# RECORD b3ec523ce2f4119ea59d806bdaeb62c8215fe9a8
# TITLE: The implicit value of tree cover in the U.S.: A meta-analysis of hedonic property value studies
# YEAR: 2016.0  VENUE: nan  DOI: 10.1016/j.ecolecon.2016.04.016
# AUTHORS: S. D. Siriwardena; K. Boyle; T. Holmes; P. Wiseman
# PAGES: 9  CANDIDATE TABLE PAGES (1-based): [5, 7, 9]
# ABSTRACT: Trees in residential neighborhoods and communities provide benefits for homeowners that are capitalized into residential property values. In this paper, we collected data from hedonic property value studies and merged these data with ancillary spatial data describing forest and socio-economic characteristics surrounding each study area to conduct a meta-analysis of the impact of tree canopy cover on the value of residential properties. The meta-analysis suggests that property-level tree cover of about 30% and county-level tree cover of about 38% maximize the implicit price of tree cover in property values. Currently, tree cover in the original study areas was about 14%, on average, around or near study properties. The empirical results, therefore suggest under investment of tree cover on private property from the perspective of individual property owners and from a societal perspective. The findings also have implications for community forest programs regarding planting trees and protection of mature trees to address potential changes in tree abundance, species diversity and stand age due to development and climate change.


===== PAGE 1 (score 17, numeric-paren 0, table-words 0, var-words 17) =====

Analysis
The implicit value of tree cover in the U.S.: A meta-analysis of hedonic
property value studies☆
Shyamani D. Siriwardena a,⁎,1, Kevin J. Boyle b,1, Thomas P. Holmes c,1, P. Eric Wiseman d,1
a Department of Forest Resources and Environmental Conservation, 313 Cheatham Hall (0324), Virginia Tech, Blacksburg, VA 24061, USA
b Virginia Tech Program in Real Estate, 420 Bishop-Favrao Hall (0715), Virginia Tech, Blacksburg, USA
c USDA Forest Service, Southern Research Station, Research Triangle Park, NC, USA
d Department of Forest Resources and Environmental Conservation, Virginia Tech, Blacksburg, USA
a b s t r a c t
a r t i c l e
i n f o
Article history:
Received 18 March 2015
Received in revised form 1 March 2016
Accepted 3 April 2016
Available online 4 May 2016
Trees in residential neighborhoods and communities provide beneﬁts for homeowners that are capitalized into
residential property values. In this paper, we collected data from hedonic property value studies and merged
these data with ancillary spatial data describing forest and socio-economic characteristics surrounding each
study area to conduct a meta-analysis of the impact of tree canopy cover on the value of residential properties.
The meta-analysis suggests that property-level tree cover of about 30% and county-level tree cover of about
38% maximize the implicit price of tree cover in property values. Currently, tree cover in the original study
areas was about 14%, on average, around or near study properties. The empirical results, therefore suggest
under investment of tree cover on private property from the perspective of individual property owners and
from a societal perspective. The ﬁndings also have implications for community forest programs regarding plant-
ing trees and protection of mature trees to address potential changes in tree abundance, species diversity and
stand age due to development and climate change.
© 2016 Elsevier B.V. All rights reserved.
Keywords:
Hedonic valuation
Tree canopy cover
Community forestry
Forest matrix
Climate change
1. Introduction
Tree cover, the land area covered by tree crowns, in residential
neighborhoods is an environmental attribute that provides a suite of
ecosystem services (e.g., landscape esthetics, wildlife habitat, energy
conservation, stormwater control) to homeowners and communities
(Nowak et al., 2010). Most trees in urban and suburban neighborhoods
are located on private property where the costs of tree planting and
maintenance are borne by homeowners, while the beneﬁts of trees
can be enjoyed by the greater community. The disparity between who
pays the costs (homeowners) and who receives the beneﬁts from
trees on privately-owned property (homeowners and the community)
suggests that the level of tree cover on privately-owned land may be so-
cially sub-optimal. A better understanding of the costs and beneﬁts of
trees, within the context of multiple social-ecological settings (Pickett
et al., 2011), can help community planners design tree planting and
protection programs to improve the quality of life for people and justify
the costs of such programs.
Although the economic beneﬁts of tree canopy cover (tree cover
hereafter) are not directly priced, one component of their economic
value, the implicit value to property owners, can be estimated using
hedonic property value models. Hedonic models decompose the total
price paid for a property into the “implicit” prices of property character-
istics. The implicit price of a characteristic, such as the amount of tree
cover, is the amount home buyers pay for a small change in the level
of the characteristic, holding all other property characteristics constant
(Taylor, 2003).
Previous hedonic studies have found that the value of forested areas
varies with the degree of forest fragmentation and urbanization. For
example, larger forest patches are valued more in the urban core,
while a more fragmented and diverse landscape is preferred in rural–
urban interfaces (Cho et al., 2008). Trees are valued more in areas
where they are scarce and less so in heavily forested areas (Netusil
et al., 2010). Large trees with full canopies have been shown to enhance
retail settings (Wolf, 2004) and large shade trees are preferred by
residents in areas with hot summers (Schroeder et al., 2006). Temporal
variations of the value of trees has also been observed by Cho et al.
(2009) where the amenity value of trees increased over time with
declining forest patch size and density.
Forested areas can also present real or perceived risks that lower
property values. Locations that have experienced or where there are
Ecological Economics 128 (2016) 68–76
☆This research was supported by the Southern Research Station of the U.S. Forest
Service (12-JV-11330143-106) and Virginia Agricultural Experimental Station.
⁎ Corresponding author.
E-mail addresses: shyadil@vt.edu (S.D. Siriwardena), kjboyle@vt.edu (K.J. Boyle),
tholmes@fs.fed.us (T.P. Holmes), pwiseman@vt.edu (P.E. Wiseman).
1 Siriwardena, Boyle and Wiseman are PhD candidate, Professor and Associate
Professor, respectively, at Virginia Tech. Holmes is an Economist with the U.S. Forest
Service.
http://dx.doi.org/10.1016/j.ecolecon.2016.04.016
0921-8009/© 2016 Elsevier B.V. All rights reserved.
Contents lists available at ScienceDirect
Ecological Economics
journal homepage: www.elsevier.com/locate/ecolecon


===== PAGE 5 (score 36, numeric-paren 0, table-words 9, var-words 14) =====

purchasers place on tree cover. These socioeconomic variables and
the other ancillary variables discussed above are likely to be ﬁxed
for any given hedonic study, but vary spatially across studies and
provide the potential for a meta-analysis to provide insights on the
value of tree cover beyond what can be observed in any single
hedonic study.
4. Model Speciﬁcation
The ﬁrst model to be estimated is the selection model (Eqs. (4a) and
(4b)). The selection Eq. (4a) is speciﬁed as:
h Z; γ
ð
Þ ¼ ϒ0 þ ϒ1onprop þ ϒ2onprop  medtrees þ ϒ3onprop  oldtrees
þ ϒ4onprop  invasive þ ϒ5onprop  temp þ ϒ6popdensity
þ ϒ7medianincome þ ω
ð5aÞ
where Z is a vector of study and context characteristics, and γ is a vector
of coefﬁcients to be estimated. Each of the variables listed in Eq. (4a) is
deﬁned in Table 2. The dependent variable is binary and indicates if an
implicit price is positive (=1) or negative (=0), and the selection equa-
tion is speciﬁed as a probit. The selection equation includes an explana-
tory variable (onprop) delineating whether the tree cover measure is on
a property (=1) or includes tree cover near a property (=0), and this
variable is interacted with other variables that may inﬂuence whether
tree cover is viewed as an amenity or disamenity (medtrees, oldtrees,
invasive, and temp). As noted previously, older trees may provide
shade, but also potentially pose a greater risk of damage or injury. Inva-
sive tree pests may affect the amenity value of trees and may increase
the risk of tree breakage. Higher temperatures may create a desire for
more tree shade for outdoor comfort and air conditioning savings,
while lower temperatures may create a desire for less tree shade to
allow winter sunlight to heat the home and hasten melting of ice and
snow. Population density and median income are included to consider
if these demographic characteristics affect whether tree cover is viewed
as an amenity.
The meta-Eq. (4b) is speciﬁed as:
gþ X; θþ


¼ θþ
0 þ θþ
1 treecover þ θþ
2 treecover2 þ θþ
3 ctreecover
þθþ
4 ctreecover2 þ θþ
5 houseprice þ θþ
6 eastern þ θþ
7 pacificnw
þ θþ
8 medtrees þ θþ
0 oldtrees þ θþ
10signif þ θþ
11popdensity
þ θþ
12medianincome þ εþ
ð5bÞ
where all variables are deﬁned in Table 2 and the θi
+ are coefﬁcients to
be estimated. The dependent variable is the consistent measure of the
implicit price (effect size) explained in the previous section. The
treecover variable represents the tree cover variable from the original
hedonic models, and ctreecover is included to consider if tree cover in
the greater study area where people work, shop, and recreate also
Table 2
Deﬁnitions and summary statistics of variables used to estimate selection equation and meta-equation.
Variable
Description
Positive observations (IP N 0)
Mean
(min, max)
Negative observations (IP b 0)
Mean
(min, max)
Dependent variable in selection equation
amenity
1 if implicit price (IP) is positive and 0 if IP is negative
0.64
(0.00, 1.00)
NA
Dependent variable in meta-equation
IP
Implicit price of tree cover for a 1% increase ($2013)
$239.41
($0.32, $2696.15)
-$156.26
(−$895.94, −$0.04)
Independent variables
onprop
1 if treecover on property, 0 otherwise
0.35
(0.00, 1.00)
0.40
(0.00, 1.00)
medtreesa
% of county forests aged 40 to 119 years
69
(8, 91)
64
(45, 91)
oldtreesa
% of county forests aged 120 years or older
11
(0.00, 44)
16
(0.00, 44)
invasive
1 if invasive species or pathogen, 0 otherwise
0.80
(0.00, 1.00)
0.70
(0.00, 1.00)
temp
Annual days with temperature above 90 °F
15
(0.00, 76)
14
(3, 45)
popdensity
County population density (people/mile2)
886
(7, 2715)
1245
(7, 2715)
medianincome
County median income
$56,084
($34,000, $81,330)
$53,911
($35,613, $81,330)
treecover
% tree cover on or near property
14
(0.10, 61)
12
(0.00, 40)
ctreecover
% county-level tree cover
44
(13,66)
48
(13, 66)
houseprice
Mean sale price of properties ($2013)
$285,138
($24,576, $494,894)
$249,200
($24,576, $494,894)
easternb
1 if study area is east of Mississippi River, 0 otherwise
0.69
(0.00, 1.00)
0.47
(0.00, 1.00)
paciﬁcnwb
1 if study area is in Oregon or Washington, 0 otherwise
0.21
(0.00, 1.00)
0.42
(0.00, 1.00)
signif
1 if tree cover signiﬁcant at 10% level, 0 otherwise
0.66
(0.00, 1.00)
0.47
(0.00, 1.00)
Minimum and maximum of the variables are given in parenthesis.
a Percent of county forest below 40 years of age is the omitted category.
b Omitted category includes the following study locations in the mountain west: Greater Flagstaff area, Arizona; Grand County, Colorado; and Flathead, Lake, Sanders, Lincoln
and Missoula Counties, Montana.
72
S.D. Siriwardena et al. / Ecological Economics 128 (2016) 68–76


===== PAGE 6 (score 26, numeric-paren 0, table-words 8, var-words 10) =====

affects the implicit value of tree cover on or near a property. Both of
these variables are included as linear and squared terms (treecover2
and ctreecover2) to allow for nonlinear relationships between the im-
plicit value and the tree cover measures.
The mean value of house prices (houseprice) from the original
hedonic studies is included to see if implicit values of tree cover are
affected by selling prices of properties. Binary variables for studies
conducted in the eastern U.S. and Paciﬁc Northwest are included to
control for spatial ﬁxed effects in different areas of the country
(e.g., the risk of forest ﬁres). The omitted category includes studies
conducted in the mountain west. Tree age is included because the
age of trees affects the size of the crown and could affect the implicit
price of tree cover. The omitted category is young trees (b40 years of
age). Finally, a variable is included to see if the signiﬁcance of
implicit price estimates in the original hedonic studies affects the
magnitudes of the estimates in meta-equation; all implicit price
estimates, statistically signiﬁcant and insigniﬁcant, from the hedonic
studies are included in the meta-data.
The selection model estimates Eqs. (5a) and (5b) jointly
(Wooldridge, 2002), and we conducted a test for the presence of
selection from only using studies with positive implicit prices in
the estimation of Eq. (5b):
Ho : ρ ¼ 0 vs Ha : ρ ≠0
ð6Þ
where ρ is the correlation between the error terms in Eqs. (5a) and
(5b) (ω and ε+, respectively). If the null hypothesis cannot be
rejected, then the meta-equation (5b) can be estimated solely,
without the selection equation.
A number of robustness analyses are conducted to provide insight
into how much conﬁdence can be placed in the meta-equation coefﬁ-
cient estimates and thus conclusions drawn from the meta-analysis.
Given that most studies provided multiple implicit-price estimates
and these multiple estimates from individual studies cannot be as-
sumed to be independent observations, the meta-equation is estimated
using OLS, study ﬁxed effects, random effects and multilevel mixed-
effects. In addition, horizontal robustness checks, excluding individual
observations or studies (groups of observations) are also conducted
(Boyle et al., 2013). The primary focus of these robustness analyses is
to consider if the exclusions of observations from the data affects the
estimates of the coefﬁcients on treecover and ctreecover. The results of
these robustness analyses are reported below.
5. Results
Summary statistics for the variables used in the estimation are
reported in Table 2. These data indicate that 64% of the observations
have positive implicit prices. Of the studies with positive implicit
prices, the average implicit price for a 1% change in tree cover is
$239, and the comparable ﬁgure for negative implicit prices is −
$156. Values of the variable treecover for the positive implicit prices
range from 0.1% to 61% and the comparable range for negative
implicit prices is 0% to 40%.
5.1. Selection Equation
Estimation of the Heckman sample selection model (Eqs. (5a) and
(5b)) resulted in the conclusion that the null hypothesis of no correla-
tion (Eq. (6)) could not be rejected (p = 0.93). As there is no sample
selection, the primary focus of the analyses reported in this paper is
the meta-equation. The results from the estimation of the Heckman
selection equations are reported in the Appendix A.
5.2. Meta-equation Estimation Results and Discussion
The meta-equation is estimated with White's consistent standard
errors (Nelson and Kennedy, 2009).6 In this estimation, we found
that popdensity and medianincome were insigniﬁcant. This is driven
in large part by the study-speciﬁc auxiliary data being correlated,
which reduces the efﬁciency of the estimation of all coefﬁcient esti-
mates. Thus, we test the null hypothesis that the coefﬁcients on
popdensity and medianincome are jointly insigniﬁcant in the meta-
equation:
Ho : θþ
11 ¼ θþ
12 ¼ 0 vs Ho : θþ
11 ≠0; θþ
12 ≠0
ð7Þ
We could not reject this null hypothesis (F = 2.11, p = 0.13). Given
this result, the estimation results reported here are for Eq. (5b) with the
popdensity and medianincome variables excluded.7
The treecover, treecover2, ctreecover and ctreecover2 variables all have
signiﬁcant coefﬁcient estimates (Table 3). The linear terms of these var-
iables have positive coefﬁcient estimates and the squared terms have
negative coefﬁcients. This supports the intuition from previous research
that people value tree cover up to a point, but beyond this point tree
cover declines in value. This could occur for a variety of reasons,
including:
• some shade is good, but too much shade blocks sunlight for lawns and
passive solar heating;
• tree cover provides privacy, but too much cover may hamper home
security or obscure vistas;
• trees make a home more attractive, but too much tree cover may lead
to buildup of mildew and algae on exterior structures and excessive
cleanup of leaves and litter;
• trees attract watchable wildlife, but too much tree cover may attract
nuisance wildlife;
• trees provide cover from harsh wind and rain, but may break apart in
inclement weather or create a wildﬁre hazard;
There could be a variety of other explanations, but these considerations
are consistent with what other investigators have found.
It is interesting to note that the implicit prices of tree cover are
higher for eastern study sites (eastern) than for mountain west study
sites (the omitted category in the estimation), and implicit prices for
Paciﬁc Northwest studies are not signiﬁcantly different from mountain
west studies. This difference could be due to a higher risk of forest
ﬁres in the mountain west and Paciﬁc Northwest, a preference for
more vistas in the west, a preference for more privacy in the east, or
other explanations.
The coefﬁcient for older forests (oldtrees) is positive and signiﬁcant,
indicating signiﬁcantly higher implicit prices for study sites with older
versus younger forests (the omitted category) in the county where the
studies were conducted. This suggests people value living in areas of
mature forests that provide more shade and may be visually appealing.8
6 To account for multiple observations from most studies, we estimated the meta-
equation using ﬁxed-effects, random-effects, and mixed-effects estimation. The ﬁxed-
effects estimation did not work because of collinearity issues. The random-effects and
mixed-effects estimations did not change the quantitative and qualitative results for the
meta-equation coefﬁcient estimates.
7 We estimated the meta-equation using negative implicit prices as the dependent var-
iable and none of the explanatory variables were signiﬁcant.
8 This is an interesting outcome when compared to the result from the selection equa-
tion reported in the Appendix. Older trees increase the probability of a negative implicit
price, but also increase the magnitude of positive implicit prices. These results suggest
the concurrent amenity/disamenity of older trees; more shade and a desirable esthetic,
but higher perceived risks with more maintenance costs.
73
S.D. Siriwardena et al. / Ecological Economics 128 (2016) 68–76


===== PAGE 7 (score 58, numeric-paren 11, table-words 21, var-words 3) =====

5.3. Tree Cover than Maximizes Property Values
Using the results shown in Table 3, the levels of treecover (tc) and
ctreecover (ctc) that maximize implicit prices were computed as:
∂IPþ
∂tc ¼ 0 and ∂IPþ
∂ctc
ð8Þ
The level of treecover that maximizes implicit prices is 30% with a
standard error of 15% and the comparable level for ctreecover is 38%
with a standard error of 3%.9
It is worth noting that treecover ranges from 0.1% to 61% in the data
used to estimate the meta-equation and the comparable range for
ctreecover is 13% to 66% (see Table 2). Thus, the levels of treecover and
ctreecover that maximize implicit prices are within the observed data
on tree cover. Increasing tree cover beyond these levels does not result
in a negative implicit price; it is just that the implicit prices decline from
their maximum values.
An additional consideration is that the mean value of treecover is
14% (Table 2), while the value that maximizes implicit price for this
variable is 30% (more than one standard deviation below the value
that maximizes implicit prices). Further, 14% tree cover on or near
a property is much less for tree cover in the county where the prop-
erties are located (44%, Table 2). These relationships support the no-
tion that individual property owners underinvest in trees on their
property from both the perspective of individual property owners
and a societal perspective.
5.4. Robustness of Meta-equation Coefﬁcient Estimates
Conducting robustness checks in any empirical analysis is always
important, but it is particularly important for this analysis when estima-
tion is done with only 68 observations. With a small sample size, a single
observation or group of observations (study) has the potential to inﬂu-
ence estimation results.
Removal of single observations and re-estimation of the meta-
equation did not induce major changes in the estimates of the key coef-
ﬁcients for treecover, treecoversq, ctreecover and ctreecoversq. None of
these coefﬁcient estimates changed sign, the average change in the
magnitude of the estimated coefﬁcients for these variables is only 4%,
and there were three occurrences where treecover and treecoversq
became insigniﬁcant (Table 4). The only variable that exhibited
substantial sensitivity to removing observations is houseprice with an
average effect of 19%, but this variable is insigniﬁcant in the original
meta-equation reported in Table 3 and only becomes signiﬁcant in
two instances in the observation-removal robustness analysis.
In contrast, all coefﬁcient estimates for treecover, treecoversq,
ctreecover and ctreecoversq exhibited more variability when studies
(sets of observations) are removed from the estimation. Recall,
there are thirteen studies with positive implicit price estimates and
the number of observations range from one to ten per study
(Table 1). Here the coefﬁcient estimates for treecover, treecoversq,
ctreecover and ctreecoversq sometimes changed signiﬁcance or sign
up to two times; the average change in coefﬁcient estimates ranges
35% (treecover) to 56% (ctreecover) (Table 5).
Given these study sensitivity results, it is logical to ask which studies
are affecting the estimation. This analysis is conducted by looking at the
changes in the level of tree cover that maximizes implicit prices
(Table 6). Recall the level of treecover that maximizes implicit prices,
based on the full estimates with all observations, is 30% and the compa-
rable value for ctreecover is 38%. The good news is that only three studies
resulted in changes of the level of treecover by more than 10%, but the
bad news is that the effects were quite large: 51% (Holmes et al.,
2006), −26% (Netusil et al., 2010) and −11% (Drake-McLaughlin and
Netusil, 2011). The effects on the level of ctreecover were generally
smaller; again, two studies resulted in greater than 10% changes in the
maximum implicit price(−13% — Holmes et al., 2006 and 15% —
Holmes et al., 2010). It is important to note that studies with the largest
percentage deviations in the levels of treecover that maximize implicit
prices, when their observations are removed from the estimation,
have some of the largest numbers of observations in the data (9 —
Holmes et al., 2006; 10 — Holmes et al., 2010; 7 — Netusil et al., 2010;
7 — Drake-McLaughlin and Netusil, 2011). The key insight from the
study robustness checks is that while coefﬁcient estimates varied, in
most cases the linear and square term coefﬁcient estimates moved in
a relationship that does not change the levels of treecover and ctreecover
that maximize implicit prices.
The violations of robustness do not imply that the inﬂuential
studies should be removed from the analysis, nor do they imply
that the meta-analysis results are invalid, but they do suggest that
the ﬁndings reported here should be interpreted as preliminary
and with caution. More studies are needed to add spatial breadth
and location-speciﬁc depth of implicit price estimates to conﬁrm
the preliminary ﬁndings presented here. It is crucial that the future
studies focus on percent tree cover as the key variable so that the
new empirical estimates can be merged with the empirical estimates
used here to estimate a new meta-equation.
6. Conclusions
This study ﬁnds a nonlinear relationship between the implicit
value of tree cover and two types of tree canopy cover: tree cover
on or near a property, and tree cover in the county where a property
is located. This nonlinear relationship allows the computation of the
9 The standard errors were approximated using the Marcov Chain Monte-Carlo
(MCMC) approach using the Gibbs sampler (Plassmann and Khanna, 2007). This method
generates a large number of samples from posterior distributions of coefﬁcient estimates
for treecover and ctreecover and posterior distributions of turning points of treecover and
ctreecover that maximize the implicit prices. Posterior means and standard deviations
are calculated for these ﬁnite sample approximations of the sampling distribution of the
levels of treecover and ctreecover that maximize implicit prices.
Table 3
Meta-equation coefﬁcient estimates (positive implicit prices, IP+).
Variables
Coefﬁcient estimates
treecover
22.32⁎
(12.64)
treecover2
−0.37⁎⁎
(0.18)
ctreecover
84.10⁎⁎
(35.94)
ctreecover2
−1.11⁎⁎
(0.47)
houseprice
−0.0002
(0.0007)
eastern
1393.97⁎⁎
(681.76)
paciﬁcnw
351.75
(239.73)
oldtrees
41.07⁎⁎
(17.66)
medtrees
5.08
(4.63)
signif
100.54
(98.15)
intercept
−2971.9⁎⁎
(1176.81)
n
68
R 2
0.27
Robust standard errors are given in parentheses.
⁎ Indicates signiﬁcance at the 10% level.
⁎⁎ Indicates signiﬁcance at the 5% level.
74
S.D. Siriwardena et al. / Ecological Economics 128 (2016) 68–76


===== PAGE 8 (score 18, numeric-paren 0, table-words 3, var-words 12) =====

amount of property-level and county-level tree cover that maxi-
mizes the implicit value of tree canopy cover: 30% for tree cover on
or near a property, and 38% at the county level. This pattern of results
suggests that property owners prefer more tree cover in the general
area where they work, shop and recreate, as a public good, than
on their own property, as a private good. This discrepancy may
be due to the fact that homeowners generally pay all of the
planting and maintenance costs and bear most of the perceived
risks (i.e., wildﬁre and storm damage) for trees located on their
property. When these costs and risks are either borne by other
private landowners, or are shared by all taxpayers for trees lo-
cated on public land, higher levels of county-level tree cover
are preferred.
The ﬁnding that the implicit value of tree cover is maximized at
about 38% for county-level cover is consistent with the often-cited
40% tree cover espoused by American Forests as an ecological goal for
communities on the East Coast and Paciﬁc Northwest of the United
States. The average urban tree cover in the U.S. is currently about 35%
(Nowak et al., 2010). Therefore, in many areas, overall tree cover
could be rightly increased with a concomitant positive effect on residen-
tial property values.
The preference for old trees (more than 120 years old) found in
this meta-analysis is another interesting result. A simplistic view of
land development may be to cut down old trees and then plant new
ones upon completion of construction rather than invest in protec-
tion and restorative care of old trees; here again, this may be indic-
ative of underinvestment in community tree cover. These veteran
trees are much more ecologically valuable than young trees for car-
bon storage, air pollution abatement, stormwater interception, and
energy conservation, and these results provide evidence that they
have more amenity value too. Thus, communities might wish to
consider public outreach programs, zoning policies, and tax incen-
tives to protect and maintain older tress. Although such endeavors
come at a cost for the community, these costs might be recouped in
part through additional real estate tax revenues on higher-
appraising residential property.
While the ﬁndings of this study may be helpful to local ofﬁcials
involved in planning community forestry programs, we expect that
decision-making by both community planners and households
will become more complex in the future due to various outcomes
anticipated from a changing climate. On the one hand, a hotter cli-
mate may increase the value of tree cover for its ability to shade
and cool homes. This preference may be most acute in urban
areas that are vulnerable to “heat island” effects. Further, tree can-
opy and forested riparian buffers intercept stormwater, helping to
mitigate the ﬂooding and water quality effects of runoff. On the
other hand, climate change may increase the frequency and sever-
ity of extreme weather events, which could lead to an increase in
storm damage to trees, causing increased power outages and debris
cleanup costs. This may push communities to think twice about
maintaining large, mature trees in residential areas. Consequently,
the future may bring complex challenges to residential property
owners and community forest planners alike. Because tree-cover
considerations may vary by geographic location, preferences for
property-level and county-level tree cover may become even
more nuanced and, over time, local planners may ﬁnd it useful to
engage residents to discuss and reﬁne the goals of community for-
estry programs.
Table 4
Observation robustness (removing one observation at a time from estimation).
Variables
Number of
times sign
changed
Number of times
signiﬁcance changed
(at 10%)
Absolute, percent changes in
coefﬁcient magnitudes
Average
(min, max)
treecover
0
3
4%
(0, 52)
treecover2
0
3
4%
(0, 39)
ctreecover
0
0
4%
(0, 21)
ctreecover2
0
0
4%
(0, 19)
houseprice
2
0
19%
(0, 177)
eastern
0
1
5%
(0, 52)
paciﬁcnw
0
6
5%
(0, 40)
oldtrees
0
1
4%
(0, 31)
medtrees
0
2
7%
(0, 179)
signif
0
0
7%
(0, 55)
Minimum and maximum of the variables are given in parentheses.
Table 5
Study robustness (removing one study at a time from estimation).
Variable
Number of
times sign
changed
Number of times
signiﬁcance changed
(at 10%)
Absolute, percent changes in
coefﬁcient magnitudes
Average
(min, max)
treecover
0
3
35%
(1, 97)
treecover2
2
2
46%
(3, 167)
ctreecover
1
1
56%
(1, 291)
ctreecover2
1
1
55%
(0, 294)
houseprice
5
1
225%
(4, 930)
eastern
2
2
65%
(6, 272)
paciﬁcnw
1
2
66%
(2, 315)
oldtrees
1
2
59%
(7, 244)
medtrees
2
3
88%
(0, 443)
signif
0
0
19%
(0, 95)
Minimum and maximum of the variables are given in parentheses.
Table 6
Changes in implicit prices by study for study robustness.
Study
Percent changes
treecover
ctreecover
Coley (2005)
4%
−2%
Dimke (2008)
NA
−2%
Drake-McLaughlin and Netusil (2011)
−11%
1%
Holmes et al. (2006)
51%
−13%
Holmes et al. (2010)
2%
15%
Kim and Wells (2005)
2%
−1%
Mansﬁeld et al. (2005)
0%
5%
Netusil et al. (2010)
−26%
0%
Paterson and Boyle (2002)
NA
−3%
Price et al. (2010)
−1%
1%
Sander et al. (2010)
2%
−2%
Sander and Haight (2012)
1%
−1%
Stetler et al. (2010)
7%
−4%
NA indicates coefﬁcient estimate of treecoversq and ctreecoversq have the wrong signs and
are insigniﬁcant so tree cover that maximizes implicit prices could not be calculated.
75
S.D. Siriwardena et al. / Ecological Economics 128 (2016) 68–76


===== PAGE 9 (score 46, numeric-paren 22, table-words 1, var-words 18) =====

Appendix A. Heckman Selection Equation Estimation
Table A1
Joint estimation of selection equation and meta-equation.
Variable
Coefﬁcient
Selection equation
onprop
3.27
(4.03)
onprop⁎medtrees
0.003
(0.06)
onprop⁎oldtrees
−0.19⁎⁎
(0.07)
onprop⁎invasive
−2.87
(2.69)
onprop⁎temp
0.01
(0.05)
popdensity
0.0004⁎⁎
(0.0002)
medianincome
0.00002⁎
(0.00001)
intercept
−1.03
(0.71)
Meta-equation
treecover
19.09
(11.92)
treecover2
−0.33⁎
(0.17)
ctreecover
49.23
(40.06)
ctreecover2
−0.70
(0.51)
houseprice
−0.002
(0.002)
eastern
735.87
(1132.8)
paciﬁcnw
397.96
(667.9)
medtrees
11.72
(8.55)
oldtrees
19.53
(23.83)
signif
112.53
(94.06)
popdensity
−0.21
(0.15)
medianincome
0.005
(0.02)
intercept
−1696.32
(1151.4)
ρ
−0.05
(0.52)
Robust standard errors are given in parentheses.
⁎ Indicates signiﬁcance at the 10% level.
⁎⁎ Indicates signiﬁcance at the 5% level.
References
Bentz, B., 2008. Western U.S. bark beetles and climate change. U.S. Department of Agricul-
ture, Forest Service, Climate Change Resource Center (http://www.fs.fed.us/ccrc/
topics/bark-beetles.shtml, accessed 11.02.2014).
Boyle, K.J., Parmeter, C.F., Boehlert, B.B., Paterson, R.W., 2013. Due diligence in meta-
analyses to support beneﬁt transfers. Environ. Resour. Econ. 55 (23), 357–386.
Cho, S.-H., Poudyal, N.C., Roberts, R.K., 2008. Spatial analysis of the amenity value of green
open space. Ecol. Econ. 66 (2), 403–416.
Cho, S.-H., Kim, S.G., Roberts, R.K., Jung, S., 2009. Amenity values of spatial conﬁgurations
of forest landscapes over space and time in the Southern Appalachian Highlands.
Ecol. Econ. 68 (10), 2646–2657.
Cho, S.H., Lambert, D., Kim, S., Roberts, R., Park, W., 2011. Relationship between value of
open space and distance from housing locations within a community. J. Geogr. Syst.
13 (4), 393–414.
Coley, M.C., 2005. House and Landscape Value: An Application of Hedonic Pricing Tech-
nique Investigating Effects of Lawn Area on House Selling Price (MS Thesis) The Uni-
versity of Georgia.
Dale, V.H., Joyce, L.A., McNulty, S., Neilson, R.P., Ayres, M.P., Flannigan, M.D., Hanson, P.J.,
Irland, L.C., Lugo, A.E., Peterson, C.J., Simberloff, D., Swanson, F.J., Stocks, B.J.,
Wotton, B.M., 2001. Climate change and forest disturbances. Bioscience 51 (9),
723–734.
Dimke, K.C., 2008. Valuation of Tree Canopy on Property Values of Six Communities in
Cincinnati, Ohio (PhD Dissertation) The Ohio State University.
Drake-McLaughlin, N., Netusil, N.R., 2011. Valuing walkability and vegetation in Portland,
Oregon. In: von Haefen, R.H. (Ed.), Twenty-second Interim Report and Proceedings
from the Annual Meeting, W2133, pp. 173–201.
Frankel, S.J., 2008. Forest Plant Diseases and Climate Change. U.S. Department of Agricul-
ture, Forest Service, Climate Change Resource Center (http://www.fs.usda.gov/ccrc/
topics/plant-diseases.shtml, accessed 11.02.2014).
Ghermandi, A., Nunes, P.A., 2013. A global map of coastal recreation values: results from a
spatially explicit meta-analysis. Ecol. Econ. 86, 1–15.
Heckman, J.J., 1979. Sample selection bias as a speciﬁcation error. Econometrica 47 (1),
153–161.
Holmes, T.P., Murphy, E.A., Bell, K.P., 2006. Exotic forest insects and residential property
values. Agric. Resour. Econ. Rev. 35 (1), 155–166.
Holmes, T.P., Murphy, E.A., Bell, K.P., Royle, D.D., 2010. Property value impacts of hemlock
woolly adelgid in residential forests. For. Sci. 56 (6), 529–540.
Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N., McKerrow, A.,
VanDriel, J.N., Wickham, J., 2007. Completion of the 2001 national land cover database
for the conterminous United States. Photogramm. Eng. Remote Sens. 73 (4), 337–341.
Hope, D., Gries, C., Zhu, W., Fagan, W.F., Redman, C.L., Grimm, N.B., Nelson, A.L., Martin, C.,
Kinzig, A., 2003. Socioeconomics drive urban plant diversity. Proc. Natl. Acad. Sci. 100
(15), 8788–8792.
Huggett Jr., R.J., 2003. Fire in the Wildland-Urban Interface: An Examination of the Effects of
Wildﬁre on Residential Property Markets (PhD Dissertation) North Carolina State
University.
Iverson, L.R., Prasad, A.M., 2001. Potential changes in tree species richness and forest com-
munity types following climate change. Ecosystems 4 (3), 186–199.
Johnston, R.J., Besedin, E.Y., Stapler, R., 2014. Enhanced geospatial data for meta-analysis
and environmental beneﬁt transfer: an application to water quality improvements.
Presented at the Meta-analysis of Economics Research Network, MAER-Net 2014
Athens Colloquium, Athens, Greece.
Kim, Y.S., Wells, A., 2005. The impact of forest density on property values. J. For. 103 (3),
146–151.
Mansﬁeld, C., Pattanayak, S.K., McDow, W., McDonald, R., Halpin, P., 2005. Shades of
green: measuring the value of urban forests in the housing market. J. For. Econ. 11
(3), 177–199.
Nelson, J.P., Kennedy, P.E., 2009. The use (and abuse) of meta-analysis in environmental and
natural resource economics: an assessment. Environ. Resour. Econ. 42 (3), 345–377.
Netusil, N.R., Chattopadhyay, S., Kovacs, K.F., 2010. Estimating the demand for tree cano-
py: a second-stage hedonic price analysis in Portland, Oregon. Land Econ. 86 (2),
281–293.
Nowak, D.J., Greenﬁeld, E.J., 2012. Tree and impervious cover change in US cities. Urban
For. Urban Green. 11 (1), 21–30.
Nowak, D.J., Stein, S.M., Randler, P.B., Greenﬁeld, E.J., Comas, S.J., Carr, M.A., Alig, R.J., 2010.
Sustaining America's urban trees and forests: a forests on the Edge report. Gen. Tech.
Rep. NRS-62. U.S. Department of Agriculture, Forest Service, Northern Research
Station.
Paterson, R.W., Boyle, K.J., 2002. Out of sight, out of mind? Using GIS to incorporate visi-
bility in hedonic property value models. Land Econ. 78 (3), 417–425.
Pickett, S.T., Cadenasso, M.L., Grove, J.M., Booned, C.G., Groffman, P.M., Irwin, E., Kaushal,
S.S., Marshall, V., McGrath, B.P., Nilon, C.H., Pouyat, R.V., Szlavecz, K., Troy, A., Warren,
P., 2011. Urban ecological systems: scientiﬁc foundations and a decade of progress.
J. Environ. Manag. 92 (3), 331–362.
Plassmann, F., Khanna, N., 2007. Assessing the precision of turning point estimates in
polynomial regression functions. Econ. Rev. 26 (5), 503–528.
Prasad, A., Iverson, L.S., Matthews, S., Peters, M., 2009. Atlases of tree and bird species hab-
itats for current and future climates. Ecol. Restor. 27 (3), 260–263.
Price, J.I., McCollum, D.W., Berrens, R.P., 2010. Insect infestation and residential property
values: a hedonic analysis of the mountain pine beetle epidemic. Forest Policy Econ.
12 (6), 415–422.
Sander, H., Haight, R.G., 2012. Estimating the economic value of cultural ecosystem
services in an urbanizing area using hedonic pricing. J. Environ. Manag. 113,
194–205.
Sander, H., Polasky, S., Haight, R.G., 2010. The value of urban tree cover: a hedonic property
price model in Ramsey and Dakota Counties, Minnesota, USA. Ecol. Econ. 69 (8),
1646–1656.
Schroeder, H., Flannigan, G., Coles, R., 2006. Residents' attitudes toward street trees in the
UK and US communities. Arboricult. Urban For. 32 (5), 236–246.
Stetler, K.M., Venn, T.J., Calkin, D.E., 2010. The effects of wildﬁre and environmental
amenities on property values in northwest Montana, USA. Ecol. Econ. 69 (11),
2233–2243.
Taylor, L.O., 2003. The hedonic method. In: Champ, P., Boyle, K., Brown, T. (Eds.), A primer
on nonmarket valuation. Kluwer Academic Publishers, New York.
Troy, A.R., Grove, J.M., O'Neil-Dunne, J.P., Pickett, S.T., Cadenasso, M.L., 2007. Predicting
opportunities for greening and patterns of vegetation on private urban lands. Envi-
ron. Manag. 40 (3), 394–412.
U.S. Department of Labor, Bureau of Labor Statistics, 2015. Consumer Price Index Data
from 1913 to 2016. (http://www.usinﬂationcalculator.com/inﬂation/consumer-
price-index-and-annual-percent-changes-from-1913-to-2008/, accessed July
2013).
Wolf, K.L., 2004. Trees and business district preferences: a case study of Athens, Georgia,
US. J. Arboric. 30 (6), 336–346.
Woodall, C.W., Oswalt, C.M., Westfall, J.A., Perry, C.H., Nelson, M.D., Finley, A.O., 2009. An
indicator of tree migration in forests of the Eastern United States. For. Ecol. Manag.
257 (5), 1434–1444.
Wooldridge, J., 2002. Econometric Analysis of Cross Section and Panel Data. MIT Press,
Cambridge.
76
S.D. Siriwardena et al. / Ecological Economics 128 (2016) 68–76