# RECORD 38d51e7656f94829f9f43fba3ef936f46b613e29
# TITLE: A Hedonic Analysis of the Impact of Marine Aquaculture on Coastal Housing Prices in Maine
# YEAR: 2017.0  VENUE: Agricultural and Resource Economics Review  DOI: 10.1017/age.2017.19
# AUTHORS: Keith S. Evans; Xuan Chen; Christina A. Robichaud
# PAGES: 26  CANDIDATE TABLE PAGES (1-based): [16, 17, 20]
# ABSTRACT: Converting coastal waters to farmed production of seafood may generate conflicts with other resource users. This study explores the impact of marine aquaculture development on coastal homeowners. Using single-family home sales from 2012–2014 and spatial data on coastal aquaculture activity, we employ hedonics to assess the impacts of mariculture development in three study areas of Maine, USA. Our results suggest modest impacts on residential property values with significant spatial variation across study areas. This spatial variation represents a challenge for managers and highlights the potential benefits from coordinating the development of aquaculture to balance resource users' objectives with industry growth.


===== PAGE 1 (score 6, numeric-paren 0, table-words 0, var-words 6) =====

A Hedonic Analysis of the Impact of
Marine Aquaculture on Coastal
Housing Prices in Maine
Keith S. Evans, Xuan Chen, and Christina A. Robichaud
Converting coastal waters to farmed production of seafood may generate conflicts
with other resource users. This study explores the impact of marine aquaculture
development on coastal homeowners. Using single-family home sales from 2012–
2014 and spatial data on coastal aquaculture activity, we employ hedonics to
assess the impacts of mariculture development in three study areas of Maine,
USA. Our results suggest modest impacts on residential property values with
significant spatial variation across study areas. This spatial variation represents a
challenge for managers and highlights the potential benefits from coordinating
the development of aquaculture to balance resource users’ objectives with
industry growth.
Key
Words:
coastal
waters,
hedonic
pricing
model,
marine
aquaculture,
mariculture, property values
Aquaculture is an important source of fish protein. While wild-capture
production has flat-lined since the mid-1980s, due to excessive fishing
pressure and changing ocean conditions, world production from aquaculture
has grown exponentially to meet market demand (World Bank 2013, Food
and Agriculture Organization of the United Nations (FAO) 2016). Although
China has represented the majority of this growth, generating more than 60
percent of production by volume (FAO 2016), early research suggests that a
wide-range of marine production opportunities exist for the United States
(Knapp 2008, Valderrama and Anderson 2008, Kite-Powell, Rubino, and
Keith Evans is an Assistant Professor in the School of Economics and School of Marine Sciences,
University of Maine, Orono, ME 04469. Xuan Chen is an Assistant Professor in the School of
Economics, University of Maine, Orono, ME 04469. Christina Robichaud is a graduate student in
the School of Economics, University of Maine, Orono, ME 04469. Correspondence: Keith S. Evans
▪School of Economics ▪5782 Winslow Hall, Room 206 ▪Orono, ME 04469 ▪Phone 207.581.3178 ▪
email: keith.evans@maine.edu.
The authors thank Maine Multiple Listings Service for providing access to the sales transactions
data. We would also like to thank the participants at the 2016 NAREA workshop and two
anonymous reviewers for helpful suggestions. This project was supported by the National
Science Foundation under EPSCoR award #IIA-1355457 and the USDA National Institute of
Food and Agriculture under Hatch projects #ME021603 and #ME021704.
The views expressed are the authors’ and do not necessarily represent the policies or views of
any sponsoring agencies.
Agricultural and Resource Economics Review 46/2 (August 2017) 242–267
© The Author(s) 2017. 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.
https://doi.org/10.1017/age.2017.19 Published online by Cambridge University Press


===== PAGE 16 (score 150, numeric-paren 124, table-words 1, var-words 11) =====

Table 3. Parameter Estimates from the Box-Cox Hedonic Pricing Model. Sales prices are transformed using the Box-
Cox transformation parameter λ. Robust standard errors are reported in parentheses. Parameter estimates for
municipality and year fixed effects are available upon request.
Control variables
Casco Bay
Damariscotta
Penobscot Bay
Bathrooms
0.324***
0.323***
0.583***
0.584***
1.080***
1.110***
(0.036)
(0.036)
(0.131)
(0.132)
(0.167)
(0.171)
Living Space (1,000s square feet)
1.092***
1.091***
1.247***
1.249***
1.338***
1.371***
(0.033)
(0.033)
(0.139)
(0.140)
(0.221)
(0.226)
Lot size (100s acres)
3.137***
3.133***
5.604***
5.609***
11.074***
11.376***
(0.580)
(0.579)
(0.686)
(0.687)
(1.563)
(1.602)
Age (50 years)
0.218***
0.218***
0.270***
0.269***
0.589***
0.601***
(0.023)
(0.023)
(0.051)
(0.051)
(0.076)
(0.078)
Age squared
0.005***
0.005***
0.004***
0.004***
0.012***
0.012***
(0.001)
(0.001)
(0.001)
(0.001)
(0.002)
(0.002)
Cabin (0/1)
0.317
0.311
0.717*
0.714*
1.685**
1.744**
(0.250)
(0.249)
(0.368)
(0.369)
(0.783)
(0.810)
Distance to water (miles)
0.171***
0.171***
0.237
0.244
0.396
0.301
(0.041)
(0.041)
(0.172)
(0.173)
(0.252)
(0.259)
Distance sq to water (miles)
0.018***
0.018***
0.067**
0.068**
0.112***
0.101**
(0.006)
(0.006)
(0.031)
(0.031)
(0.043)
(0.044)
View of water (% 2–mile buffer)
2.282***
2.289***
3.396***
3.397***
5.458***
6.140***
(0.141)
(0.142)
(0.473)
(0.474)
(0.751)
(0.789)
Waterfront (0/1)
1.333***
1.333***
1.545***
1.549***
2.848***
2.941***
(0.101)
(0.101)
(0.172)
(0.172)
(0.297)
(0.304)
Govt. Beach/Dock (0/1)
0.221*
0.219*
0.118
0.115
0.865***
0.886***
(0.121)
(0.121)
(0.269)
(0.270)
(0.309)
(0.316)
Continued
Evans et al.
A Hedonic Analysis of the Impact of Marine Aquaculture 257
https://doi.org/10.1017/age.2017.19 Published online by Cambridge University Press


===== PAGE 17 (score 111, numeric-paren 84, table-words 3, var-words 9) =====

Table 3. Continued
Control variables
Casco Bay
Damariscotta
Penobscot Bay
Elevation (100s feet)
0.101***
0.101***
0.328***
0.329***
0.029
0.038
(0.027)
(0.027)
(0.122)
(0.122)
(0.081)
(0.083)
Seasonal homes (% census tract)
0.007**
0.007**
0.024
0.025
0.068
0.087
(0.003)
(0.003)
(0.057)
(0.057)
(0.063)
(0.067)
Median income† ($10,000s)
0.004
0.005
1.245***
1.241***
0.613*
0.639*
(0.015)
(0.015)
(0.450)
(0.451)
(0.343)
(0.352)
Spending per student ($1,000s)
0.048
0.049
0.043
0.042
0.141*
0.136*
(0.032)
(0.032)
(0.049)
(0.049)
(0.073)
(0.074)
Sale in winter (0/1)
0.144***
0.144***
0.065
0.067
0.054
0.060
(0.041)
(0.041)
(0.149)
(0.149)
(0.195)
(0.201)
Constant
19.804***
19.778***
29.587***
29.613***
25.349***
25.787***
(0.203)
(0.203)
(2.639)
(2.643)
(1.852)
(1.916)
Aquaculture Variables and Transformation Parameter
Neighborhood attitudes
0.005
0.012
0.778
0.770
0.000
0.061
(0.058)
(0.055)
(0.512)
(0.512)
(0.335)
(0.336)
Aquaculture index (base)
0.885
0.694***
23.392**
(2.922)
(0.215)
(10.882)
Aquaculture index (alternate)
7.894
4.342***
57.609**
(9.837)
(1.367)
(24.721)
Transformation parameter (λ)
0.100***
0.099***
0.110**
0.111**
0.141***
0.143***
(0.010)
(0.010)
(0.022)
(0.022)
(0.017)
(0.017)
Number of observations
5698
5698
1238
1238
1644
1644
Adjusted R2
0.733
0.733
0.598
0.598
0.590
0.590
F–statistics (Wald)
307.754
307.968
39.255
39.200
35.260
35.568
† Median household income ($10,000s) in the home’s census tract.
Significance levels: *** p < 0.01, ** p < 0.05, and * p < 0.10
Agricultural and Resource Economics Review
258 August 2017
https://doi.org/10.1017/age.2017.19 Published online by Cambridge University Press


===== PAGE 18 (score 13, numeric-paren 0, table-words 1, var-words 11) =====

We find more interesting patterns related to the Aquaculture index, which
captures the spatial arrangement of leases in relation to a home. Our results
suggest variation in the impact on housing prices across the three regions.
We fail to find evidence of any impact in Casco Bay, while we find statistically
significant evidence for the other study areas. This pattern is robust to our
specification of the Aquaculture index (base and alternate form). It is difficult
to directly compare the magnitude of coefficients across models, as each is
transformed
using
a
different
value
for
the
Box-Cox
transformation
parameter.
Instead,
the
following
focuses
on
differences
in
sign
and
significance across regions.
For
example,
Aquaculture
index
is
significant
and
positive
for
the
Damariscotta River region. That is, after controlling for structural and
neighborhood characteristics, and attitudes surrounding aquaculture, houses
with “more” coastal aquaculture command a higher price on average,
suggesting that aquaculture may be viewed as an amenity in the region. In
Penobscot Bay, we find the opposite result, with coastal aquaculture lowering
sales prices – coastal aquaculture may be viewed as a disamenity in this
housing market. This pattern is consistent with the information on concerns
raised at public lease hearings (Table 2). Note that, on average, there are
more than twice as many concerns raised about the development of
mariculture and how it affects riparian homeowners and coastal users in
Penobscot Bay.
Given the evidence that aquaculture capitalizes into residential property
values in two study areas, we explore the relative magnitude of these
marginal impacts – are they large or small in these study areas? That is,
while statistically significant, are these results economically significant? A
comparison of the marginal effects for the components of the Aquaculture
index will provide insight into this,and answer our second research question.
Based on the design of this index, there are three margins of interest: the
density effect, the acreage effect, and the proximity effect (equation 3). The
density effect measures the marginal impact of an additional lease sited near
a home, holding the number of acres of aquaculture production and
proximity constant.8 The acreage effect and proximity effect have similar
interpretations. Given the Box-Cox transformation and the form of the
Aquaculture index, these margins are nonlinear and depend on a combination
of
parameter
estimates,
the
Box-Cox
transformation
parameter,
and,
importantly, the sales price of a home; this generates heterogeneity in the
marginal impacts across the sample. We calculate margins at the house-level
using
sample
values.
Further,
we
restrict
attention
to
houses
with
aquaculture sited within a 2-mile buffer zone of the home. Given skewness in
prices (evident from the estimates of the Box-Cox transformation parameter),
8 This is similar in theme to the expected contribution of an additional bedroom, ceteris paribus,
to the sales price of home, which holds square footage of the home constant.
Evans et al.
A Hedonic Analysis of the Impact of Marine Aquaculture 259
https://doi.org/10.1017/age.2017.19 Published online by Cambridge University Press


===== PAGE 20 (score 28, numeric-paren 12, table-words 3, var-words 8) =====

Table 4. Median Marginal Willingness-to-Pay (MWTP) ($) for coastal aquaculture by study area and model. Estimates
calculated from the subset of observations that contained aquaculture within a 2-mile buffer zone. Density depicts the
margin for an additional lease holding the number of acres and distance of leases to houses constant – increasing the
density of leases contained in the existing acreage. Acreage depicts the margin for an additional acre of aquaculture,
holding the number of leases and distance from housing constant. Proximity depicts the margin for moving 1 meter closer
to aquaculture holding the number of leases and acres constant. One-tailed p-values simulated using 100,000 draws from
the parameters estimates (Table 3).
Study area
Model
Density
Acreage
Proximity
Casco Bay
Base
39.64
54.71
0.01
Alternate
122.62
228.04
0.05
Damariscottta
Base
1.69***
93.77***
0.01***
Alternate
3.59***
156.05***
0.01***
Penobscot Bay
Base
1,588.79**
704.57**
0.78***
Alternate
1,006.27***
637.75***
0.44***
Significance levels: *** p < 0.01, ** p < 0.05, and * p < 0.10
Evans et al.
A Hedonic Analysis of the Impact of Marine Aquaculture 261
https://doi.org/10.1017/age.2017.19 Published online by Cambridge University Press


===== PAGE 21 (score 13, numeric-paren 0, table-words 0, var-words 13) =====

government agencies and other invested stakeholders (e.g., NGOs) (Knapp
2012, Knapp and Rubino 2016). Researchers argue that we could reduce
these tensions through improved availability of interdisciplinary research and
communication of the positive benefits of aquaculture to these groups of
users (Bricknell and Langston 2013, NOAA 2016b). These challenges
highlight the importance of acceptance by coastal resource users for the
successful expansion of mariculture. Resource managers and regulators with
a deeper understanding of the impacts of marine development (e.g.,
mariculture) could design spatial plans that balance concerns across a suite
of users and promote this emerging industry. To this end, quantifying the
impacts of mariculture development on coastal residential property values
provides resource managers with valuable information for integrative marine
spatial planning.
This paper presents an important addition to the literature and provides one
of the first empirical analyses of the effect of marine aquaculture on coastal,
residential property values. This revealed-preference study complements the
stated-preference work on public perceptions toward aquaculture (Shafer,
Inglis, and Martin 2010, D’Anna and Murray 2015, Jodice et al. 2015). Further,
the design of our hedonic model allows us to address questions of direct
interest to policy makers and land managers regarding the development of
mariculture. Specifically, we measure not just the impact from proximity to
aquaculture, but also from the spatial configuration of its siting, e.g., the
density and acreage of leases across the local seascape. The results from this
work provide insights into the impact of decentralized expansion of marine
aquaculture. Finally, our results can also inform future exploratory analyses
of transcripts from aquaculture lease hearings. For example, an exploration of
the patterns of concerns raised at these hearings, and how they vary by lease,
lessee, and community characteristics could
be important for coastal
managers. This type of analysis could provide information about the efficacy
of public participation in the siting process, as well as provide insight into
designing a siting framework that improves social acceptance of mariculture.
While there is a rich body of work that has evaluated the impact of land-based
farming (Abeles-Allison and Connor 1990, Palmquist, Roka, and Vukina 1997,
Le Goffe 2000, Herriges, Secchi, and Babcock 2005, Ready and Abdalla 2005,
Kim and Goldsmith 2009), few studies have attempted to investigate the
impact of water-based farming on residential property values (Northern
Economics 2010, Jodice et al. 2015, Sudhakaran 2015). Our work fills this
hole in the research. Our results suggest wide variation in how marine
development of aquaculture impacts property values, both across and within
study areas. In Casco Bay, we fail to find statistical evidence of impacts from
marine aquaculture. Recall that this is an urban area for Maine, with an
active working waterfront for shipping, recreation, and commercial fishing. It
is possible that mariculture is insubstantial relative to these other coastal
uses. In the Damariscotta River region, we find statistically significant and
positive effects of the development of marine aquaculture. However, these
Agricultural and Resource Economics Review
262 August 2017
https://doi.org/10.1017/age.2017.19 Published online by Cambridge University Press