goals and to plan for and implement ownership transitions in the face of climate change and other shocks. We continue to improve empirical models designed to forecast landscape outcomes and simulate policy impacts. During the reporting period, Maine researchers published research on improving the designation of urban-rural gradients and land-scape scale policy simulations. Our findings suggest that commonly used urban-rural metrics are not capturing differences over time and space that matter for conservation and natural resource management. We initiated discussions with regional stakeholders to better understand the implications of these findings for future extension activities and policy analyses, including the sensitivity of policy simulations to assumptions about land-use gradients. We also contributed to ongoing projects studying the resource management decisions of Maine’s maple and honey producers and assessing the potential for small producers to scale up their enterprises.

 

University of Rhode Island:

• Analyzed survey data regarding resident preferences for belonging in a multi-state greenhouse gas mitigation program. We hypothesized that status quo bias would impact policy preferences and results confirmed that. Article was published in Environmental Research Letters.

 

Virginia Tech University:

• Coastal flood risk: Moeltner continued work on Bayesian matching estimators for housing markets, with application to flood risk in New England. The main objective for this research is to determine the optimal number of matched control observations for each treated property, along with other matching parameters. This past year Moeltner expanded my model diagnostics to also include distance statistics, balance statistics, and regression adjustment effects. Moeltner continued to run and compare a variety of different specifications.

 

University of Wyoming:

• Project: 2019 “The Comprehensive Survey of Snowmobile and Off-Road Recreational Vehicle (ORV) Use Trends and Economic Impacts of the Wyoming State Trails System.” Wyoming Dept. of Commerce, Division of State Parks and Historic Sites, Cheyenne Wyoming, $114,000 PI: C. T. Bastian, Co-PIs: R.H. Coupal, A.M. Nagler, and D.T. Taylor) (Project Awarded Funding Nov. 2019). Duration Jan. 2020- Dec. 2022.
  • Milestones: Survey research completed regarding 2020-2021 snowmobile season recreational use, and survey research completed regarding 2021 ORV recreational use.
  • Activities: Task 1.1 Preliminary analysis of snowmobile survey has been completed for the state trails report. A report regarding descriptive statistics of survey responses, trails usage, and economic impacts of snowmobiling in Wyoming was drafted. The snowmobile survey data will be used in an MS thesis project to be completed in 2023.
    

Objective 2: Economic Valuation

 

Colorado State University:

• Jesse Burkhardt is working with the Colorado Department of Parks and Wildlife (CPW) to estimate Colorado residents’ willingness to pay for wolf reintroduction. We are also evaluating preferences for wolf management practices. The results will be used by the CPW to design a wolf management plan and to inform a cost-benefit analysis.
• Jesse Burkhardt estimated the value of conformity in landscape choices using a hedonic model.
• Manning completed a project with w4133 member, Amy Ando, to value the loss of bats to agricultural producers in the US.

 

Louisiana State University

• Penn worked with Petrolia on comparing convenience and probability samples for the value of beach condition information among Gulf of Mexico beachgoers. Penn also published methodological work on increasing comprehension of discrete choice experiment instructions and hypothetical bias mitigation.

 

Michigan State University

• Developed a statewide water quality valuation model with spatially-explicit water quality models developed at multiple water shed scales from hydrologic unit codes of HUC 10 to HUC 4.
• Made progress on developing an integrated assessment model for non-point pollution control in Michigan linking on-farm economic decision making and behaviors to on and off-farm impacts to assess benefits and costs of policies to address agricultural non-point source pollution in freshwaters including inland lakes and streams and the Great Lakes.

 

Mississippi State University:

• Investigators Interis and Yun have explored methods of combining observations of both discrete and marginal willingness to pay values in benefits transfer analysis in order to increase benefit transfer validity. This work will be presented at the 2022 annual meeting of the Northeastern Agricultural and Resource Economics Association.

 

Oregon State University:

• Alix-Garcia and collaborators published a paper testing the validity of indirect surveying as a method to collect household data. They compare household and informant reports of assets, develop poverty indices from both, test their performance as regression covariates, and examine errors in reporting and targeting resulting from using indirectly reported variables. Informant-based targeting indices are highly correlated with household measures and can be reasonable substitutes for self-reported indices in simple regressions. They can also be used to assign a simulated anti-poverty program with similar error rates to related methods. In their setting, eliminating direct household surveys would have reduced survey costs by 50%.
• Kling, Dundas and interdisciplinary collaborators published a paper using a choice experiment to examine public preferences for coastal dune ecosystem restoration in the US Pacific Northwest, a public good whose natural state is now rare. Respondents are asked to choose among hypothetical projects that vary by project size, restoration quality, recreation access, flooding risk, and cost. Restoration quality is defined as closeness to the natural ecosystem. We find that increasing restoration quality results in significantly higher welfare gains than increasing the size of restoration area. Maintaining recreation access is preferred, and programs with recreation restrictions yield positive willingness-to-pay only if accompanied by the highest restoration quality.
• Kling and Daniel K. Lew (NOAA Fisheries) co-supervised Russel Dame's (OSU Ph.D. student) research on Alaskan recreational angler behavior. They develop a model of angler site choice that allows for the prediction of days spent on site in addition to site choice. This model allows for more comprehensive welfare measures associated with harvest rates. The model can also predict recreational harvest mortality, which is a key metric for fisheries management for halibut and salmon species. Dame will present this research at the upcoming International Institute of Fisheries Economics and Trade (IIFET) meeting in Spain.

 

University of Georgia:

• Launched survey to assess novel application of “consequentialism” in valuing private goods: Landry, Schaffer, Roy, and Campbell. “Willingness to Pay for Local Herbs: Combatting Hypothetical Bias with a New Variant of Consequentiality”
• Directed research into the effects of dam removal on riverine property values: Krasovskaia, Landry, Lewis, and Filipski. “Hedonic Analysis of Dam Removal: Evidence from Matching Methods in Maine, USA”

 

University of Illinois Urbana-Champaign:

• Ando finalized and submitted a paper with Liqing Li on the influence of early childhood experience on the values people place on grassland restoration.
• Ando worked with graduate student Fred Nyanzu and University of Rhode Island colleague Corey Lang to write and focus group the survey for a paper that estimates the values people place on local food and peri-urban conservation of natural and farmland areas.
• Ando submitted, revised, and resubmitted a paper to Landscape and Urban Planning with Oregon co-authors including Sahan Dissanayake and Noelwah Netusil on the values people place on different features of green roof installations in urban areas.
• Ando did the data analysis with Kaylee Wells, Sahan Dissanayake, and partners at the USDA for a paper on the values of grassland restoration projects in the tallgrass prairie region. They have also drafted a survey to estimate the benefits of landscape level changes in grassland prevalence across the entire U.S.
• Ando revised and resubmitted a paper with Dale Manning to the Journal of the Association of Environmental and Resource Economists estimating the losses to the agricultural sector from reduced bat populations caused by White-Nose Syndrome.
• Ando worked with a team of coauthors including W4133 partners from Colorado and Mississippi on a paper submitted to the Review of Environmental Economics and Policy showing how methods commonly used in environmental economics – including non-market valuation techniques – can support policies and management actions that perpetuate racial inequity in access to environmental quality.
• Ando gave several presentations about research related to this objective:
  • August 2021. “Critical variation in preferences for environmental goods.” Invited keynote talk, 27th Ulvön Conference on Environmental Economics (virtual), Sweden.
  • July 2021. “Distributional issues in environmental valuation findings and practice.” LEEPout seminar series; Land, Environment, Economics and Policy Institute; University of Exeter (virtual), UK.
  • April 2021. “Early Exposure to Nature and WTP for Conservation.” L. Li and A. Ando. Dept. of Agricultural Resource and Economics, University of California Berkeley.

 

University of Maine:

• Changing recreation behaviors during COVID-19 have sparked considerable interest in improving documentation and analysis of recreation visits and economic valuation of changing recreation opportunities. Working with regional stakeholders, we made progress on two research projects during the reporting period. In collaboration with the Land Trust Alliance, Maine researchers prepared a review of recent literature that used cell-phone and/or social media data to assess recreation visitation and, in some instances, model recreation demand. We shared the results with the Land Trust Alliance and worked together to summarize the pros and cons of using such data resources for different types of land trusts and valuation work. In addition, Maine researchers are collaborating with state and regional trail organizations to pilot the use of satellite/cell-phone based data streams to document trail use in different recreational settings. We are comparing and contrasting counts based on survey and in-place counters with results based on satellite/cell phone data streams. As we continue to make progress on this research, we will support the development of improved recreation demand valuation methods, and also call attention to the use of and services provided by diverse trail systems and park lands in Maine and beyond.

 

University of Rhode Island:

• Analyzed exit poll and aggregate voting data to assess bias stemming from use of aggregate data. Due to voter selection and omitted spatial variables, using aggregate data to estimate preferences leads to biased results. Article was published in Journal of Environmental Economics and Management.
• Analyzed how implicit value of open space varies across cities. We used the hedonic price method and data from over 200 MSAs in the Eastern United States and estimate MSA-specific valuation. We then examine variation in MWTP and city-level determinants. Valuation is largely local – relatively large changes in income or existing conservation lead to modest changes in MWTP – suggesting validity of benefit transfer across regions. However, geographic features that naturally limit development do correlate with MWTP. As a result, we examine geographic features as instrumental variables, and find that on average steep slope and water/wetlands yield valuation coefficients of opposite sign, consistent with a LATE interpretation. Article was published in Land Economics.

 

Virginia Tech University:

• Bayesian Meta-Regression: During the 2020/2021 reporting period, I continued work on a novel Bayesian Locally-Weighted Meta-Regression (BLWRMRM) approach to value water quality improvements as a result of USDA conservation efforts on agricultural lands. This analytical framework produces more accurate benefit transfer estimates than typical one-model-fits-all MRM strategies. I applied the model to an existing (and growing) meta-data set on willingness-to-pay (WTP) for water quality improvements from a variety of source studies. I developed programming code to test model performance and generate locally-weighted benefit transfer (BT) predictions for the monetary value of water quality improvements. I found that the new framework performs better in all statistical aspects considered than conventional meta-regression and BT based on a single model.
• Free-form choice experiments: Building on results from the preceding research year, I developed models that also include water quality information based on survey respondents' browsing through the interactive maps offered in our online questionnaire. This led to important discoveries, most notably that not only water quality near a respondent's home matters, but also quality in potentially distant areas that matter to the household for various reasons. If this factor is ignored, resulting willingness-to-pay estimates will be biased downwards.

 

 

Objective 3: Integrated Policy and Decision-Making

 

Colorado State University:

• Jesse Burkhardt is working with Lynker Technologies to estimate the causes and consequences of water shortages in Colorado.

 

Mississippi State University:

• Investigator Yun as a co-PI awarded the USDA-NIFA AFRI grant for the project, “Dynamic Portfolio of Spatial and Temporal Forest-Based Biodiversity Conservation.” The lead PI of the grant is a W4133 investigator, Dr. Sunghoon Cho at University of Tennessee, Knoxville.
• Investigator Petrolia supervised an MS student (Freedom Enyetornye) working on valuation of oyster reef restoration across the Gulf Coast. Investigator Yun is on the thesis committee, and student will present his thesis proposal on June 30, 2022.
• Investigator Petrolia supervised an MS student (Barbara Okai) working on an application of benefits transfer to conduct a benefit-cost analysis of coastal restoration projects in Alabama and Mississippi. Investigators Interis and Yun are on the thesis committee, and student will present her thesis proposal on June 10, 2022.
• Investigator Petrolia completed a project quantifying and monetizing ecosystem services provided by four oyster habitats: bottom oyster beds, off-bottom oyster farms, non-harvested bottom reefs, and living shorelines.  A spreadsheet-based tool was constructed for resource managers to calculate estimates.
• Investigators Petrolia and Yun continued work on a project focused on valuation of oyster reef restoration in the Gulf Coast, funded by NOAA Restore Science.

 

Montana State University:

• Completed working paper on Ricardian climate impact estimation and submitted to academic journal.
  • Bigelow, D.P. and M. Jodlowski. 2022. "Self-reporting and aggregation bias in Ricardian climate impacts: Evidence from observed farmland sales." (under review at American Journal of Agricultural Economics).

 

Oregon State University:

• Lewis, Kling, Dundas and a collaborator at NOAA Fisheries (D. K. Lew) published a paper developed a new empirical approach combining a choice experiment and a structural model to estimate two key parameters in a dynamic willingness-to-pay function for threated species abundance. The method enables estimation of the current marginal benefit of increases in threatened species abundance and the rate implicitly used to discount future marginal benefits and is demonstrated with an application to a threatened Coho salmon along the Oregon coast.

 

University of Maine

• Maine researchers contributed to multiple interdisciplinary research projects focused on improving land, forest, and water resource management as well as community economic development through integrated policy and decision-making. Working with colleagues from Ohio, Connecticut, and Virginia, Maine researchers drafted multiple manuscripts summarizing our research on the vulnerability and resilience of US rural forested communities and the roles of ecosystem services in these changing communities. Maine researchers also made progress in our interdisciplinary research of municipal capacity to respond to natural hazards and other crises. During the reporting period, Maine researchers published one manuscript assessing the response of 100 Maine municipalities to the COVID-19 pandemic. Working collaboratively with external partners, including state agency, regional planning and economic development, and non-profit organizations, Maine researchers hope to learn from the experiences of municipalities during the pandemic to improve future municipal programs and decision-support tools aimed at helping communities adapt to climate change and other shocks. Lastly, Maine researchers linked project research to ongoing integrated policy and decision-making coursework and research at the University of Maine by participating in two NSF-funded interdisciplinary and convergent graduate student training programs (One Health and System Approaches for Understanding and Navigating the New Arctic) and two University of Maine System funded interdisciplinary and engaged undergraduate research projects.

 

University of Tennessee

• The objective of the research is to identify optimal conservation investment allocations for target regions and species within those regions under uncertain conditions. We develop a two-step approach using modern portfolio theory (MPT) to estimate percentages of conservation investment (referred to as ‘portfolio weights’) for counties and taxonomic groups in the central and southern Appalachian region under climate and market uncertainties.

 

University of Wyoming:

• Task 3.2 Economic Assessment of Nonpoint Source Pollution Management in the Tongue River Basin, WY. Local BMPs matched with WTP levels derived from meta-analysis. Balint, K., R. Coupal, K. Hansen, G. Paige and D. McLeod. Ongoing.
  • Milestones: Task 3.2 Completion of Balint’s MS Thesis (2019) “Economic Assessment of Nonpoint Source Pollution Management in the Tongue River Basin, WY.” Presentation at W4133 February 2020 Athens, GA Manuscript in process at present.

Publications (60 total)

Objective 1: Resource Management (35)

1. Adhikari, R.K., R.K. Grala, D.R. Petrolia, S.C. Grado, D.L. Grebner.   "Landowner concerns related to availability of ecosystem services and environmental issues in the southern United States."  Ecosystem Services 49(June): 101283.
2. Adhikari, R., R. Grala, S. Grado, D. Grebner, D.R. Petrolia. “Landowner Satisfaction with Conservation Programs in the Southern United States.”  Sustainability 14(9): 5513.
3. Adhikari, R., R. Grala, D.R. Petrolia, S. Grado, D. Grebner, and A. Shrestha.   "Landowner Willingness to Accept Monetary Compensation for Managing Forests for Ecosystem Services in the Southern United States."  Forest Science 68(2): 128-44.
4. Ando, A.S. and C. Reeser. 2021. “Homeowner willingness to pay for a pre-flood buyout agreement.” Land Economics uwpress.org/content/early/2022/01/20/le.98.4.052721-0056.full.pdf
5. Bernstein, A., T. Loch-Temzelides, A. Ando, P. Daszak, M. Vale, B. Li, C. Chapman, M. Kinnaird, K. Nowak, H. Li, J. Busch, M. Castro, C. Zambrana-Torrelio, L. Xiao, L. Kaufman, S. Pimm, L. Hannah, A. P. Dobson. “The scale of justified investments in primary pandemic prevention.” Science Advances.
6. Bigelow, D.P., D.J. Lewis, and C. Mihiar. 2022. "A major shift in U.S. land development avoids significant losses in forest and agricultural land." Environmental Research Letters, 17 (2).
7. Caffey, R.H., D.R. Petrolia, I. Georgiou, M. Miner, H. Wang, and B. Kime. "The economics of sediment quality on barrier shoreline restoration."  Revise & resubmit.
8. Canales E., J.S. Bergtold and A. Featherstone. 2021. Farm efficiency and productivity growth: The effect of commodity prices. Journal of Applied Farm Economics 4(1): DOI: 10.7771/2331-9151.1053
9. Chen, Y., Lewis, D.J., and B. Weber. 2021. “Amenities and skill sorting: A case study of land conservation policy.” The Annals of Regional Science, 67: 649-669. org/10.1007/s00168-021-01060-3.
10. Goeb, J., and F. Lupi, 2021, Showing pesticides’ true colors: The effects of a farmer-to-farmer training on pesticide knowledge, Journal of Environmental Management. 279: 111821.
11. Gutierrez A, Penn J, Tanger S, and M Blazier. 2022. “Conservation Easement Landowners’ Willingness to Accept for Forest Thinning.” Forest Policy & Economics. 135: 102627.
12. Hanson, E., A. Nagler, J. Ritten, B.S. Rashford.   Farm-level Economics of Bioenergy in the Upper Missouri River Basin. Journal of the American Society of Farm Managers and Rural Appraisers 2022: 16-30.
13. Hashida, Y., and D.J. Lewis. 2022. “Estimating welfare impacts of climate change using discrete-choice models of land management: An application to western U.S. forestry.” Resource and Energy Economics, 68: 101295. org/10.1016/j.reseneeco.2022.101295
14. Hrozencik, Aaron, Dale T. Manning, Jordan F. Suter, Christopher Goemans (2021). Impacts of Block-Rate Energy Pricing on Groundwater Demand in Irrigated Agriculture. American Journal of Agricultural Economics.
15. Jensen, A. J., S. J. Dundas, and J. T. Peterson. 2022. Phenomenological and Mechanistic Modeling of Recreational Angling Behavior Using Creel Data. Fisheries Research 249: 106235. doi: 1016/j.fishres.2022.106235
16. Jones Ritten, C., A. Nagler, K.M. Hansen, D.E. Bennett, and B.S. Rashford.   Incorporating Landowner Preferences into Successful Migratory Species Conservation Policy.  Western Economics Forum 21(1): 83-94.
17. Kaminski, A., Bauer, D.M., Bell, K.P. et al. Using landscape metrics to characterize towns along an urban-rural gradient. Landscape Ecology 36, 2937–2956 (2021). https://doi.org/10.1007/s10980-021-01287-7
18. Laird, H., C. Landry, S. Shonkwiler, and D.R. Petrolia.   "Riders on the Storm: Hurricane Risk and Coastal Insurance and Mitigation Decisions."  Journal of Ocean and Coastal Economics 8(1): Article 3.
19. Landry, C.E., D. Turner, and D.R. Petrolia.   "Flood Insurance Market Penetration and Expectations of Disaster Assistance."  Environmental & Resource Economics 79: 357-86.
20. LandryE., S. Anderson, E. Krasovskaia, and D. Turner. 2021. “Willingness to Pay for Multi-Peril Hazard Insurance” Land Economics (Nov) 97(4). doi:10.3368/le.97.4.072820-0115R1
21. Logan, L.H., R.S. Gupta, A.W. Ando, C. Suski, A.S. Stillwell. 2021. “Quantifying tradeoffs between electricity generation and fish populations via population habitat duration curves.” Ecological Modelling 440: 109373. org/10.1016/j.ecolmodel.2020.109373
22. Lupi, F., R. von Haefen and L. Cheng. 2021. “Distributional effects of entry fees and taxation for financing public beaches in Michigan.” Land Economics.
23. Mihiar, C., and D.J. Lewis. 2021. "Climate, adaptation, and the value of forestland: A national Ricardian analysis of the United States." Land Economics, 91(4): 911-932. org/10.3368/le.97.4.011620-0004R1.
24. Moffette, F, J. Alix-Garcia, K. Shea, A.H. Pickens. 2021. Freely available deforestation alerts can reduce emissions from land-use change. Nature Climate Change 11(11): 913-14.
25. McCollum, C.J., S.M. Ramsey, J.S. Bergtold and G. Andrango. 2021. Estimating the supply of oilseed acreage for sustainable aviation fuel production: taking account of farmers' willingness to adopt. Energy, Sustainability and Society 11:33.
26. Nagler, A., J. Bannon, and B.S. Rashford.   Landowner and Economic Benefits from Migration Corridor Designation in Wyoming.  UW Extension Bulletin, B-1374.2, August 2021.
27. Petrolia, D. R., S. G. Ishee, S. D. Yun, J. R. Cummings, and J. Maples, 2022, “Do Wind Hazard Mitigation Programs Affect Home Sales Values?” Journal of Real Estate Research, forthcoming.
28. Reem H., P. Newton, M. Ihalainen, A. Agrawal, J. Alix-Garcia, S. E. Castle, J. T. Erbaugh, M. Gabay, K. Hughes, S. Mawutor, P. Pacheco, G. Schoneveld, J.A. Timko. 2021. Levers for alleviating poverty in forest. Forest Policy and Economics 132: 102589
29. Rouhi Rad, M., D.T. Manning, J.F. Suter, C. Goemans (2021). Policy Leakage or Policy Benefit? Spatial Spillovers from Conservation Policies in Common Property Resources.  Journal of the Association of Environmental and Resource Economists.
30. Sheng, G., J.S. Bergtold and E. Yeager. 2021. Assessing the joint adoption and complementarity between in-field conservation practices of Kansas farmers. Agricultural and Food Economics 9: org/10.1186/s40100-021-00201-8 .
31. Sampson G.S., A. Al-Sudani and J.S. Bergtold. 2021. Local irrigation response to ethanol expansion in the High Plains Aquifer. Resource and Energy Economics 66: org/10.1016/j.reseneeco.2021.101249.
32. Sloggy, M., J. Suter, M. Rouhi Rad, D.T. Manning, and C. Goemans (2021). Changing climate, changing minds? The effects of natural disasters on public perceptions of climate change.  Climatic Change.
33. Smith D., Caffey R., S. Midway, and J. Penn. 2022. “Economic values of potential regulation changes for the declining Southern Flounder fishery in Louisiana.” Marine and Coastal Fisheries. 14(2): e10195.
34. Suter, J., M. Rouhi Rad, D. Manning, C. Goemans, and M. Sanderson (2021). Groundwater Depletion, Climate, and the Incremental Value of Groundwater. Resource and Energy Economics.
35. von Haefen, R., and F. Lupi, 2021, How does congestion affect the evaluation of recreational gate fees? An application to Gulf Coast Beaches. Land Economics.

 

Objective 2: Economic Valuation (15)

1. Alix-Garcia, K. RE Sims, and L. Costica. 2021. Better to be indirect? Testing the accuracy and cost-savings of indirect surveys. World Development 142: 105419
2. Barrett, L.T., S.J. Theuerkauf, J.M. Rose, H.K. Alleway, S.B. Bricker, M. Parker, D.R. Petrolia, and R.C. Jones.   "Sustainable growth of non-fed aquaculture will generate valuable ecosystem benefits."  Ecosystem Services 53(February):  101396.
3. Blanchette, A., Lang, C., & VanCeylon, J. (2021). Variation in Valuation: Open Space and Geography. Land Economics, 011720-0005R.
4. Burkhardt, J., N. W. Chan, B. Bollinger, and K. Gillingham. "What is the value of conformity? Evidence from home landscaping and water conservation."  J. Agric. Econ(2021).
5. Landry C.E., Turner, D., and T. Allen. 2021. “Hedonic Property Prices and Coastal Beach Width” Applied Economic Policy & Perspectives. org/10.1002/aepp.13197
6. Lang, C., & Pearson-Merkowitz, S. (2022). Aggregate data yield biased estimates of voter preferences. Journal of Environmental Economics and Management, 111, 102604.
7. Lang, C., Weir, M., & Pearson-Merkowitz, S. (2021). Status quo bias and public policy: evidence in the context of carbon mitigation. Environmental Research Letters, 16(5), 054076.
8. Manning, D. and A. Ando. Ecosystem Services and Agricultural Land Rental Markets: The Producer Cost of Bat Population Crashes. Journal of the Association of Environmental and Resource Economists (forthcoming).
9. Nguyen, T., D. M. Kling, S. J. Dundas, D. Hacker, D. K. Lew, P. Ruggiero, and K. Roy. Forthcoming. Quality over Quantity: Non-market Values of Restoring Coastal Dunes in the US Pacific Northwest. Land Economics. doi: 10.3368/le.040721-0036R
10. Penn J. and W Hu. 2021. “Videos Improve Attention and Cheap Talk in Online Surveys.” Frontiers in Economics in China. 16(2): 347-376.
11. Penn, J., D.R. Petrolia, J.M. Fannin. "Comparing Representative and Convenience Samples for Values of Beach Condition Information."  Working paper.
12. Petrolia, D.R., D. Guignet, J.C. Whitehead, C. Kent, K. Amon, and C. Caulder.   "Nonmarket Valuation in the Environmental Protection Agency's Regulatory Process."  Applied Economic Perspectives & Policy 43(3): 952-69.
13. Dahal, R.P., R.K. Grala, S. Gordon, I.A. Munn, and D.R. Petrolia. 2021.  "Geospatial Heterogeneity in Monetary Value of Proximity to Waterfront Ecosystem Services in the Gulf of Mexico."  Water 13(17): 2401.
14. Yun, S. D. and B. M. Gramig, 2022, “Spatial Panel Models of Crop Yield Response to Weather: Econometric Specification Strategies and Prediction Performance,” Journal of Agricultural and Applied Economics, 54(1): 53-71.
15. Yun, S. D. and A. Kim, 2021, “Economic Impact of Natural Disasters: A Myth or Mismeasurement?” Applied Economics Letters, 29(10): 861-866.

 

Objective 3: Integrated Policy and Decision-Making (10)

1. Blachly B, Ferreira S, Herman B, Grace Ingham A, Landry C.E. and A.J. Perry. 2021. Developing an Inventory of US Army Corps of Engineers’ Nature-based Infrastructure Projects. DOER Technical Notes Collection (DOER-EWN-XX), U.S. Army Engineer Research and Development Center, Vicksburg, MS. doer.el.erdc.dren.mil.
2. Berenter, J., Morrison, I., and Mueller, J.M. (2021). “Valuing user preferences for geospatial fire monitoring in Guatemala.” 13:12077.
3. Jennings, K.S., A.N. Wlostowski, R.E. Bash, J. Burkhardt, C.W. Wobus, and G. Aggett. "Data availability and sector‐specific frameworks restrict drought impact quantification in the Intermountain West." Wiley Interdisciplinary Reviews: Water(2022): e1586.
4. Kang, N., C. Sims, and S. Cho. 2022. “Spatial and taxonomic diversification for conservation investment under uncertainty.” Environmental Conservation, 1-8. doi:10.1017/S0376892922000194.
5. Levesque, V.R., Johnson, E.S., and K.P. Bell. 2021. Municipal Capacity to Respond to COVID-19: Implications for Improving Community Resilience in Maine. Maine Policy Review2 (2021): 62 -71, digitalcommons.library.umaine.edu/mpr/vol30/iss2/8.
6. Lewis, D. J., D. M. Kling, S. J. Dundas, and D. K. Lew. 2022. Estimating the Value of Threatened Species Abundance Dynamics. Journal of Environmental Economics and Management 113: 102639. doi: 1016/j.jeem.2022.102639
7. Mueller, J.M., Loomis, J.B., Richardson, L., and Fitch, R.A. (2021) “Valuing impacts of proximity to Saguaro National Park on house prices.” Applied Economic Perspectives and Policy. org/10.1002/aepp.13196
8. Petrolia, D.R., W.C. Walton, and J. Cebrian.   "Oyster Economics:  Simulated Costs, Market Returns, and Nonmarket Ecosystem Benefits of Harvested and Non-Harvested Reefs, Off-Bottom Aquaculture, and Living Shorelines." Marine Resource Economics 37(3).
9. Shaw, G. R., and S. J. Dundas. 2021. Socio-Economic Impacts of the Southern Flow Corridor Restoration Project: Tillamook Bay, Oregon. Garibaldi, OR: Tillamook Estuaries Partnership. 47 pp. doi: 25923/w8c8-v180
10. Wlostowski, A.N., Jennings, K.S., Bash, R.E., Burkhardt, J., Wobus, C.W. and Aggett, G., (2021). Dry landscapes and parched economies: A review of how drought impacts nonagricultural socioeconomic sectors in the US Intermountain West. Wiley Interdisciplinary Reviews (WIRES): Water, p.e1571.

 

2022 Meeting Abstracts

 

SESSION 1: Stated Preference Methods

 

Title:               Controlling for attribute non-attendance using self-reported attribute non-attendance and response time information

Authors:         Jake Kennedy, Sahan Dissanayake, Randy Bluffstone

W4133 obj.:   2 (task 2.1)

Presenter:      Jake Kennedy

Email:             jakekennedy@pdx.edu

Abstract:       

The use of choice experiment surveys to elicit preferences and values for non-market goods and policies has increased significantly in the last decade. At the same time there is growing concern of the presence of attribute non-attendance (ANA), the occurrence where a respondent ignores one or more attributes. Previous studies have cast doubt on the trustworthiness of self-reported attendance information and shown improved model fit using inferred attribute non-attendance methodologies. However, existing literature has also shown a correlation between response time and cognitive effort as well as response time and self-reported attendance. We further contribute to this literature first by looking at the connection between response time and self-reported attendance information and second by exploring how that information can be jointly used to improve estimation in choice experiments. Our application is from an online survey conducted in the summer of 2020 concerning park management strategies for Forest Park, a public park located in Portland, Oregon. The data set consists of 822 responses from a choice experiment consisting of six attributes including the cost attribute and thirty-six unique choice tasks that were divided into six unique surveys. From our exploration of the literature, this study also contributes the first evidence of the impact of attribute non-attendance from the context of recreational preferences for urban greenspace. Initial exploration of the data shows pervasive ANA ranging from 35.9-50.7% of responses ‘at least sometimes’ not attending to an attribute with a particularly concerning 32.2% of responses ignoring the cost attribute. Finally, preliminary results show statistically significant but mixed directional effects on response time from self-reported attendance information per attribute prompting further exploration.

 

Title:               Assessing Hypothetical Bias in Nudging: Willingness to Pay for Consultation towards Improved Forest Management

Authors:         Sapana Bastola, Jerrod Penn, Michael Blazier

W4133 obj.:   1 (task 1-1), 2 (Task 2-1)

Presenter:      Jerrod Penn

Email:             jpenn@agcenter.lsu.edu

Abstract:           

Researchers have applied the principles from behavioral economics and psychology to nudge people towards a desired outcome. Stated Preference studies have also begun to adopt nudges, commonly framed as informational treatments, significantly altering willingness to pay in a wide array of settings. However, the observed effect of information treatments in SP studies may be subject to hypothetical bias such that the real impact is potentially smaller. To explore this issue, we examine to what extent the effect of nudge information treatments in Stated Preference studies is overstated due to hypothetical bias by eliciting both hypothetical and real WTP.

We examine the presence of HB for “Social Norm” and “Financial Incentives” information. These represent two different, but commonly implemented information nudges. This occurs in the context of understanding family forest landowners’ maximum WTP to receive a consultation with a private forest consultant, an important step towards obtaining a forest management plan and forest certification. While family forest landowners own more than 275 million acres of forestland in the US, adoption of among these de facto forest managers of sustainable forest management practices remains extremely low.

Each participant was randomly assigned to one of three treatments: 1) Social Norm 2) Financial Incentive, and 3) Control. In Social Norm, landowners received information about the number of certified farms and certified acres of forestland specific to their county as well as the same information in the three leading counties in their state. In Financial Incentive, landowners are told about three financial incentives of working with a private industry forester. Landowners in Control did not receive either of the information nudges. We used a payment card approach to elicit landowners’ WTP to receive consultation in both elicitation settings and utilized a Becker-DeGroot-Marschak (BDM) auction mechanism in the real setting.

The survey was distributed via mail in Spring 2021 to 4507 family forest landowners with property in four Louisiana parishes and four Arkansas counties with 10-600 acres of land. We received 578 responses from landowners, yielding a response rate of 14.1% and roughly equal participation in the three treatments.

Our preliminary results show that hypothetical WTP is higher than real WTP indicating the presence of hypothetical bias. Compared to control, Social Norm significantly increases WTP for consultation. However, the effect of Social Norm information is significantly higher in real elicitation than in hypothetical elicitation. Financial Incentive has no significant effect on WTP and there is no significant difference in the effect of Financial Incentive on hypothetical and real WTP.

 

Title:   Comparing Water Quality Valuation Across Probability and Non-Probability Samples     

Authors: Frank Lupi 

W4133 obj.:   Task 1-1 & Task 2-1

Presenter: Frank Lupi                      

Email: lupi@msu.edu             

Abstract:

Choosing a sample is a critical step in valuation survey design. While probability-based sampling is traditionally preferred for representing the general population, non-probability online samples offer faster implementation and lower costs. Non-probability online samples are criticized for potential biases that may not be mitigated by balancing samples to “represent” population demographics (Baker et al. 2013). For stated preference (SP) valuation, a superb review of best practices recommends probability-based sampling but also notes several SP studies with mixed findings (Johnston et al. 2017). More recently, some popular non-probability sample sources have undergone a “crisis” in response quality (Chmielewski and Kucker 2020).

This paper compares SP results between an addressed-based probability sample and two non-probability samples (Qualtrics and MTurk). Our SP survey was developed according to suggestions by Johnston et al. (2017). The probability sample is an address-based sample (ABS) of the general population from the USPS postal delivery file that was implemented in a push-to-web mail-invitation design. The survey received a 23% response rate yielding ~2,200 observations. The two non-probability online opt-in samples include 1,237 respondents from MTurk and 3,095 from Qualtrics. Amazon’s. Topically, the research compares values for freshwater ecosystem services in Michigan using referendum choice questions. We find that while the samples differ in some demographics, such as the relative youth of the MTurk respondents, across 95% of attitudinal variables the samples were substantively very similar. Most importantly though, the samples consistently differed in terms of key economic outcomes like total and marginal WTP for attribute changes.

 

Title: Text Analysis as a complementary approach to identifying protest responses

Authors: Brian Vander Naald, S. Kolstoe    

W4133 obj.: Task 2-1

Presenter: Brian Vander Naald        

Email: brian.vandernaald@drake.edu          

Abstract:

The protection of at-risk species, prevention and mitigation of invasive species, and climate change are all topics that have the potential to trigger strong responses from respondents in stated choice surveys. Some strong responses may actually be biased (e.g., protest or warm glow) or suggest scenario rejection, traditionally all reasons for excluding responses from analysis. The current standard is to use purposeful questions to identify these responses. However, this strategy may be overlooking information from open-ended responses. Open-ended responses have traditionally been examined on a case-by-case basis, and it is unclear whether subjective researcher classification of such responses can be replicated consistently. We contribute to the conversation around classifying open-ended responses by using a machine learning algorithm and exploring how it performs.  

Our data consists of 1030 observations from a 2020 binary choice DCE that elicits preferences for attributes of a hike (including cost, hike availability, invasive species risk, and types of birds encountered) in a unique intact tropical ecosystem. Under assumptions of the RUM, respondents from this representative sample choose the hike with attributes that maximize their utility. From these data, we calculate WTP estimates for the probabilities of encountering different birds, getting on the hike, and introducing an invasive species.

We use a form of text analysis called sentiment analysis to identify and categorize the open-ended responses to the open-ended comments question at the end of the survey as positive, negative or neutral. Using latent class analysis, we explore whether classifying sentiments mechanically using sentiment analysis may be used as a complementary approach to identifying protest votes in a discrete choice experiment. To the extent it helps refine WTP estimates, this approach has the potential to be useful to policymakers.

 

SESSION 2: Water Quality

 

Title:   Spatial dimensions of water quality value in New England River networks 

Authors: Robert J. Johnston, K. Moeltner, S. Peery, T. Ndebele, Z. Yao, S. Crema, W. M. Wollheim, E. Besedin

W4133 obj.: Task 2-2           

Presenter: Robert J. Johnston           

Email: rjohnston@clarku.edu           

Abstract:

To provide insight into how spatial dimensions of water quality changes influence WTP, we integrated a water quality model with a novel map-based, interactive choice experiment architecture to estimate households’ WTP for water quality improvements over a river system. Among our hypotheses is that allowing people to interact with dynamic GIS maps in a choice experiment can provide detailed spatial information that improves the accuracy of WTP prediction, by identifying individualized areas where improvements might have high value to each respondent. We developed the approach to estimate values linked to realistic, predicted scenarios of water quality change flowing from a set of policy actions throughout a large river system. We estimated values for changes over a study domain covering approximately 95,800 miles of rivers and streams in Connecticut, Massachusetts, Rhode Island, Vermont, New Hampshire and Maine.

Each choice-experiment question paired a possible environmental policy scenario with a hypothetically binding household cost, compared to a “business as usual” (BAU) status-quo with no change in household cost. Scenarios illustrated both current conditions and prospective changes in three different water quality measures, representing: (a) safety for human use, (b) support for aquatic life and (c) a multi-metric indicator of overall water pollution. To represent these measures under (1) contemporary conditions, (2) the BAU, and (3) the policy scenario, the survey included nine individual water-quality maps. All maps had dynamic GIS zoom capability, enabling respondents to click a URL to voluntarily view conditions at any desired magnification and area. We developed a novel survey architecture to capture data on how respondents maneuvered through each of these maps prior to answering choice questions. The resulting map-interaction database provided information on all map extents viewed by each respondent, for each water quality map, and the length of time that each frame was viewed. We applied the map-interaction tracking data to infer locations where water quality might be salient to each respondent, based on the time spent viewing each map extent.

The survey was implemented in May 2021 using a mixed-mode, address-based, push-to-web sample. The analysis draws from 1,292 respondents who engaged with maps. A discrete-choice, random-utility model was estimated in WTP-space using Bayesian model search and averaging. Results provide robust evidence that map interactions convey information related to choices and WTP estimates. WTP estimates are influenced by regionwide quality improvements and improvements surrounding each respondent’s home, as anticipated, but also in additional, individualized locations identifiable solely via each respondent’s dynamic map interactions. These spatial effects are pertinent solely for improvements to rivers at low baseline quality and vary over different water quality measures.

These results show that the presented architecture for choice experiments—which captures data on how people interact with maps of water quality change—can reveal individualized areas wherein some types of change have high value, beyond effects related to distance-decay. The model allows these areas, and associated welfare effects, to be identified even if the reasons why these areas are highly valued by each household are not suspected in advance by researchers. Results suggest that spatial dimensions of WTP are more complex than is typically acknowledged in the literature and include components that cannot be identified via standard methods.          

 

Title:   Empirical evaluation of non-point source pollution abatement – Is an improvement detectable?   

Authors: Max Pohlman, K. Zipp      

W4133 obj.:   Task 1-1 & Task 3-2

Presenter: Max Pohlman      

Email: maxpohlman@psu.edu          

Abstract:

Nonpoint source pollution in waterbodies has been a longstanding issue in the United States. These pollutants are defined as detectable, but unable to be traced back to their source. One of the biggest causes of nonpoint source pollution is runoff from agriculture; rainwater flows over the land, picking up excess nutrients as it flows, before winding up in an adjacent body of water. While each source of pollution is likely small, the pollution flows through water networks and coalesces in larger bodies of water.

In recent years, the most common policies to address nonpoint source water pollution subsidize the use of best management practices (BMPS): actions farmers can take to reduce nutrient runoff from their land. However, the evaluation of these policies has relied on predictive modeling, with no empirical studies, to our knowledge, estimating the effectiveness of BMPs on reducing nutrient runoff.

This paper seeks to evaluate whether we can empirically detect the impact of BMPs on water quality. If we can, this could be a tremendous boon towards future policy. If there is no empirical relationship, we must wonder if investing into these kinds of policies is worth it. Can policies be classified as a success even if there’s no detectable change in water quality? If so, how does one define success?

To our knowledge, we are the first to use spatially explicit riparian buffer location data to empirically estimate the impact of a BMP on water pollution at a large scale. We use geospatial techniques to calculate relative watersheds for each monitoring station, defined as the land area that feeds into each station. Over 25 years, we observe the installation of 7,106 buffers in Pennsylvania and connect them with 32,725 nitrogen readings from 1,685 water quality monitoring stations from the USGS and EPA.

We construct a two-way fixed-effects (TWFE) model evaluating the changes in percent feasible buffer area on nitrogen levels. We use month x year and watershed as our temporal and spatial fixed effects, along with various climate, land use, and biological covariates.

Recent research has shown that TWFE estimates often cannot be interpreted as causal due to treatment effect dynamics (de Chaisemartin and D’Haultfoeuille 2020, Goodman-Bacon 2021, Callaway and Sant’Anna 2021). Thus, we also used group-time average treatment effects (GTATE) (Callaway and Sant’Anna 2021) to overcome the shortcomings of the TWFE model.

Our primary result from the TWFE model shows that Pennsylvania’s goal of having 95,000 acres of buffers planted by 2025 (a 50% increase in buffer acreage) would be associated with a 2% further decrease in detected nitrogen. Additionally, our GTATE estimator finds that the addition of buffers caused watersheds that received them to have 11% less nitrogen than watersheds that did not receive buffers. We discuss two major issues with interpreting these results – (1) the potential endogeneity of buffer location and (2) the correlation between buffers and other unobserved conservation practices.  

 

Title: Locally-weighted meta-regression and benefit transfer

Authors: Klaus Moeltner, R. Puri, R. J. Johnston, E. Besedin, J. A. Balukas, A. Le, I. Morin       

W4133 obj.: Task 1-1, Task 2-3 & Task 3-1

Presenter: Klaus Moeltner   

Email: moeltner@vt.edu       

Abstract:

We propose a locally-weighted regression approach to analyze meta-data and generate benefit transfer predictions in an environmental valuation context. We introduce the concept of locally-weighted meta-regression, provide econometric underpinnings, and discuss the construction of weight functions. We illustrate the use of cross-validation to decide between weight functions and show how this framework can be applied in a benefit transfer setting. For our application on willingness-to-pay for water quality improvements we find that the proposed approach brings substantial gains in predictive accuracy in a leave-one-out setting, and vast improvements in predictive efficiency for benefit transfer. These gains are relatively robust to the choice of weight function, suggesting strong potential of this method to generalize to other meta-regression contexts.        

 

Title: Climate benefits of nutrient management                   

Authors: Bryan Parthum, C. Moore, E. Kopits, J. Beaulieu

W4133 obj.: Task 1-1 & Task 1-4

Presenter: Bryan Parthum    

Email: Parthum.Bryan@epa.gov                                          

Abstract:  N/A     

 

SESSION 3: Valuation of Grasslands & Wilderness

 

Title: The use value of grasslands in the Tallgrass Prairie Region of the United States                 

Authors: Kaylee Wells         

W4133 obj.: Task 1-1, Task 1-3 & Task 2-1

Presenter: Kaylee Wells       

Email: kkwells2@illinois.edu           

Abstract:

Much is known about how grasslands (i.e., prairies) provide benefits to society through ecosystem services like carbon sequestration, erosion control, and recreational opportunities (Zhao, Liu and Wu 2020; Chang et al. 2021). Much less is known about the value of these benefits to society because they are largely non-market in nature (Gascoigne et al. 2011). This research is the first to apply the travel cost method to the problem of estimating the value of and demand for grassland recreation. Using data on year-round trips made by Iowa, Illinois, and Minnesota residents to grasslands in the Tallgrass Prairie Region of the United States I estimate trip count models and the use (or recreational) value of these sites. I estimate zero-inflated negative binomial and Poisson hurdle models, which also provide insight into the demographics most likely to visit a grassland for recreation. Grasslands in Iowa, Illinois, and Minnesota are of particular interest because it is estimated that nearly all the native tallgrass prairie in these three states has been converted by agriculture or development (Samson and Knopf 1994). As such, this research could inform the administration of many national public and private programs concerned with preserving, protecting, and restoring grasslands like the USDA’s Conservation Reserve Program and the work of the Nature Conservancy. It could also provide useful information for state-run programs like Minnesota’s Walk-In Access hunting program, which offers additional monetary incentives to landowners with land enrolled in conservation programs who allow public hunting on that land.    

 

Title: Long run evolution of wilderness value: A combined cross-section time series analysis of backcountry hiking             

Authors: Jeffrey Englin, T. Holmes 

W4133 obj.: Task 1-3           

Presenter: Jeffrey Englin      

Email: jeffrey.englin@asu.edu         

Abstract: N/A – presentation canceled         

Title: Public access, ecological integrity, and people’s heterogeneous values of grassland restoration                

Authors: Amy Ando, S. Dissanayake, R. Iovanna, K. Wells           

W4133 obj.: Task 1-1, Task 1-3 & Task 2-1            

Presenter: Amy Ando          

Email: amyando@illinois.edu           

Abstract:

Grassland, or prairie, ecosystems provide many benefits to society including species habitat, carbon sequestration, soil erosion control, and recreational opportunities. At the same time, grassland ecosystems in North America are disappearing, with grassland loss in most areas exceeding 80% since the mid-1800’s; in Illinois the loss is 99.9%.  The USDA protects and restores grasslands through the Conservation Reserve Program. However, it is more costly for farmers to plant high quality grassland habitat on CRP acres, and much remains unknown about public willingness to pay (WTP) for grassland restoration projects and how that varies with grassland quality and public access to restored grasslands. This study quantifies the relationship between the value of grassland restoration and its ecological quality and public access. We use a choice experiment survey of residents in the Tallgrass Prairie region that includes area restored, ecological quality, public access with and without hunting, and annual cost to households as attributes. There were two experimental treatments in survey administration: with and without video (to see if video of the thing being valued changes respondents’ expressed preferences) and with the cost attribute expressed in terms of sales tax or income tax (to see if this affected responses from low-income people who do not pay income tax.) We find that people have large MWTP for acres of grassland restored, and that the premium people are willing to pay for diverse grassland habitat is high. We also find that people have large MWTP to have walking and biking, while the average MWTP to have hunting also available is negative. Findings yield interesting insights into how preferences for open space and recreation vary among different groups of people.     

            

SESSION 4: Land Conservation

 

Title: Does temporary land retirement promote organic adoption? Evidence from expiring Conservation Reserve Program contracts                

Authors: Daniel Bigelow, H. Wing, K. Fuller          

W4133 obj.: Task 1-1           

Presenter: Daniel Bigelow   

Email: daniel.bigelow@montana.edu           

Abstract:

The Conservation Reserve Program (CRP) was established in 1985 in an effort to reduce agricultural production on highly erodible and otherwise environmentally-sensitive farmland. Through the CRP, producers can voluntarily retire land from production for a contract period of 10-15 years in exchange for a yearly rental payment. At its 2007 peak, CRP enrollment exceeded 35 million acres but has since been steadily declining. Over the same period, organic agriculture has seen tremendous revenue growth and is increasingly promoted as a potential solution to various environmental problems associated with conventional agricultural production.

Despite its appeal, there remain important barriers that may dissuade conventional producers from adopting organic practices. Chief among these is the three-year transition period that prospective organic producers must get through before their products can be marketed under the USDA organic certification label. During this three-year transition period, producers incur the lower yields associated with organic production without receiving the price premium typically associated with organic products. The CRP provides a unique opportunity for producers to forgo this transition process. CRP land is, by definition, not being used for production and therefore incidentally complies with organic transition standards, allowing it to be certified in the year that it exits the program.

In this paper, we study the extent to which CRP enrollment promotes uptake of organic production practices. Primary econometric results come from a set of finite distributed-lag models, which identify the effect of the number of expiring CRP contracts on the number of new certified organic operations, conditional on county fixed effects, time fixed effects, state-specific linear trends, and various county controls. Our analysis is based on a unique county-level panel dataset of organic certifications from USDA’s Organic Integrity database combined with contract-level CRP information from the USDA’s Conservation Contract Maintenance System. We exploit the 10-15-year gap between CRP enrollment and contract expiration as part of our identification strategy, which rests on the assumption that producers did not enroll in the CRP with the specific intention of converting their land to organic production upon expiration of the contract.

By examining the connection between the CRP and organic agriculture, our paper provides new evidence on the fate of land in expiring CRP contracts and furthers our understanding of the importance of the transition process in organic conversion decisions. Preliminary results indicate that CRP contract expiration leads to increases in organic adoption, an effect that is most prominent in the southern region of the US. These results suggest that easing the transition period constraints faced by landowners can promote greater uptake of organic production practices, highlighting an important potential co-benefit of the CRP.  

 

Title: The distributional impacts of land conservation         

Authors: Corey Lang, J. VanCeylon, A. Ando        

W4133 obj.: Task 2.1

Presenter: Corey Lang         

Email: clang@uri.edu           

Abstract:

Wealth inequality and environmental justice issues have become more prevalent in recent years and are important policy considerations. Evidence for inequitable environmental outcomes is well-documented in research where poorer and more diverse communities are more likely to be exposed to environmental hazards like poor air quality or superfund sites compared to wealthier and whiter neighborhoods (Kahn 2002, Currie et al. 2011, Lang and Cavanagh 2018, Banzhaf et al. 2019). While environmental justice priorities are currently more present in government decision-making processes, these policies and efforts do not always lead to socially equitable outcomes.

The purpose of this paper is to understand how the benefits accrued from land conservation are distributed among households. When land is conserved, this creates a place-based amenity that is capitalized into the housing market. Nearby homeowners benefit from this creation or transfer of wealth with little or no expense. Because neither home purchase location nor land conservation decisions are random, the wealth generated is unlikely to be benefit socioeconomic groups equally.  

 
Title: Observed and unobserved heterogeneity in conservation decisions: Implications for the voluntary supply of agricultural greenhouse gas abatement                 

Authors: Dale Manning, M. Rouhi Rad, D. Mansfield, S. Ogle      

W4133 obj.: Task 1-1, Task 2-1 & Task 3-1

Presenter: Dale Manning     

Email: Dale.Manning@colostate.edu           

Abstract:

Agriculture has the physical potential to contribute to global greenhouse gas mitigation efforts.  Current policy initiatives focus on voluntary efforts to influence agriculture’s impact on the climate.  Despite this, little is known about the economic viability of agricultural greenhouse gas abatement at a large scale.  Therefore, we leverage novel panel data on crop and practice choice, soil carbon, and emissions over time on more than 15,000 fields across the Cornbelt Region of the United States to estimate the quantity of agricultural greenhouse gas abatement as a function of the price paid for changes in net carbon emissions.  To do this, we develop a discrete choice modeling framework in which producers choose a crop (e.g., corn, soy, etc.) and practice (full-till, no-till, cover crop, etc.) from the set of available crops and practices observed in the region.  Importantly, we observe a rich set of farm and field characteristics that drive the relative profitability of alternative crops and practices.  After estimating the parameters of the choice model under alternative assumptions about producer heterogeneity in response to net profit, we use DayCent model simulations to estimate field-level changes in soil carbon and net emissions over a 15-year period under all feasible crop and practice choices.  Finally, we sample from predicted probabilities to estimate baseline crop and practice choices and compare those to choices under a range of carbon prices up to $100 per tonne.

Through this work, we make two contributions to the agricultural and resource economics literature.  First, given existing evidence on heterogeneity in producer conservation attitudes, we explore the importance of modeling assumptions regarding the heterogeneity in response to conservation incentives.  Specifically, we estimate choice models using conditional logit, mixed logit with random parameters describing the response to net revenue, latent class models, and a combination of latent class and random parameters.  Using a subset of the data, we estimate how modeling assumptions affect the estimated abatement supply curve.  After identifying the model specification that allows for sufficient heterogeneity while minimizing computational time, we apply the model to our study area.  This leads to our second contribution: we provide an empirically relevant estimate of the cost of incentivizing agricultural greenhouse gas abatement at scale.  This can inform current and future policy efforts to encourage greenhouse gas abatement on agricultural land in the US.

This work will benefit greatly from input from W4133 members.  We hope to learn from the group’s experience with choice models commonly used in valuation exercises.       

 

SESSION 5: Stated Preference Applications

 

Title: Households’ Willingness to Pay for Different Attributes of Alternative Sources of Residential Electricity: Evidence from a Large Discrete Choice Experiment     

Authors: Dilek Uz, J. Mamkhezri     

W4133 obj.: Task 1-4           

Presenter: Dilek Uz  

Email: dilekuz@unr.edu       

Abstract:

Consistent with the dramatic decreases in costs, economic incentives offered at various governmental levels, and rising social awareness of the environmental issues, the US residential

rooftop solar capacity grew rapidly in the past decade. As of December 2011, total installed

residential solar capacity was about 0.8 GW spread over 159 thousand customers while this

number reached 16.5 GW with more than 2.5 million customers in December 2020.

There are multiple avenues through which residential solar is supported within the US

energy and environment policy such as tax credits and rebates both at the federal and state

level. Perhaps the most important and debated mechanism is net energy metering (NEM)

which is an arrangement between a power utility company and its customers where the

customer buys or rents the equipment to generate electricity onsite for own consumption   and sell any unused portion to the grid. Generating electricity in such decentralized manner

with relatively small-scale equipment is referred to as distributed generation (DG). First net

metering law in US was enacted in Minnesota in 1983 (Wan and Green [1998]). More than

two decades of relatively slow uptake is followed by a rapid growth in net metering both in

terms of customers and capacity as a response to decreasing costs and financially attractive

net metering terms. However, as the capacity skyrocketed, economic feasibility and fairness

concerns (more details below) grew proportionately, leading to a reversal in the state policies

regarding the lucrativeness of the net metering agreements. This reversal is the main source

of ongoing controversy around rooftop solar in the US.

The support mechanisms for residential solar are promoted on the grounds that they provide public benefits such as avoided generation from fossil fuels and reduction in transmission losses via bringing together the point of generation and consumption. The proponents assert that these supports are needed to incentivize private investments in renewable energy and diversify the energy sources while helping the local economy and the environment.3

On the other hand, economists raise concerns that NEM causes unfair cost shifting unto

non-solar households who, on average, tend to be on the lower end of the income distribution

compared to solar households (Inzunza and Knittel [2020]). Furthermore, studies showed

that public benefits of residential solar are highly location specific and overall, not necessarily

high enough to justify the costs of supporting the NEM programs (Brown and Bunyan [2014];

Cohen and Callaway [2016]). Consistent with this, compensation scheme - how much the

NEM customer should be compensated for their exports to the grid, has been the subject of

ongoing policy debates in many states. As of October 2016, 44 states (plus the District of

Columbia) implemented mandatory NEM policies with varying rules whereas this number

has retracted to 37 states (plus District of Columbia) as of August 2021. In the third quarter of 2021 alone, 27 states took a total of 55 actions related to DG compensation rules

(Proudlove et al. [2021]). In December 2021, the public utilities commission of California– the

state with the largest installed residential solar capacity in the US, has issued a proposed

decision to cut back the compensation that the net metering customers receive for their

exports into the grid.

The usual scenario for a rooftop solar panel sale involves a customer soliciting information from one or more vendors. Afterwards, a member of the sales team meets with individual customers during which they deliver their sales pitch. It usually includes some information about financial advantages of owning solar panels which are derived using some strong underlying assumptions about future energy prices. Additionally, the sales representative often makes remarks about the importance of being independent from the utility company as well

as the environmental benefits associated with this investment.

As mentioned above, in addition to the public benefits, rooftop solar panels offer a

variety of private benefits to their owners. These include financial benefits such as potential

reduction in energy bills and appreciation for home value as well as subjective or psychological

benefits like “warm glow”7, the ability to signal environmental values, and the feeling of

being more self-dependent. Understanding how much US households value these private

benefits is vital for rooftop solar producers who need to be able to develop effective marketing

strategies to expand their customer base beyond the early adopters in a policy environment

where support for NEM has the potential to gradually weaken. The findings could also

inform the policy makers, who are tasked with formulating compensation schemes while

balancing fairness and meeting the demands of highly vocal residential solar proponents.

In an ideal setting to investigate how much consumers value different attributes of various home energy options, the researchers would observe the households making actual investment decisions while facing different alternatives with varying attributes. However, unfortunately, such a dataset does not exist. Furthermore, the existing records on past household meetings with solar customer representative and the choices of potential customers are proprietary to the rooftop solar vendors. The next best alternative is to simulate the choice settings with consumers. With this in mind, we develop and implement a large stated-preference

discrete choice experiment (N = 1033) in order to disentangle the willingness to pay (WTP)

for distinct private benefit sources of owning rooftop solar panels. Participants, who are

selected from the demographic groups that are more likely to purchase solar panels based

on existing demographic data on rooftop solar owners, are asked to choose from different

energy options with varying payment schedules and other attributes (explained in more

detail below). The study also features a randomized control trial (RCT) component where

a randomly selected subset of the participants (treatment group) are presented with the net

present value (NPV) of each of the options together with the payment schedule which is

composed of the down-payment and net monthly out-of-pocket costs. The control group, on

the other hand, is only presented with the payment schedule without the NPV information.

This study contributes to the current body of knowledge on WTP for renewable energy

and household investment decisions in various ways. To the best of our knowledge, this is

the first study that i) differentiates the WTP for distinct attributes of a renewable energy

source, ii) investigates the importance of information provision in investment decisions within

the renewable energy setting, and iii) incorporates decision makers’ political affiliations as

a control variable. Furthermore, we expand the current WTP methodology by presenting

individual-level WTPs coupled with their confidence intervals. In order to do this, we first estimate a random coefficients (mixed) logit model using the Hierarchical Bayes (HB) method.

We then construct and provide visualisations for the 95% confidence intervals of the WTP

for various attributes by taking the 2.5th and 97.5th percentiles of the set of estimates created

using the parameter draws that the HB method yields. For this exercise, we implement the

“individual parameters” approach explained in Revelt and Train [1998] in order to conduct

an individualized refining to the coefficient draws obtained from the parameter draws.

Our study leads to a number of important findings. First, we show that there is a

considerable amount of heterogeneity in consumer WTPs for different attributes where fixed

coefficients logit fails to capture. This is evident not only in the significant standard deviation

estimates inherent in the mixed logit estimations but also in the individual level confidence

intervals for the WTP estimates we construct. Second, when presented with the NPV

information for the options they are facing, the WTPs that the consumers demonstrate for

the attribute we are studying are up to an order of magnitude less than the case where this

information is suppressed. This finding has important implications for the future of rooftop

solar in the US where the current policy environment around supporting NEM is highly

controversial. In other words, without the financial advantages of NEM policies, rooftop

solar industry may face serious challenges expanding their customer base especially if their

marketing pitch solely depends on selling “energy independence” to households. Another

important insight to be gained here is that our results raise potential concerns regarding the

reliability of WTP figures estimated in investment related discrete choice settings where the

NPV information is not provided.

 

Title: Willingness to Pay for Green Energy, Green Jobs and Green Views: A Choice Experiment on Offshore Wind Energy              

Authors: Lee Parton, S. Lutzeyer, D. J. Phaneuf, L. O. Taylor        

W4133 obj.: Task 2-1           

Presenter: Lee Parton           

Email: leeparton@boisestate.edu       

Abstract:  

We conduct a choice-experiment estimating the changes in electricity prices which North Carolina residents would be willing to face to have wind energy augment the grid, and how willingness to pay for green energy varies depending on the visual, social and environmental impacts associated with offshore wind development. Respondents were presented with offshore wind farm scenarios at different distances from the coast and information regarding the emissions reductions and job creation that accompany different development scenarios. We explore the extent to which job creation and carbon reductions can offset visual disamenities and differences in prefences across coastal and inland residents. We find preference heterogeneity dependent on the siting location, i.e., whether the wind farm is visible from a hypothetical resort town or along the Cape Hatteras National Seashore. Using a latent class model, we find a class of respondents indicating positive willingness to pay (WTP) for wind farm development visible from a tourist town and negative WTP for wind farms visible from the Cape Hatteras National Seashore. Conversely, we find a class of respondents who prefer a hypothetical wind farm be constructed along the National Seashore as opposed to being visible from a hypothetical tourist town.  

Title: Comparing Money vs Labor Payments in Individual and Group Settings for Choice Experiments in Developing Country Settings             

Authors: Sahan Dissanayake, S. Vidanage  

W4133 obj.: Task 1-1 & Task 2-1    

Presenter: Sahan Dissanayake          

Email: sdissan2@gmail.com 

Abstract:

The use of stated preference methods in developing countries is growing with the increasing concern about the environment as economies develop. At the same time using monetary payment vehicles and estimating a Willingess to Pay (WTP) can be problematic in rural or low incomes areas in developing countries. Many respondents in these areas regularly engage in barter and paying with labor and do not use monetary paymetns for all transactions. This distinction from urban areas with a monetary economy and with most settings in developed countries can impact results from valuation studies as the WTP elicited from rural and low-income areas is likely to be low even though respondents may have a high value and be willing to pay through other means. In response to these concerns, a growing number of stated preference studies explore using both monetary and non-monetary payment options. We contribute to this literature by exploring how the use of monetary vs labor payment options can impact values eclitied from choice experiment studies coundcuted in rural developing country settings. Our application is a choice experiment survey to value restoring an ancient irrigation system known as cascading tank systems in Sri Lankan. The cascading tank systems are designed to complement the surrounding landscape and has parallels to the Japanese land use system of Satoyama. In Sri Lanka, these irrigations systems were created over 1500-2000 years ago but are still functioning today and provide irrigation for nearly 40% of the total irrigable area of the country. At the same time these ancient systems are degrading and there a large number of efforts to restore these systems which the FAO recently identified as a Globally Important Agricultural Heritage Systems (GIAHS). We conduct a choice experiment to understand the WTP/willingesss-to-contribute of rural households to restore these irrigiation systems. We also contribute to the literature on the applications of choice experiments in developing country settings by comparing data gathering methods, specifically individual surveys vs group information session. We find that in the individual survey settings respondents are more willing to contribute labor (compared to an equivalent monetary payment). Based on the early results we find that there is no difference between the group and individual survey settings when the survey is presented as a monetary payment but for the labor payment treatment, the group setting results in a positive payment coefficient for the labor payment attribute (i.e., respondents are willing to contribute labor when the survey is conducted in a group setting). We are currently exploring the heterogeneity of these results across respondents and plan to include these new results in the presentation as well.

 

Title: Household Preference for Impure Public Goods - an Application of Community Gardens  

Authors: Liqing Li, D. Long

W4133 obj.: Task 2-1

Presenter: Liqing Li 

Email: liqing@Fullerton.edu

Abstract:

Urban agriculture (UA) has grown as a planning priority in many cities in the U.S. As an essential form of UA, community gardens create a space for residents to grow plants. While they share similarities with other types of urban greenspace by providing ecosystem services as public goods. individuals who are often motivated to participate in garden management in hope of obtaining the private consumption benefits (i.e., harvested vegetables and fruits), which in turn contributes to the provision of the public goods such as biodiversity. Community gardens are thus impure public goods.

While many studies have examined the benefits of urban agriculture and other types of urban greenspace like parks, few have focused on community gardens. Even though community gardens are seen as a tool addressing social equity issues like food insecurity among low-income communities and unbalanced distributions of urban green spaces (Siegner et al., 2018), preferences for local and healthy food are typically associated with highly educated and wealthy communities (Bellemare and Dusoruth, 2020). It is therefore unclear whether there is a demand for such food among low-income households whom the gardens intend to benefit. More importantly, it is critical to evaluate whether public goods co-benefits encourage individual’s willingness to contribute to community gardens. We fill the gap in the literature by estimating individuals' preferences for community gardens with both private and public characteristics. Specifically, we explore the heterogeneity in their monetary willingness to pay (WTP) and non-monetary willingness to volunteer (WTV).

We will conduct a discrete choice experiment (DCE) to estimate households’ WTP and WTV for community garden in Los Angeles, California. The status quo scenario is a piece of vacant land with no community garden. The respondents are presented with a total of six discrete-choice questions. Each choice question includes five varying attributes with the corresponding attribute levels: monetary cost, time cost, ecosystem services, food allocation, and garden scale.

To quantify the individual level valuation for community gardens while accounting for unobserved heterogeneity, we estimate a mixed multinomial logit (MMNL) model. To explore heterogeneity in households’ preferences, we include interaction terms of factors such as neighborhood diversity with choice-specific attributes in the MMNL. Second, to illustrate how preferences vary across different neighborhoods, we recover the conditional individual-specific means of marginal WTP and WTV for every respondent and then calculate the zip code level valuation for each attribute.

Taken together, our analysis examines benefits offered by community gardens and factors affecting the provision of impure public goods with both public and private goods component. Our study contributes to several strands of literature. First, we offer insights on whether and how public co-benefits affect people’s valuation of impure environmental public goods (MacKerron, 2009). Second, we contribute to the urban agriculture and greenspace literature by investigating the heterogeneity in households’ preferences such as how diversity or social cohesion affects their evaluation (Printezis & Grebitus, 2020). Lastly, our study complements a recent yet rapidly developing literature on measuring valuations of environmental assets using non-pecuniary payment modes like time (Ando et al., 2020).            

 

SESSION 6: Valuation of Coastal Resources

 

Title: The Effect of Managed Retreat on Coastal Property Markets: Evidence from the NY Rising Buyout and Acquisition Program 

Authors: Steven Dundas, Y. Hashida           

W4133 obj.: Task 1-2 & Task 1-4    

Presenter: Steven Dundas    

Email: steven.dundas@oregonstate.edu      

Abstract:  

Understanding the economic impacts to a community of removing housing and infrastructure from hazard-prone locations is an open empirical question. Using data from a buyout and acquisition program in the U.S. state of New York and residential property transactions between 1995 and 2020, we recover hedonic estimates of the property value impacts of government-acquired properties in a coastal hazard area. Our identification strategy exploits the spatial proximities between sales and bought-out houses and the timing of different stages of the buyout process - closing, demolition (to be converted to open space), auction, and redevelopment. Our results suggest that buyouts have a large negative effect on housing prices for adjacent houses, with that effect becoming smaller at the block and neighborhood scale and then dissipating completely beyond 800 m. These findings are consistent between different difference-in-differences specifications, including two-way fixed effects, an event study, an event study that addresses concerns of treatment effect heterogeneity and variation in treatment timing, and a triple difference model. The event study specifications suggest housing market impacts likely dissipate after four to seven years. We also find heterogeneity in impacts across space and at different stages of the buyout process. Our findings are relevant for policy by demonstrating the housing market implications of different buyout methods, the role of information in the buyout process, and by highlighting potential avenues to minimize negative impacts of managed retreat. 

 

Title: Preferences for Buyouts and Rentbacks on the East Coast                

Authors: Anna Perry, C. Landry      

W4133 obj.: Task 1-2 Task 1-4 & Task 3-2 

Presenter: Anna Perry          

Email: anna.perry1@uga.edu

Abstract:

We use survey data to assess coastal residents’ willingness to accept (WTA) a property buyout and willingness to pay (WTP) to rentback their property from the state government after a buyout (obviating financial risk of property loss). Placing risk of property loss in the expected utility framework, we implicitly define WTA and WTP as a function of flood experience, risk perception, risk preference, affect (worry), income, wealth, and other factors. We utilize univariate and bivariate Tobit and Cragg models to analyze the data. We find WTA/sqft is increasing in income and a wealth proxy (likely reflecting housing quality), while risk perceptions have no significant effect (potentially due to endogeneity). WTP/sqft is increasing in expected hurricane damages and worry, but not influenced by income or wealth. We present a foundation for structural estimation and explore policy implications.         

Title: The Impact of Air Pollution on Outdoor Recreation: An Application to Shoreline Fishing Along the Atlantic and Gulf Coast         

Authors: Roger von Haefen, Y. Liu 

W4133 obj.: Task 1-1, Task 1-3 & Task 2-1

Presenter: Roger von Haefen           

Email: rhhaefen@ncsu.edu   

Abstract:

            A large travel cost literature investigates how improvements in water quality impact recreation demand, but relatively few studies investigate how improvements in air quality impact recreation demand. This paper aims to fill this gap by exploiting remotely sensed PM2.5 data for the entire East and Gulf Coast regions and 15 years of shoreline fishing data from the Marine Recreational Information Program (MRIP). Using a two-stage linked modeling framework (Parsons and Kealy, 1995), we first estimate random utility maximization (RUM) models of site choice separately by year and two-month wave, and then estimate a Poisson participation model with fixed effects by year, wave, and origin zip code. Our preliminary results suggest that air pollution has a negative and statistically significant impact on recreation demand, and in addition to refining our air pollution data, our next steps include constructing welfare measures for several policy scenarios.         

                     

SESSION 7: Recreation

 

Title: Using cell phone data in discrete choice recreation demand models: A case study of Rocky Mountain National Park            

Authors: Jude Bayham, R. Schmidt 

W4133 obj.: Task 2-1

Presenter: Jude Bayham       

Email: Jude.Bayham@colostate.edu

Abstract:

Nonmarket valuation has been a cornerstone of environmental economics for decades. Data collection generally requires the development and administration of surveys that cost both time and money. Over the past five years, technology has enabled the passive collection of

high-resolution mobile device geolocation data. Many studies have used mobile device location

data in research related to the COVID-19 pandemic, and some have used data to document

trends in recreation activity. However, none (to our knowledge) have used individual-level

device data to estimate discrete-choice recreation demand models.

We construct a dataset to estimate the welfare impacts of closing Rocky Mountain National Park in Colorado for several months during the early stage of the COVID-19 pandemic in the Spring of 2020. Specifically, we construct visitation diaries of over 2000 devices believed to reside in three counties near RMNP to 41 recreation sites in Northern Colorado. We use google maps to estimate travel time and distance from the devices common evening location (home) to all recreation sites. We estimate a Random Utility Model using data in the summer months following RMNP’s reopening. We use the model to simulate a closure of RMNP and quantify the redistribution of visits to substitute recreation sites and measure the welfare impacts of the closure. We compare the results of the RUM to a single site travel cost model to highlight the importance of capturing substitute sites in measuring the impacts of a site closure.

We find that people did substitute other recreation sites during RMNP closure. The results of the RUM support the intuition that accounting for substitution reduces the welfare cost of a site closure. While the results are interesting and have policy implications for large national park closures, the contribution of the research is to demonstrate how this emerging data source can be used to estimate welfare impacts for almost any recreation site in the US. The presentation will also cover the limitation of the data including the substantial data processing costs.        

 

Title: COVID-19 mandates and recreation in Idaho            

Authors: Sophie Croome      

W4133 obj.: Task 1-3           

Presenter: Sophie Croome   

Email: sophiecroome@boisestate.edu          

Abstract:

Significant survey data exists for national and state park visitation and several studies have measured consumer preferences and visitation patterns (Kupfer, 2021; Volenec, 2021; Wood, 2013; Yan, 2021). However, municipal park visitation remains largely understudied due to the difficulty and costliness associated with data collection and analysis. This study utilizes high frequency mobile device location data to measure changes in municipal and state park visitation caused by COVID-19 response policies. We exploit spatial and temporal variation in COVID-19 mandates at the county level in the U.S. state of Idaho and at the state level in the United States to identify the causal effect of mandates on park visitation. The research finds that people were more likely to recreate in, and come from, areas with less restrictions. Visitation rates were about seven percent lower in areas with a mask mandate than would be expected if no policies were in place. Estimates of visitation patterns based on visitors’ origin states indicate that of the people who recreate in Idaho, a state with limited COVID-19 response, the visitation rate was about 21 percent less for states with mask mandates than that of a state without a mask mandate. This research brings insight on behavioral response to restrictions and on recreational choice behavior.            

Title: Estimating changes in value flows for recreational resources: evidence from the COVID-19 pandemic                       

Authors: Matt Sloggy, D. Long, J. S nchez

W4133 obj.: Task 1-3

Presenter: Matt Sloggy        

Email: matthew.sloggy@usda.gov   

Abstract:

Several studies examining recreation patterns during the COVID-19 pandemic have found substantial drops in visitation to parks (e.g., Landry et al. 2021) but booms in visitation to

forests (Derks et al., 2020). However, few have examined other kinds of recreational activities,

namely visiting public wilderness areas where social distancing is much easier. Increases in

visitation to public lands may raise costs for agencies, and insufficient funding can degrade

ecosystem services due to overuse and deferred maintenance. It is thus important to examine if

current trends will continue following the pandemic, which provides critical information

needed to coordinate the proper amount of investment. Additionally, recreation is an important

economic activity, stimulating purchases in a variety of industries. This study fills the gap in the

literature by analyzing how the COVID-19 pandemic has changed visitation, group size, and

time spent recreating in two wilderness areas, Desolation Wilderness (DW) and the Inyo

National Forest (INF), located in California.

We create a unique dataset by combining visitation permit data with data on smoke conditions and weather. The final dataset contains a total of 49,843 permits across both wilderness areas. The average group size remains constant across the two years (2.9 persons per permit). The total visitation in our sample is 61,865 in 2019 and 81,621 in 2020. We also obtain weather precipitation and temperature data from the Parameter-elevation Regressions on Independent Slopes Model dataset, income and population data from the American Community Survey, and smoke cover data from NOAA.

We estimate a travel cost model in which the slope of the demand curve is a function of the pandemic. The estimation results show that pandemic has increased visitation on average, and this increase is primarily driven by increases in visitation to the INF. Furthermore, we find that the slope of the demand curve for recreation is a function of whether the pandemic is occurring or not (whether it is later than March 13th, 2020). This indicates that the pandemic limited the choice set of recreators, leading to an increase in the value of sites that are accessible. We also find the demand curve of visitation to the wilderness areas became less elastic. This further implies that the value of visitation increased during the pandemic.

Though past research has examined visitor permit data, this is the first that analyzes the impacts of the COVID-19 pandemic on recreation in wilderness areas. We make several contributions to the literature. We extend work that finds increases in visitation to forests by estimating the extent to which the pandemic changed the value of the resource. We nuance similar research demonstrating decreases in recreation by highlighting the importance of the variety of recreation. Additionally, we contribute to the recreation literature by further demonstrating how the use and value of recreational resources change during times of crisis, that is, a pandemic. Taken together, these results provide useful information that helps forest managers implement efficient strategies to maintain access to the public during unexpected events.      

 

Title: Seasonal variation in recreational value by activity in a state and national scenic river corridor – The case of the Little Miami                     

Authors: Brent Sohngen, A. Longstreth, C. Barnett, H. Taft, B. Schiemann          

W4133 obj.: Task 1-3

Presenter: Brent Sohngen    

Email: sohngen.1@osu.edu  

Abstract:

This study develops estimates of the value of recreational activities in the Little Miami River corridor, a state and national scenic river, in Southwestern Ohio.  The corridor includes a multi-use pathway that forms part of the Ohio river to Lake Erie recreational trail, as well as many local and state parks that provide access to the river and its largely intact riparian corridor. The river corridor flows to the east of the Cincinnati-Dayton Metropolitan area, a heavily populated and industrialized region in the state of Ohio. State parks and state nature preserves provide unique outdoor recreational activities, as do locally managed parks. Despite long-held policy interest in protecting the river and the corridor, little is known about the number of people who recreate there annually or the value they provide.  Working with a set of over 20 volunteers, we designed a stratified random sampling procedure to determine the number of visitors to 51 locations along a 50 mile stretch of the corridor. Sampling at the sites occurred form April 1, 2021 to October 31, 2021, and involved counting cars and leaving a flyer on each car requesting individuals to respond to a short survey.  The data suggest that there were over 750,000 visitors to the 51 locations over the 7-month survey period.  Using a zonal travel cost model, we estimate average consumer surplus of $12.03 per trip.  We explore variation in consumer surplus across the seasons and across various activities, including biking, hiking, paddling, and fishing.             

 

SESSION 8: Measurement Error and Environmental Exposure

 

Title: Remotely Incorrect? Accounting for Nonclassical Measurement Error in Satellite Data on Deforestation              

Authors: Jennifer Alix-Garcia, D. Millimet 

W4133 obj.: Task 1-1

Presenter: Jennifer Alix-Garcia        

Email: jennifer.alix-garcia@oregonstate.edu.           

Abstract:

Historically, data on land use have been obtained from ground surveys. However, this has changed over time. There are now over 2,000 satellites orbiting Earth. In combination with the growth in computer processing power, this has driven an explosion in the availability of data - on land use as well as other environmental attributes - derived from remote sensing methods. Remotely sensed data offers numerous advantages. First, satellite data can provide global coverage. Second, whereas survey data suffers from many sources of error (enumeration errors, response biases), these are necessarily eliminated through satellite collection. However, absence of familiar types of data error does not imply the absence of all error. Systematic mismeasurement in remotely sensed measurements is likely present. This paper details the nature of these errors, discusses the econometric implications, and identifies an estimator that could help ameliorate the problem.

To begin, we describe the process by which satellite data moves from sensors to usable data and review the econometric implications of misclassification of the dependent variable on several binary choice estimators common in the land use literature. We then discuss alternative estimators, representing extensions of the misclassification binary choice model proposed in Hausman et al. (1998). In particular, we consider two extensions. First, we allow for the misclassification rates to depend on covariates (Lewbel 2000). Here, the covariates capture environmental attributes affecting the accuracy of satellite classifications. Second, we use the scobit family of binary choice models, which nests the logit model as a special case (Nagler 1994). The scobit introduces an additional shape parameter into the link function. This additional flexibility has proven useful when the outcome is of the rare-events type (Golet, 2014), which is the case for deforestation.

We provide evidence of nonclassical measurement error in both binary and continuous measures of forest cover using two satellite-based measures of forest cover for Mexico near the same time period and based upon imagery from the same type of sensor. Correlations between the differences in two measures and environmental (slope, elevation, biome) and sensor attributes are statistically and economically significant, consistent with the data suffering from nonclassical measurement error.

To investigate the practical performance of the estimators considered we examine the impact of a program of payments for ecosystem services on deforestation in Mexico over the period 2003-2015. Prior to analyzing the data, we undertake a Monte Carlo study designed to mimic the panel data. The simulations lead to three primary conclusions. First, ignoring misclassification introduces significant bias. Second, current approaches in the literature designed to deal with misclassification do not eliminate bias. Third, our extensions of the Hausman et al. (1998) estimator perform quite well.

In our application, we also obtain three main findings. First, the satellite-based measure that we use under-reports the true extent of deforestation. Topography and the availability of images are important determinants of misclassification. Second, ignoring misclassification results in bias of the average marginal effects, although the bias need not always be toward zero. In our application the estimated treatment effect is reasonably estimated, but the average marginal effects are significantly attenuated for the other covariates in the model, especially those that are determinants of misclassification.         

 

Title: In Utero Environmental Exposures and Birth Outcomes: Evidence from Large-Scale Animal Feeding Operations                 

Authors: Laura O. Taylor, N. Abashidze, J. Hochard          

W4133 obj.: Task 1-1

Presenter: Laura O. Taylor   

Email: laura.taylor@gatech.edu       

Abstract:

We estimate the causal effects of hog farm facilities on birth outcomes in surrounding communities using rich, spatially explicit, individual-level data that links infant births to detailed geographic information on hog farm operations.  The contribution that in utero exposure to pollution has on birth outcomes is important for policy makers, medical professionals, and families to understand as they work to reduce the incidence of preterm births and other negative birth outcomes. A small number of recent papers have used observational data and credible research designs to find negative effects of in-utero exposure to pollution generated by industrial activities on birth outcomes.  Our research extends this work to an important source of community exposure in rural areas: large-scale animal feeding operations (AFOs).  With more than 450,000 AFOs in the U.S., the potential impact of AFOs on birth outcomes could be significant and more heavily borne by low-income or minority populations that can be correlated spatially with AFOs. 

Our research focuses on birth outcomes from mothers living near pork producing AFOs. Hog farming is a $22.5 billion industry in the U.S. with more than 60,000 farms raising over 72 million hogs in 2017.  Most hogs in the country are raised in AFOs, where large numbers of animals, their waste, and other byproducts are concentrated over a small land area.  AFOs have generated public concerns regarding odor and potential health effects due to airborne particulate matter and toxic gases present in hog waste storage lagoons (e.g., ammonia, methane, hydrogen sulfide, and carbon monoxide), especially when their contents are sprayed seasonally on neighboring fields as part of waste management plans.  The causal effects of hog AFOs on public health outcomes are not well understood despite the scale of the industry and conflict between AFO operators and neighboring communities.

We exploit exogenous variation in daily wind direction during a woman’s pregnancy relative to prevailing winds to identify the effects of hog farm exposure on birth outcomes. Evidence suggests that in-utero exposures to hog farm facilities reduce birth weight, shorten gestation periods and increase the likelihood of preterm birth. These airborne exposures are concentrated within three miles of these facilities and are driven by the season with greatest sprayfield activity. Taken together, these findings shape a landscape of birth outcomes that reveal “hotspots” of vulnerable populations where the social costs of environmental exposures are highest.              

 

Title: Addressing measurement error from movers in a travel cost analysis with revealed and stated preference data          

Authors: Sonja Kolstoe, K. Poisal    

W4133 obj.: Task 1-3 & Task 2-1

Presenter: Sonja Kolstoe      

Email: sonja.kolstoe@usda.gov        

Abstract:

To conduct a travel cost study, information is needed on the respondent’s origin and recreation destination. Travel cost models rely on travel cost being measured with minimal measurement error to avoid issues of internal validity of estimates of this parameter. To date, to the best of our knowledge, little attention has been paid to measurement error in the respondent’s origin if they moved during the study period. Moving potentially changes the overall distance to the recreation site, introducing the possibility of measurement error in the absence of data about when and where the respondent moved. To address this problem in count models, we propose using a weighted travel cost measure that accounts for when respondents moved, i.e., movers, as well as when they took trips. We incorporate this correction into a travel cost model combining revealed and stated preference data on coastal recreation use in Delaware and use it to value changes in sea-level rise accounting for respondents moving in our sample. We use responses from an online survey on zonal trip behavior for 2019 and 2020 as well as contingent behavior data on the hypothetical reduction in the size of coastal public lands because associated with sea-level rise. We use a zonal approach to model trip frequency using actual and stated data from 2019 and 2020 under three different area reduction scenarios of 25%, 50% and 75% due increasing sea-level rise. We estimate per-trip welfare estimates and well as calculate changes in welfare due to differences in behavior based on the contingent behavior scenarios. On average movers in our sample move closer to Delaware coastal public lands; non-Delaware residents moved about 64 miles closer on average and Delaware residents moved about 28 miles closer on average. We find evidence of a statistical difference in consumer surplus between the unweighted and weighted travel cost estimates for non-Delaware residents, but not for Delaware residents. These results suggest that it may be important to gather information on if and when respondents move within the study time and study area in order to have the data available to minimize measurement error in travel cost. Future studies may want to consider asking questions about moving and seasonal visitation behavior when asking for past year’s trips to adjust the travel cost for this change in origin.