Meta-Analysis of Wetland Values: Modeling Spatial Dependencies Randall S. Rosenberger Oregon State University Meidan Bu Microsoft
Overview Spatial relationships in metadata Spatial econometric modeling Application to wetland valuation studies in North America Sensitivity analysis to intra-study dependence Conclusions
Research questions Are wetland values correlated across space? What is the spatial relationship of wetland welfare estimates? geographic closeness ecological linkages socio-economic characteristics of local people
Spatial Relationships Proximity matters – location, location, location Hedonic values increase with proximity to positive amenities Hedonic values decrease with proximity to disamenities Spatial heterogeneity matters (50km radius) Previous wetland values MRA results Marginal values increase with local GDP Marginal values increase with population density Marginal values decrease with resource density
Statistical Problems Locational aspects lead to: Spatial heterogeneity Metadata augmentation – GDP, population & resource density Omitted variable problem Spatial dependence Spatial lag – correlation in dependent variable Omitted variable problem – biased, inconsistent estimates Spatial error – correlation in errors Uncorrelated error problem – inefficient estimates
Spatial Modeling
The Empirical Model
Spatial Weight Matrix Definition
Spatial Weight Matrices W defined as Threshold (Euclidean) distances Ecological similarity Economic similarity
Threshold Distance W Any two sites within a threshold are considered neighbors
Ecological similar neighbors Any two sites located in the same boundary are considered neighbors The USGS Hydrologic Unit 2 (HUC2) unit (n=21)
Economic similar neighbors Any two sites sharing the same socioeconomic attributes (i.e., latent demand) are considered neighbors local education level population density within 50km radius county level average personal income local GDP Multivariate hierarchical clustering analysis local education level Group observations into clusters (n=40) that have similar values of measured variables
Multivariate hierarchical clusters
An economic similarity cluster
Wetland Metadata Wetland welfare estimates from primary studies conducted in North America through studies, 163 value estimates Explanatory variables Study attributes Valuation methodology Wetland ecosystem type Ecological functions valued Geographic and socio-economic characteristics
Results – Methodology, Ecosystem Spatial model OLSThreshold distance Ecological similarity Economic similarity 50km lag100km lag150km lag Estimate Intercept * * * ** * Wetland area (ha) - log scaled ** * * Economic literature dummy1.19 ** * 0.94 * 0.95 * * 0.99 * Regional study dummy ** 0.82 * Valuation methodology (Travel Cost Method as the reference group) CVM1.66 ** 2.01 *** 2.08 *** 2.06 *** 1.84 *** 1.60 ** Choice Experiment3.14 ** 3.54 *** 3.49 *** 3.52 *** 3.39 *** 3.34 *** Hedonic Price7.13 *** 6.80 *** 7.24 *** 7.40 *** 6.71 *** 6.86 *** Market Price1.99 ** 2.13 ** 2.13 ** 2.28 ** 2.20 ** 1.81 ** Replacement Cost4.27 *** 4.26 *** 4.22 *** 4.45 *** 4.49 *** 3.99 *** Production Function ** 1.85 ** 1.89 ** 1.85 ** 1.49 * Wetland ecosystem type (Estuarine as the reference group) Riverine2.00 ** ** 1.82 ** 1.81 ** 1.62 ** Palustrine Lacustrine *
Results – Ecosystem Functions Spatial model OLSThreshold distance Ecological similarity Economic similarity 50km lag100km lag150km lag Estimate Ecological function valued Preservation2.89 ** 2.77 ** 3.00 *** 3.08 *** 3.04 *** 2.82 ** Restoration Water quality *2.17*1.95*1.88*1.78 Flood control & water supply Amenity-2.69**-2.64***-2.52**-2.46**-2.32**-2.64*** Recreational fishing & hunting2.58**2.28**2.54**2.62**2.58**2.34** Non-consumptive recreation3.26***3.05***3.34***3.42***3.11***3.01*** Biodiversity Commercial fishing & hunting
Results – Geographic/Socioeconomic Spatial model OLSThreshold distance Ecological similarity Economic similarity 50km lag100km lag150km lag Estimate Geographic and socio-economic information Ramsar Site dummy Wetland area in 50km radius (ha)/ log scaled-0.26 * ** ** ** ** ** Population in 50km radius -log scaled0.40 ** 0.26 * 0.27 * * 0.25 * 0.38 ** Education (county level)0.06 ** 0.07 *** 0.06 *** 0.06 ** 0.08 *** 0.06 ** Distance to city (km)0.07 ** 0.10 *** 0.08 *** 0.08 *** 0.08 *** 0.08 *** ** *** *** *** *** ***
Results – Test Statistics Spatial model OLSThreshold distance Ecological similarity Economic similarity 50km lag100km lag150km lag Estimate N163 R2R Likelihood ratio test statistic P-value for the likelihood ratio test <0.000*** *** *** *** ** AIC
Recap – Spatial MRAs Positive spatial correlation for all three neighborhood criteria Threshold distance neighbors are strongest correlation Spatial correlation exists as far as 150km Economic similarity defined neighbors has the weakest correlation Covariate estimates are robust to spatial dependence, although magnitude varies some
Intra-study Correlation What about confounding intra-study correlation? An unbalanced panel meta-dataset with 163 observations from 80 wetland sites 39 wetland sites report multiple measures (max = 16 obs.)
Bootstrap Sensitivity Analysis Bootstrap draw one observation per wetland site Form 1000 sub-datasets Repeat spatial MRAs Test the significance of spatial correlation for every combination Count the number of significant LLR results Test the robustness of the spatial correlation
Sensitivity Analysis Results Weight Matrix Significant LLR p ≤ 0.05 Binomial test Significant LLR p ≤ 0.10 Binomial test 50 km threshold933p < p < km threshold908p < p < km threshold747p < p < 0.00 Ecological similarity49p = p = 0.24 Economic similarity12p = p = 1.00
Recap – Sensitivity Analysis Significant evidence of spatial correlation exists in threshold distance defined neighbors Inconclusive evidence of spatial correlation in ecological and economic defined neighbors Ecological similarity – HUC2 may be too large Economic similarity – intra-study correlation
Conclusions Spatial correlation exists, although partial effects are robust to specifications Threshold distance is robust to intra-study correlation Future issues: Other spatial models (e.g., spatial error specification)? What are the implications for international benefit transfers? Are results consistent for other spatially dependent metadata?
Q&A We hope you enjoyed this tour of spatial econometric modeling in an MRA framework THANK YOU!
The Parking Lot - Descriptives MeanMeanSt. Dev.MinMax Wetland welfare estimate/ha/year in 2010 USD – log scaled Wetland area (ha) - log scaled Economic literature dummy Regional study dummy
The Parking Lot - Descriptives MeanMeanSt. Dev.MinMax Valuation methodology (binary variables) CVM Choice Experiment Travel Cost Hedonic Price Market Price Replacement Cost Production Function
The Parking Lot - Descriptives MeanMeanSt. Dev.MinMax Wetland ecosystem type (binary variables) EstuarineEstuarine RiverineRiverine Palustrine Lacustrine
The Parking Lot - Descriptives MeanMeanSt. Dev.MinMax Ecological function valued (binary variables) Preservation Restoration Water quality Flood control & water supply AmenityAmenity Recreational fishing & hunting Non-consumptive recreation Biodiversity Commercial fishing & hunting
The Parking Lot - Descriptives MeanMeanSt. Dev.MinMax Geographic and socio-economic characteristics Ramsar Site dummy Wetland area in 50km radius (ha) Population in 50km radius Education (county level) Distance to city (km)
The Parking Lot – Best Fit Model We also isolated the best fit (i.e. largest LLR) single observation model from among the 1000 bootstrapped samples These results follow: Inferences remain consistent across models Magnitudes of effects are not robust to model specification Likely due to small observations – n = 80
Best Fit Single Observation Models Spatial lag model OLSThreshold distance weight Ecological similarity weight Economic similarity weight 50km lag100km lag150km lag Estimate Intercept Wetland area (ha) - log scaled-0.30*** ***-0.25*** *** Economic literature dummy Regional study dummy1.39* *0.90* * Valuation methodology (Travel Cost Method as the reference group) CVM *1.36* Choice Experiment4.85**5.31***3.84***4.56*** ** Hedonic Price11.68***10.51***12.29***13.49***10.21***10.75*** Market Price *2.65**3.17** ** Replacement Cost3.80**3.49***2.57**3.32***2.41*3.38** Production Function *2.08**2.35**
Best Fit Single Observation Models Spatial lag model OLSThreshold distance weight Ecological similarity weight Economic similarity weight 50km lag100km lag150km lag Estimate Ecological function valued Preservation6.64**6.74***2.59**2.91**5.41**1.67 Restoration **2.46* ***-1.34 Water quality ** **-0.24 Flood control & water supply ** **1.83 Amenity ***-7.68*** *** Recreational fishing & hunting7.15**7.38***2.78**3.18**6.96**1.70 Non-consumptive recreation7.41**7.65***3.41***3.76***7.13***2.71** Biodiversity8.06**9.34***5.66***6.27***7.85***4.06*** Commercial fishing & hunting6.06**4.79** **-0.09
Best Fit Single Observation Models Spatial lag model OLSThreshold distance weight Ecological similarity weight Economic similarity weight 50km lag100km lag150km lag Estimate Geographic and socio-economic information Ramsar Site dummy Wetland area in 50km radius (ha)/ log scaled *-0.22**-0.23*-0.14 Population in 50km radius -log scaled Education (county level) *0.06**0.04* Distance to city (km) ***0.07**0.07** *0.04 Education * City ***-0.002**-0.002** **-0.001
Best Fit Single Observation Models Spatial lag model OLSThreshold distance weight Ecological similarity weight Economic similarity weight 50km lag100km lag150km lag Estimate N80 R-square64.47% Rho LLR test statistic P-value for the LLR test<0.000***<0.000***<0.000***<0.000 ** *<0.000*** AIC