Transcript Linking Forests to Faucets with a Distant Municipal Area

Linking Forests to Faucets with a
Distant Municipal Area:
Investigating Public Support for
Water Security and Watershed
Protection
Dadhi Adhikari
Janie Chermak
Jennifer Thacher
Robert P. Berrens
Study Area
Survey Approach






Two Focus Groups & 100 pre-test surveys
Population: 190,000 Albuquerque households, obtained from
County Assessor Data provided by ABCWUA
Sampling frame: 104,000 home owners
Sample selected: 2596, proportional to HH# by zipcodes
Five contacts (two mailed questionnaires 9/2013 - 10/2013
Responses received through mail and internet
Mailed surveys: 2596
Undelivered surveys: 133
Net mailed surveys: 2463
Responses (n=911): (a) Mail-751 (b) Online-160
Response Rate: 37%
Willingness to Pay Study

Good to be Valued: Creating water source protection fund to
conduct land treatments on 30,000 acres/yr in watershed lands
north of Albuquerque to reduce the risk of high-severity wildfire.

Payment Vehicle : Annual fee collected through water utility bills,
property taxes, or insurance premium taxes.

Valuation Question: Currently overgrown brush and trees are
removed from approximately 3000 acres/year in the larger
watershed. The University of New Mexico is trying to figure out at
what level, if any, metropolitan Albuquerque homeowners would
support a Water Source Protection Fund to conduct land treatments
on 30,000 acres/year in the same area and reduce the risk of highseverity wildfire. A required annual fee of all homeowners could be
targeted for this purpose. Different people might be willing to pay
different amounts to the Water Protection Fund. What is the most
your household would be willing to pay per year to the Water
Source Protection Fund ? Fill in the blank
$---------------- per year
Uncertainty

Delivery Uncertainty- Two questions
◦ On a scale of 0 to 10, where 0 means “Not at all likely” and 10 means
“Highly likely” and 5 is halfway in between, how likely do you feel it is
that wildfires will impact your supply of drinking water if fire-prone land
in the watershed are not treated to reduce wildfire risk? Circle one.
◦ Suppose the Water Source Protection Fund is put in place and funds are
targeted to minimize the risk of high-severity wildfire in the forested
area north of Albuquerque. On a scale of 0 to 10, where 0 means “Not
at all effective” and 10 means “Highly effective” and 5 is halfway in
between, how effective do you feel the program would be in ensuring
the sustainability of maintaining metropolitan Albuquerque’s supply of
water ?

Preference Uncertainty – After the WTP question
◦ On a scale of 0 to 10, where 0 means “Completely uncertain” and 10
means “Completely certain” and 5 is halfway in between, how certain
are you of your answer to question 11 (wtp question)?
Descriptive Statistics
Variable
Description
Obs
Mean (St.D)
WTP
Willingness to Pay ($/year)
825
PERCEPCC
Perception on how serious the climate change problem is,
measured in 1-4 likert scale (1- Not Serous, 4- Extremely serious)
755
62.60
(81.86)
2.41
(1.03)
PERCEPWS
Perception on how serious the water supply problem is, measured
in 1-4 likert scale (1- Not Serous, 4- Extremely serious)
762
3.27
(0.80)
PERCEPPW
Perception on how serious the payment for water problem is,
measured in 1-4 likert scale (1- Not Serous, 4- Extremely serious)
761
2.36
(0.99)
PERCEPWFR
Perception on how serious the wildfire fire risk problem is,
measured in 1-4 likert scale (1- Not Serous, 4- Extremely serious)
763
3.10
(0.86)
PERCEPTAX
Perception on how serious the tax problem is, measured in 1-4
likert scale (1- Not Serous, 4- Extremely serious)
762
2.58
(1.00)
WHITE
1 if white, 0 otherwise
825
MALE
1 if male, 0 otherwise
824
FAMSIZE
Family Size
825
INCOME
Income (measured in $1000)
758
EDU-UG
1 if has undergraduate education, 0 otherwise
818
EDU-G
1 if has graduate level education, 0 otherwise
818
YEARNM
Numbers of years lived in the New Mexico
792
OPNPRCBRN
1 if the respondent supports prescribed burning method of treating
forest, 0 otherwise
Delivery uncertainty measured in 0-10 likert scale, 0 being fully
certain and 10 being fully uncertain.
813
0.84
(0.37)
0.62
(0.49)
2.35
(1.22)
80.35
(44.11)
0.37
(0.48)
0.35
(0.48)
35.11
(0.48)
0.73
(0.44)
1.68
(1.70)
DELIVUNCRN
761
WTP Distribution
Regression Results
PERCEPCC
PERCEPWS
PERCEPPW
PERCEPWFR
PERCEPTAX
WHITE
GENDER
FAMSIZE
INCOME
EDU-UG
EDU_G
YEARNM
OPNPRCBRN
OLS
W/O Uncertainty
OLS
With Uncertainty
TOBIT
W/O Uncertainty
TOBIT (2)
With Uncertainty
Double Hurdle
W/O Uncertainty
Double Hurdle
With Uncertainty
0.142**
(2.21)
0.346***
(3.73)
-0.209***
(-2.79)
0.395***
(4.47)
-0.363***
(-4.67)
0.116
(0.60)
0.0937
(0.73)
-0.0227
(-0.45)
0.00442***
(3.11)
0.268
(1.63)
0.0897
(0.50)
-0.00588*
(-1.73)
0.410**
(2.56)
0.0469
(0.71)
0.179**
(2.12)
-0.207***
(-2.98)
0.208**
(2.49)
-0.282***
(-3.81)
0.311*
(1.68)
0.0657
(0.54)
-0.0334
(-0.71)
0.00512***
(3.49)
0.225
(1.46)
-0.00313
(-0.02)
-0.00753**
(-2.30)
0.169
(1.08)
-0.349***
(-8.90)
2.766***
(5.29)
607
0.323
0.166**
(2.10)
0.413***
(3.93)
-0.238***
(-2.70)
0.458***
(4.61)
-0.416***
(-4.71)
0.113
(0.52)
0.114
(0.73)
-0.0222
(-0.37)
0.00514***
(2.77)
0.339*
(1.81)
0.120
(0.57)
-0.00736*
(-1.87)
0.476***
(2.73)
0.0340
(0.74)
0.166**
(2.50)
-0.124**
(-2.34)
0.216***
(3.50)
-0.193***
(-3.64)
0.0861
(0.66)
0.0258
(0.28)
-0.0335
(-0.93)
0.00202*
(1.88)
0.0744
(0.65)
-0.0466
(-0.37)
-0.000707
(-0.30)
0.286***
(2.68)
0.763
(1.31)
653
0.0378
(0.49)
0.184*
(1.78)
-0.223***
(-2.62)
0.222**
(2.24)
-0.326***
(-3.81)
0.341
(1.63)
0.0841
(0.57)
-0.0313
(-0.54)
0.00616***
(3.51)
0.297*
(1.65)
0.0136
(0.07)
-0.00931**
(-2.48)
0.202
(1.19)
-0.482***
(-9.04)
2.803***
(4.63)
607
2.962***
(8.26)
653
-0.00126
(-0.02)
0.191**
(2.53)
-0.171***
(-2.90)
0.120*
(1.66)
-0.195***
(-3.30)
0.214
(1.45)
0.0420
(0.41)
-0.0365
(-0.90)
0.00325***
(2.71)
0.0262
(0.21)
-0.137
(-0.96)
-0.00290
(-1.10)
0.178
(1.46)
-0.226***
(-5.11)
3.304***
(7.64)
607
-1220.5
2471.0
2538.2
-1072.4
2176.8
2247.4
-1011.2
2080.5
2210.5
-973.2
2004.4
2132.3
DELIVUNCRN
CONSTANT
No.of Obs
R-Square
Log Likelihood
AIC
BIC
1.282**
(2.44)
653
0.229
Mean and Median WTP
Models
Mean WTP
($/year)
Median WTP
($/year)
OLS (without accounting for uncertainty)
37.3
26.95
OLS (with accounting for uncertainty)
28.6
22.38
Tobit (without accounting for uncertainty)
40.1
27.38
Tobit (with accounting for uncertainty)
31.15
22.54
Double Hurdle (without accounting for uncertainty)
51.72
48.06
Double Hurdle (with accounting for ncertainty)
37.18
33.87
WTP: Albuquerque vs Santa Fe
Santa Fe: Probability of WTP
ABQ: Probability of WTP
$/Month
$/Month
96.22
0.65
86.31
88.67
1
75.54
73.11
1.5
2
60.06
56.45
47.34
Probability Without Uncertainty
Probability With Uncertainty
66%
0.65
1
1.5
2
16%
58%
20%
47%
42%
Very Will
Mean Household WTP (Tobit Model)
Without accounting for Uncertainty: approx $40/year
With accounting for uncertainty: approx $31/year
16%
23%
22%
SW Will
20%
29%
35%
SW/V Unwill
2%
2%
1%
1%
DK/ NA
Who should collect the funds?
Respondent's Preference for Institutions
36.51%
30.07%
12.62%
11.63%
9.16%
Water Utilities
Counties
Municipalities
State
Other