# RECORD 467b5db9559e6802d7f96e2e470749e1440b69fd
# TITLE: VALUING OPEN SPACE IN A MARSHLAND ENVIRONMENT: DEVELOPMENT ALTERNATIVES FOR COASTAL GEORGIA
# YEAR: 2016.0  VENUE: Journal of Agricultural and Applied Economics  DOI: 10.1017/aae.2016.25
# AUTHORS: Ajita Atreya; Warren P. Kriesel; J. Mullen
# PAGES: 20  CANDIDATE TABLE PAGES (1-based): [13, 14]
# ABSTRACT: Abstract In a coastal environment, open space can exist as land set aside by a real estate developer or as tidal marshland. In this article, we determine the relative values of both types of open spaces in a coastal county in Georgia using a spatial hedonic price framework. Results indicate that (1) there is a price premium associated with the marshlands and (2) developers have market incentives to incorporate more open space into their designs of residential subdivisions. Regarding marshlands, we also find that accessibility is an important variable that adds much more value to a property than just the proximity.


===== PAGE 1 (score 8, numeric-paren 0, table-words 0, var-words 8) =====

Journal of Agricultural and Applied Economics, 48, 4 ( 2016): 383–402
C⃝2016 The Author(s).
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence
(http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution, and reproduction in any medium, provided
the original work is properly cited.
doi:10.1017/aae.2016.25
VALUING OPEN SPACE IN A MARSHLAND
ENVIRONMENT: DEVELOPMENT
ALTERNATIVES FOR COASTAL GEORGIA
AJITA ATREYA ∗
The Wharton School, University of Pennsylvania, Philadelphia, Pennsylvania
WARREN KRIESEL
Department of Agriculture and Applied Economics, University of Georgia, Athens, Georgia
JEFFREY D. MULLEN
Department of Agriculture and Applied Economics, University of Georgia, Athens, Georgia
Abstract. In a coastal environment, open space can exist as land set aside by a real
estate developer or as tidal marshland. In this article, we determine the relative
values of both types of open spaces in a coastal county in Georgia using a spatial
hedonic price framework. Results indicate that (1) there is a price premium
associated with the marshlands and (2) developers have market incentives to
incorporate more open space into their designs of residential subdivisions.
Regarding marshlands, we also find that accessibility is an important variable that
adds much more value to a property than just the proximity.
Keywords. Coastal, Georgia, market incentive, spatial hedonic
JEL Classifications: R21, R58, Q51
1. Introduction
In most jurisdictions of the United States, land use policy is made at the local
level within the context of existing federal and state regulations. County and
city government officials scrutinize developers’ requests for zoning variations
and building permits. Local officials make their decisions under two potentially
conflicting objectives. First, they are concerned with providing adequate
financing for their jurisdiction, which can be accomplished by increasing the
extent of developed land and improving their real estate tax base. Second, they
are charged with meeting environmental quality objectives, which sometimes
requires the preservation of land in an undeveloped state.
This publication was supported in part by an institutional grant (NA10OAR4170098) to the Georgia
Sea Grant College Program from the National Sea Grant Office, National Oceanic and Atmospheric
Administration, U.S. Department of Commerce.
∗Corresponding author’s e-mail: atreya@wharton.upenn.edu
383
https://doi.org/10.1017/aae.2016.25 Published online by Cambridge University Press


===== PAGE 13 (score 86, numeric-paren 59, table-words 2, var-words 11) =====

Valuing Open Space in a Marshland Environment 395
Table 3. SARAR Model Results using an Inverse Distance Matrix Truncated at 200 Meters
(dependent variable: lnPrice)
Variables
Model 1
Model 2
Model 3
lnHousesize
0.82∗∗∗
0.75∗∗∗
0.75∗∗∗
(0.01)
(0.02)
(0.02)
lnParcelsize
0.13∗∗∗
0.13∗∗∗
0.10∗∗∗
(0.01)
(0.01)
(0.01)
Beds
−0.01
0.001
0.002
(0.008)
(0.009)
(0.008)
Age
−0.007∗∗∗
−0.007∗∗∗
−0.008∗∗∗
(0.0007)
(0.0007)
(0.0007)
Agesq
7.5e-05∗∗∗
7.3e-05∗∗∗
7.5e-05∗∗∗
(6.1e-06)
(6.0e-06)
(5.9e-06)
Fireplace
0.04∗∗∗
0.034∗∗∗
0.04∗∗∗
(0.01)
(0.01)
(0.01)
Brick
−0.013
0.01
0.01
(0.01)
(0.01)
(0.01)
Garage
0.07∗∗∗
0.09∗∗∗
0.10∗∗∗
(0.01)
(0.01)
(0.01)
Deck
0.06∗∗∗
0.04∗∗∗
0.04∗∗∗
(0.01)
(0.01)
(0.01)
Pool
0.10∗∗∗
0.10∗∗∗
0.10∗∗∗
(0.02)
(0.02)
(0.02)
lnCommonspace
0.07∗∗∗
0.06∗∗∗
0.04∗∗∗
(0.005)
(0.005)
(0.006)
lnMajorroads
0.02∗∗∗
0.008
0.002
(0.005)
(0.005)
(0.005)
lnSchool
0.03∗∗∗
0.04∗∗∗
0.04∗∗∗
(0.008)
(0.009)
(0.008)
SFHA
0.03∗
0.009
−0.0004
(0.01)
(0.015)
(0.015)
Bltpostfir
0.06∗∗∗
0.04∗∗
0.06∗∗∗
(0.01)
(0.018)
(0.018)
lnNonwhite
−0.09∗∗∗
−0.0861∗∗∗
−0.086∗∗∗
(0.006)
(0.006)
(0.006)
lnIncome
0.26∗∗∗
0.25∗∗∗
0.27∗∗∗
(0.02)
(0.019)
(0.0189)
lnPark
−0.09∗∗∗
−0.08∗∗∗
(0.008)
(0.008)
lnMarsh
−0.07∗∗∗
−0.04∗∗∗
(0.005)
(0.006)
Waterview
0.10∗∗∗
0.17∗∗∗
(0.02)
(0.02)
Wateraccess
0.38∗∗∗
(0.06)
Dock
0.51∗∗∗
(0.03)
Year2005
0.22∗∗∗
0.21∗∗∗
0.21∗∗∗
(0.016)
(0.02)
(0.01)
https://doi.org/10.1017/aae.2016.25 Published online by Cambridge University Press


===== PAGE 14 (score 46, numeric-paren 24, table-words 5, var-words 7) =====

396
AJITA ATREYA ET AL.
Table 3. Continued
Variables
Model 1
Model 2
Model 3
Year2006
0.41∗∗∗
0.40∗∗∗
0.40∗∗∗
(0.017)
(0.02)
(0.07)
Year2007
0.51∗∗∗
0.50∗∗∗
0.50∗∗∗
(0.018)
(0.018)
(0.02)
Year2008
0.37∗∗∗
0.36∗∗∗
0.37∗∗∗
(0.02)
(0.019)
(0.02)
Year2009
0.21∗∗∗
0.20∗∗∗
0.21∗∗∗
(0.02)
(0.02)
(0.020)
Year2010
0.05∗∗
0.04∗∗
0.05∗∗
(0.02)
(0.02)
(0.020)
Constant
3.67∗∗∗
5.08∗∗∗
5.07∗∗∗
(0.21)
(0.22)
(0.22)
Lambda
0.03∗∗∗
0.02∗∗∗
0.026∗∗∗
(0.003)
(0.003)
(0.003)
Rho
3.65∗∗∗
3.56∗∗∗
3.5∗∗∗
(0.10)
(0.07)
(0.0814)
Notes: N = 11,011. Standard errors in parentheses. ∗∗∗P < 0.01, ∗∗P < 0.05, ∗P < 0.1. SARAR, spatial
autoregressive model with autoregressive residuals.
(i.e., all continuous variables were transformed to their natural logarithms).5
Throughout, heteroskedasticity consistent standard errors are reported. The
estimate for λ, the spatial autoregressive coefficient, is significant and in the
expected direction indicating a positive adjacency effect; that is, a higher sale
price of adjacent properties should result in a higher average sale price, ceteris
paribus. Regarding the interpretation of the regression coefficients in the SARAR
model, a marginal effect is calculated by multiplying an estimate into a spatial
multiplier, 1/(1 −λ), as suggested by Kim, Phipps, and Anselin (2003).
All coefficients for the structural variables have their expected sign and are
statistically significant except Beds and Brick.6 The quadratic specification for the
variable Age seems to capture the diminishing marginal effects as shown by the
positive and significant Agesq variable. Housesize and Parcelsize have significant
impacts on property prices. A 1% increase in Housesize and Parcelsize increases
property prices by 0.75% to 0.82% and 0.10% to 0.13%, respectively, which is
equivalent to approximately $1,389 to $1,506 and $22 to $27 when evaluated
for an average priced property across the three models (Table 4).7 In Table 4,
model 3 results indicate that a property with a fireplace (Fireplace), garage
(Garage), deck (Deck), or pool (Pool), however, commands even higher price
5 We performed the Box-Cox transformation of the dependent variable and concluded that the
natural log of price as the dependent variable was the best specification for our model. Others have used
semiparametric models, but those models are difficult to interpret.
6 The interpretation hereafter will be in regard to model 3 unless specified otherwise.
7 Note that the spatial multiplier (1/1 −λ) is taken into account when calculating marginal effects.
https://doi.org/10.1017/aae.2016.25 Published online by Cambridge University Press


===== PAGE 15 (score 16, numeric-paren 0, table-words 1, var-words 14) =====

Valuing Open Space in a Marshland Environment 397
Table 4. Marginal Effects Evaluated for an Average Property in Chatham County
Variables
Model 1
Model 2
Model 3
Housesize
$1,506.1
$1,375.4
$1,389.1
Parcelsize
$26.7
$27.3
$21.8
Beds
−$1,292.8
$114.5
$212.1
Age
−$67.2
−$68.3
−$75.5
Agesq
$0.7
$0.7
$0.7
Fireplace
$12,055.6
$10,218.9
$11,708.3
Brick
−$3,819.6
$2,779.2
$3,106.3
Garage
$22,111.8
$27,319.3
$28,404.5
Deck
$17,009.1
$12,168.2
$12,992.6
Pool
$29,094.5
$29,150.5
$27,747.4
Commonspace
$4,476.1
$3,726.1
$3,075.0
Majorroads
$14.8
$5.8
$2.0
School
$3.4
$4.7
$5.5
SFHA
$8,206.1
$2,678.8
−$115.0
Bltpostfir
$16,501.8
$12,050.0
$16,606.7
Nonwhite
−$821.5
−$777.9
−$783.0
Income
$1.3
$1.2
$1.2
Park
−$7.6
−$7.9
Marsh
−$53.0
−$31.1
Waterview
$29,534.4
$49,282.3
Wateraccess
$112,901.3
Dock
$154,716.7
Note: Marginal effect calculation takes into account the spatial multiplier.
premiums equal to approximately $11,708, $28,404, $12,992, and $27,747,
respectively, when evaluated for an average priced home.
All coefficients for neighborhood and location variables are of the expected
sign and are statistically significant except the proximity to roads (lnMajorroads)
and location in SFHA. In Chatham County, the coefficient of lnCommonspace
indicates that a 1% increase in the percentage of common area in the
neighborhood increases the property prices on average by approximately $3,075
when evaluated for an average price home. This result indicates that developers
have an incentive to keep more common space in their design feature for greater
revenue. The proximity to school significantly decreases the price of the property,
which may be attributed to the level of noise from a school. An increase in the
proportion of nonwhites in a community is negatively related to the property
prices, whereas an increase in income is positively related to the property prices
in Chatham County.
Regarding the environmental amenity variables, all of those included in the
model have their expected positive sign and are statistically significant. The
variables lnPark, lnMarsh, and Waterview all seem to be important factors
adding value to a property. The finding that both lnCommonspace and lnMarsh
are significant, positive contributors to property prices is surprising. The initial
https://doi.org/10.1017/aae.2016.25 Published online by Cambridge University Press