1. A comparative economic assessment of dredge- and diversion-based land building CPRA Board April 16, 2014  Baton Rouge, LA Rex H. Caffey 1, Hua Wang 2, Daniel Petrolia 3 1 Professor /Director and 2 Graduate Research Assistant Center for Natural Resource Economics & Policy LSU AgCenter and Louisiana Sea Grant 3 Associate Professor, Center for Natural Resource Economics & Policy Dept. of Ag. Economics, Mississippi State University

2. “Economics” is the study of how limited or scarce resources are allocated amongst competing needs.

3. Benefit-Cost Analysis vs. Cost Efficacy Total Benefits ($) Total Costs ($) B:C Ratio = Total Costs ($) Benefits (Resource Units) C:E Ratio = ≥ 1.0

4. cost-effectiveness “...coastal wetland restoration projects in Louisiana (will) provide for the long-term conservation of such wetlands... based on the cost-effectiveness of such projects in creating, restoring, protecting, or enhancing coastal wetlands...” (1990 - Public Law 646: CWPPRA, Sec. 3952 1(b)). Efficiency as a Primary Criteria?

5. Is cost-efficacy driving project selection? (C. Aust 2006, Merino et al. 2011) “Candidate Selection Model” using CWPPRA nominee data Binary Logit, 1991-2001, n=350 Probability of Selection = f {CE, Total Cost, Size, Type, Pop, etc.) Cost ($/AAHU) Years 1-5 significant (-) Cumulative Wetland Loss not significant Total Area of Project not significant Total Cost of Project (FFC) significant (-) Rapid-Land Building Projects (MC) significant (+) Population not significant significant (+)Cost ($/AAHU) Years 10-15 Variable Pr.>z; (  =0.05)

6. Restoration Project Selection by CWPPRA (n=124)

7. An Evolving Benefits Construct Pre 2005: Restoration was “habitat-driven” under CWPPRA with a standard efficiency metric ($ per AAHU) Katrina changed things... scale of the crisis is much greater than originally thought The integration of coastal protection and restoration has shifted the definition of benefits Post 2005: Emergence of a “land-building” focus has fueled scientific/ideological debate over competing methods ($/acre)

8. Monetized Estimates of Ecosystem Services Year/Author Ecosystem ServiceValuation Method($/acre/year) Farber (1996)Fisheries ProductionContingent Valuation (Stated) $63 Bergstrom et al., (1990)Cultural/RecreationTravel Cost Method (Revealed) $91 Woodward and Wui (1990)Wetland HabitatContingent Valuation (Stated) $306 Kazmierczak (2001)Water QualityMeta (Stated and Revealed) $825 Petrolia and Kim (2010)Barrier IslandsContingent Valuation (Stated) $2,500 Costanza (2008)Hurricane ProtectionAvoided Damage (Revealed) $3,336

9. Terminal Stocks vs. Aggregate Flows How do MC and DIV projects compare? Years Acres Years Acres 50

10. Freshwater/Sediment Diversions “Although this technique helps protect and sustain existing wetlands, it could take decades for new land to be built with new diversions alone.” – CPRA Master Plan (2007) Tradeoff: Do the benefits of this more “natural” method outweigh the risks of waiting for this land to be restored?

11. Rapid Land Building (Marsh Creation) “Pumping sediments…can build marsh quickly…(but) wetlands built via pipeline may not function in the same ways as wetlands built through natural processes…(and) pumping in sediment is expensive…” – CPRA Master Plan 2007 Tradeoff: Does the risk reduction by moving benefits up in time outweigh the potentially “higher costs” of this technology?

12. Objectives (Caffey and Petrolia – CREST 2008) (Caffey, Wang, and Petrolia 2011) 1.Construct generic benefit trajectories and generic cost models for each project type 2.Develop break-even derivations via ecosystem-service- flow based benefit-cost analysis 3.Conduct sensitivity analyses using risk-adjusted case studies to illustrate trade-offs

13. Data and Methods Data: - CWPPRA, CIAP, LCA,WRDA, STATE (prior to 2012) - Authorized projects and project bids (n=146) Methods: - Generic Benefit and Costs Models (regression, mass-balance) 1. Dredge-based “Marsh Creation” (MC) 2. Diversion-based: - Fitted model from projections (DIV 1 ) - Exogenous model: Boustany 2010 (DIV 2 ) - Benefit-Cost Analysis ESV derived via break-even sensitivity analysis ESV specified for case studies using benefits-transfer

14. Total Acreage Benefits ($) Total Acreage Costs ($) B:C Ratio =...where: b = $benefits (ESV), c = $costs, t = year, and r = discount rate bt (1+r) t = T  t = 0 (1+r) t ct T  t = 0 ≥ 1.0 B:C Ratio Benefit-Cost Analysis

15. Generic Benefit Modeling: Marsh Creation Projects (MC) “Restoration Trajectory” (Percentage of Completion) T MC =1/(1+exp(-(t-0.96)/0.08)) R 2 =0.90 Time Lag Target acreage Restoration (1-2 years)

16. Generic Benefit Modeling: Diversion Projects (DIV 1 ) Target Acreage Restoration (20-50 years) “Restoration Trajectory” (Percentage of Completion) T DIV =-0.0029+0.0501*t R 2 =0.91 Time Lag

17. Parameters for Cost-Benefit Analysis Specified and Derived ComponentParameterComponentParameter Time period (year)SpecifiedNet BI Accretion Rate (%) Derived Desired Scale (acres)SpecifiedTotal Sediments-MC (cuyds) Derived Discount rate (%)SpecifiedTotal Sediments-BI (cuyds) Derived Water Flow Rate- DIV 2 (cfs) Specified Water Flow Rate- DIV1 (cfs) Derived Mob/Demob($) Specified Construction Cost-MC ($) Derived Distance (miles) Specified E&D cost-MC ($) Derived Access Dredging/Channel ($) Specified O&M cost-MC ($) Derived E&D Lag (MC) Specified Fully Funded Cost-MC ($) Derived E&D Lag (BI) Specified Construction Cost-BI ($) Derived E&D Lag (DIV) Specified E&D cost-BI ($) Derived Projected Construction Costs ($) Specified O&M cost-BI ($) Derived Projected E&D cost ($) Specified Fully-Funded Cost-BI ($) Derived Projected O&M cost ($) Specified Construction Cost-DIV1 ($) Derived Land Loss Rate (%/year) Specified E&D cost-DIV 1 ($) Derived Long-shore Transport (%/year) Specified O&M cost-DIV1 ($) Derived Ecosystem Value (Habitat) Specified/Derived Fully-Funded Cost-FWD1 ($) Derived Ecosystem Value (Water Quality) Specified/DerivedE&D cost-DIV 2 ($)Derived Ecosystem Value (Storm Surge) Specified/Derived O&M cost-DIV2 ($) Derived Fully-Funded cost-DIV2 ($) Derived

18. Unconstrained “Break-Even” Analysis What are the annual ecosystem service values required for B:C=1.0? Time (years) Scale (acres) Discount (%) Distance (miles) Break-Even ($/acre/yr) DIV 2 (Boustany 2010) DIV 1 (fitted trajectory) MC DIV 2 DIV 1 MC DIV 2 DIV 1 MC DIV 2 DIV 1 MC Break-Even ($/acre/yr)

19. Social Opposition Risks? (DIV projects) Data: Case Studies, operational regimes of diversions What does history tell us about their operation? How might location and scale affect flow? (25-80% of maxmimum) Refining for Risks Climatological Risks? ( MC and DIV projects) Data: Hurricane Landfall Probabilities ( Klotzbach and Gray 2011) “Expected valuation” construct (will it hit? + will it not hit?) What if it does hit? - Static vs. Dynamic impacts: (20-50% acreage loss)

20. Case Study Assumptions Location:Upper and Lower Estuary Project Types:MC and DIV 2 Project life time: 20 years and 50 years Target scales: 1000 and 5000 acres Time lag: 4 to 10 years Land loss rates: 0.003 to 0.006 per year Hurricane probability: 0.1 to 0.2, X HN Diversion Type: Controlled Diversion Flow:0.25 to 0.80 of capacity Mob/Demob Cost:$1,000,000 Pumping Distance: 4 miles Access Dredging:$600,000 Construction Costs 85% E&D cost 10% O&M cost 5% Ecosystem service values$4,410 per year (Surge, Habitat, W.Q.)

21. Case Study Simulations (16) MCDIV 2 Upper M-1 1000ac/20y Upper M-2 1000ac/50y Upper M-3 5000ac/20y Upper M-4 5000ac/50y Upper D-1 1000ac/20y Upper D-2 1000ac/50y Upper D-3 5000ac/20y Upper D-4 5000ac/50y Net Acres 934853467042671933216021003 NPV Costs ($) 37,798,40037,423,57547,801,52947,327,50912,035,23011,830,91612,082,69511,900,929 NPV Benefits ($) 40,687,95871,993,875203,439,791359,969,3732,399,5967,323,3287,496,97722,880,297 B-C Ratio 1.081.924.267.610.20.62 1.92 $/acre40,46943,87310,23611,09262,35936,85620,07111,865 MCDIV 2 Lower M-1 1000ac/20y Lower M-2 1000ac/50y Lower M-3 5000ac/20y Lower M-4 5000ac/50y Lower D-1 1000ac/20y Lower D-2 1000ac/50y Lower D-3 5000ac/20y Lower D-4 5000ac/50y Net Acres 8727284359363950867115202098 NPV Costs ($) 37,798,40037,423,57547,801,52947,327,50913,366,46513,151,14013,419,17913,229,091 NPV Benefits ($) 38,885,39667,044,229194,426,982335,221,1448,161,17216,722,89424,271,47652,247,394 B-C Ratio 1.031.794.077.080.611.271.813.95 $/acre43,34751,40610,96613,00626,31219,5998,8286,306

22. Results and Conclusions  Efficiency is relative: This analysis was “acreage-focused” with $-benefits expressed via ecosystem service values  Data is limited: Project data is sparse, variable, estimate-based More confidence in MC estimates (some performance data) Less confidence in DIV estimates (entirely projection-based)  Reality: Same project data available to state and feds. Refinement is needed, but not likely to change shape of restoration trajectories. MC (rapid, sigmoid, subject to erosion) DIV (slower, physical sustainability)

23. Tradeoff: Do the benefits of this more “natural” method outweigh the risks of waiting for this land to be restored? Results and Conclusions: Diversions  It depends…., but in our simulations, no. Scale, Time, Location  B-C Ratios ranged from 0.2 to 3.95 (<MC in all simulations) Efficiency generally increases with scale and time  Slow rate of restoration is an economic and ecological disadvantage  Social “sustainability” is a major constraint Limited locations for large-scale diversion Additional costs must be addressed in middle/upper estuary Opportunity cost of delay and/or flow constraint Preemptive compensation?

24. Results and Conclusions: Marsh Creation Tradeoff: Does the risk reduction by moving benefits up in time outweigh the potentially “higher” costs of this technology?  It depends….but in our simulations, yes. Scale, Time, Distance  B-C Ratios ranged from 1.03 to 7.61 (>DIV 2 in all simulations) Efficiency generally increases with scale. Breakeven cost curves for MC converged with DIV (unconstrained) at: 4,000-10,000 ac, 25-35 yrs, 10-20 miles for DIV 2 and DIV 1, But…with risks accounted for, convergence points increase. “Apparent” high costs diminish under time and risk considerations  Rapid rate of restoration is an economic and ecological advantage

25. Louisiana’s Comprehensive Master Plan for a Sustainable Coast (2012), Fig. 9. p 35 Master Plan Implications?

26.  Momentum towards large-scale diversions is grounded in… the geologic history of coastal Louisiana the predominance of biophysical sciences in restoration ...but current assertions regarding the economic supremacy of diversion efficiency are only technically accurate if we… assume the benefit projections for sediment diversions are mostly correct assume the cost projections for sediment diversions are mostly correct assume social/transition costs external to the project (or inconsequential) assume “sustainability” is defined on biophysical terms only assume diversion flow > 80% of design capacity (our study) assume a diversion trajectory of a minimum of 50 years assume diversion benefits > 17x projections (> ???X performance) assume a discount rate of zero assume an acre of wetland in 50 years is of equal value to an acre today assume “$/acre” is a better metric than “BC ratio” Conventional Wisdom on Diversions

27.  Our research is not intended as an argument for or against a particular method of restoration. DIV and MC projects are not infinitely substitutable.  Our study indicates… A.While the estimated costs of terminal stocks (i.e. acres) may indeed be higher for MC (~ 3x)… B.…analysis of benefit flows (i.e. monetized ecosystem services) suggests that MC projects may achieve higher BC ratios (within a 50 year horizon)  “Trajectory Economics” - It’s not only important to address “how” we get the benefits, but also “when” we get the benefits. The operational definition of efficiency needs to be expanded from stocks to flows.  Future revisions needed to address projects of the scale in the SMP, but the generic shape of these trajectories is not likely to change – and that shape has tremendous implications for the underlying economics. Take-Home Points

28. Thank you