Content uploaded by Bridget K. Behe
Author content
All content in this area was uploaded by Bridget K. Behe
Content may be subject to copyright.
Content uploaded by Bridget K. Behe
Author content
All content in this area was uploaded by Bridget K. Behe
Content may be subject to copyright.
Content uploaded by Bridget K. Behe
Author content
All content in this area was uploaded by Bridget K. Behe
Content may be subject to copyright.
127J. Environ. Hort. 23(3):127–133. September 2005
Landscape Plant Material, Size, and Design Sophistication
Increase Perceived Home Value1
B. Behe2, J. Hardy2, S. Barton3, J. Brooker4, T. Fernandez2, C. Hall4, J. Hicks2,
R. Hinson5, P. Knight6, R. McNiel7, T. Page8, B. Rowe2, C. Safley9, and R. Schutzki2
Department of Horticulture, Michigan State University
East Lansing, MI 48824
Abstract
Little consumer research is available to help landscape design and installation businesses develop service marketing strategies. We
investigated the effect of three components of a landscape design on the perceived value of a home. This information would be useful
in marketing lawn and landscape services to prospective clients. Our objective was to provide a consumer perspective on the value of
the components in a ‘good’ landscape and determine which attributes of a landscape consumers valued most. Using conjoint design,
1323 volunteer participants in seven states viewed 16 photographs that depicted the front of a landscaped residence. Landscapes were
constructed using various levels of three attributes: plant material type, design sophistication, and plant size. Results showed that the
relative importance increased from plant material type to plant size to design sophistication. Across all seven markets, study participants
perceived that home value increased from 5% to 11% for homes with a good landscape.
Index words: conjoint analysis, consumer, marketing.
1Received for publication February 11, 2005; in revised form April 18, 2005.
Partial funding was provided by a grant from The Horticultural Research
Institute, 1000Vermont Ave., NW., Suite 360, Washington, DC, 20005.
The authors gratefully acknowledge the assistance of many colleagues in
the USDA multi-state S-290 project for their assistance in data collection
and manuscript revisions.
2Department of Horticulture, Michigan State University, East Lansing, MI
48824.
3Department of Plant and Soil Sciences, University of Delaware, Newark,
DE 19716.
4Department of Agricultural Economics, University of Tennessee, Knox-
ville, TN 37996.
5Department of Agricultural Economics and Agribusiness, Louisiana State
University, Baton Rouge, LA 70803.
6Coastal Research and Extension Center, Mississippi State University,
Poplarville, MS 39470.
7Department of Horticulture, University of Kentucky, Lexington KY 40546-
0091.
8Department of Marketing and Supply Chain Management, Michigan State
University, East Lansing, MI 48824.
9Department of Agricultural and Resource Economics, North Carolina State
University, Raleigh, NC 27695.
Significance to the Nursery Industry
While realtors and realty-related Websites can provide in-
formation to homeowners regarding the amount of a home
improvement that can be recovered upon the sale of a home,
there is insufficient information available as to the value of
landscaping as a home improvement across multiple mar-
kets. Specifically, how much do plant size, plant material
type, and design sophistication in a landscape affect the per-
ceived value of a home? If these questions were answered,
landscape professionals would have more information to
present to potential clients on the soundness of an invest-
ment in landscape services. Our objective was to provide a
consumer perspective on the value of the components in a
‘good’ landscape and determine which landscape attributes
consumers valued most, and to see if that was consistent
across markets. Overall, participants valued the landscape
design sophistication most. Landscape professionals should
emphasize to customers that island beds and curved bed lines
add to the perceived value of a home. Results showed that in
the seven states, plant material type was relatively least im-
portant. Plant size was of intermediate importance in all
markets analyzed here, but was most important in results
previously published (18). Still, the largest affordable plant
size should be used as it consistently provided a higher per-
ceived value in all markets. Landscape professionals should
indicate the type of plants used in a landscape, but realize
that potential clients will not likely value this as much as
design sophistication and plant size. Professional landscape
mangers can show potential clients that a good landscape
adds 5 to 11% to the perceived value of a home.
Introduction
Gardening leads the nation as the leisure activity in which
more Americans participate than any other (6). For garden-
ers, a general profile emerged from the literature showing
individuals who were highly involved with gardening or those
who spent large amounts of money on gardening were more
likely to be female and older, wealthier, and more educated
than average (4, 18). Higher average home value was an-
other important characteristic of an avid gardener (18). In
another study, primary purchasers of nursery plants ranged
in age from 25 to 44 years, with a higher than average in-
come and level of education (17). Yet, gardening can be an
activity in which participants are active or passive. Active
participants may choose to ‘do-it-yourself’ while passive
participants may elect to have someone provide services of
landscape design, installation, and maintenance.
The profile of the ‘typical’ landscape service buyer can be
gleaned from The National Gardening Survey (9). The aver-
age purchasers of landscape installations were 50 years of
age or older, college educated, with a household income of
$75,000 or more. The American Nursery and Landscape As-
sociation conducted a consumer study, segmenting the mar-
ket of landscape consumers into three categories: Partners,
Pro-Purchasers, and Hardcore Do-it-Yourselfers (2). Partners
preferred to work with landscape professionals on an ad-hoc
basis. The Hardcore Do-It-Yourselfers wanted to do most of
the work themselves and were not likely clients for profes-
sional design and installation. Pro-Purchasers were the larg-
est group of customers and preferred no interaction with their
128 J. Environ. Hort. 23(3):127–133. September 2005
landscape jobs. This group consisted of single-family house-
holds with annual household incomes exceeding $100,000.
Results showed that 22% of the Pro-Purchasers (2.8 million
households) were the most likely group to purchase profes-
sional landscape services. Yet, there is little understanding
of what these consumers want from a home landscape and
what features of this highly personalized product are impor-
tant to them.
The market for landscape design and installation is con-
siderably smaller than the market for plants and garden cen-
ter products. Butterfield (9) reported that less than 2.5% of
American homes used landscape design and 3.3% purchased
installation services in 2003, percentages that have changed
little in the prior decade. Khatamian (25) reported that 9.5%
of those surveyed at garden centers, and flower, lawn and
garden shows, preferred to contract with a professional in-
stallation firm. With approximately 109 million households
in the United States in 2003, approximately two million may
be potential targets for services. In the most recently avail-
able statistics (1997), there were 37,853 landscape contrac-
tors (SIC 0781-02) and 9826 landscape designers (SIC 0781-
03) (10). The relatively small size of this market and prolif-
eration of firms willing to do this work suggests that much
competition exists.
Motivations for improving a home vary widely. For the
resident of a newly constructed home, the exterior landscap-
ing may serve as a ‘frame’ to the home, enhancing its aes-
thetics. For the resident desiring to sell a home, landscaping
could enhance market value. Realtors have data showing that
the value of a home is enhanced with the renovation of a
kitchen, bath, or bedroom, or with the addition of a deck or
patio. Realtors can also convey the return on investment for
renovations and additions. Recent estimates of recovered
costs show that 94% of a mid-priced bathroom addition can
be recovered, 74% of the cost of window replacement can be
recovered, and 79% of the cost of a family room addition
can be recovered (12).
A Weyerhaeuser publication (3) estimated that landscap-
ing added approximately 15% from a homeowner perspec-
tive, but only 7.3% by real estate appraisers. Neither data
nor methodology were reported for arriving at the percent-
ages. Nearly all other studies of landscape value investigated
one attribute in a single market. Prior work by Hardy et al
(18) added to the body of knowledge by showing that par-
ticipants at a flower and garden show perceived the well-
landscaped residence to increase in value by 12.7%. Plant
size was most important in the conjoint analysis, followed
by design sophistication, and plant material. The 12.7% in-
crease in home value was also consistent with several other
reports (20, 21, 22, 28, 29, 31, 32). Given that most publica-
tions reported a 10–15% increase in perceived home value,
we hypothesized that a consistent and similar increase would
be seen across in several markets. Using identical materials
and methodology from the Hardy et al. study (18), we ex-
pected that across all venues included, we would observe a
12% increase in perceived home value by the consumer.
Materials and Methods
Generation of plans and photographs. A researcher pho-
tographed a two-story, newly-built home in a Delaware sub-
urb, as the test home. Given the researchers’ specifications, a
commercially employed landscape architect prepared 16 flat
plans. The designer was given the factor level parameters
and definitions for each plan, and also received a set of guide-
lines that included incorporating only plants whose hardi-
ness extended from USDA plant hardiness zones 4–7. Re-
searchers did instruct the landscape designer to select plant
material common across all hardiness zones to be included
in the study. Research on cross-national preferences for tree
canopy form showed that preferences for tree form were as-
sociated with the tree shape reported as most common in the
geographical area where the respondents grew up (32). The
architect was instructed to use only common plants that were
readily available in all growing zones in the study. Computer
generated color perspective images of the home and land-
scaping were prepared from each flat plan [Adobe PhotoShop
version 5.0] (1). Each photograph depicted the home and land-
scaping as viewed from the street.
Generation of orthogonal design, factor level definitions
and conjoint analysis. Using the methodology of Hardy et
al. (18), the respondent’s overall preference for a particular
landscape was defined as the dollar value assigned to the
landscaped home by the respondent. Conjoint analysis was
previously used in horticultural studies to assess relative
importance of attributes and predict consumer demand for
blue geraniums (6) and colored bell peppers (15); investi-
gate consumer preference for packaging of edible flowers
(24); and analyze consumer preference for retail evergreen
shrubs (13). Conjoint analysis defines overall preference for
a particular product, in this case a landscape, as the sum of
the part-worths (also termed utilities) for each factor level
(7, 16, 19). By definition, the sum of the part-worths is analo-
gous to the value added to the home by the landscape as pre-
dicted by the conjoint analysis procedure. Researchers chose
plant size, diversity of plant material (type), and design so-
phistication as the factors that most comprehensively describe
all landscape attributes. In this model, the preference for plant
size plus preference for design sophistication plus preference
for the type of plant material used resulted in the overall pref-
erence (measured in dollars) for a particular landscape.
For each factor, we determined a measurable, hierarchical
set of levels for each variable. The plant size levels were
defined as small, medium, or large. Small was the smallest
available size for the product, large was the largest available
size for the product, and medium was the intermediate size
between large and small. Design sophistication levels were:
(1) foundation planting only, (2) foundation planting with
one large, oblong island planting and one or two single speci-
men or shade trees in the lawn, or (3) a foundation planting
with adjoining beds and two or three large island plantings,
all incorporating curved bedlines. The plant material types
were: (1) evergreen only, (2) evergreen and deciduous plants,
(3) evergreen and deciduous plants with 20% of the visual
area of the landscape beds planted in annual or perennial
color, or (4) evergreen and deciduous plants, 20% annual or
perennial color, and the addition of a colored brick sidewalk
entrance.
While all 36 (3 size × 3 design × 4 material levels) pos-
sible combinations of factor levels could have been used for
full profile conjoint analysis, researchers chose to reduce
respondent fatigue by minimizing the number of photographs
evaluated. By using a partial factorial design, the number of
photographs required to maintain orthogonality was reduced
from 36 to 16. Conjoint Designer [version 3.0 produced by
Bretton-Clark, 1992] (7) was used to generate the list of 16
129J. Environ. Hort. 23(3):127–133. September 2005
stimuli. Conjoint and other statistical analyses were facili-
tated using SPSS 10.0 (25).
Survey administration and instrument. Churchill and
Iancobucci (11) classify survey type by method of adminis-
tration or data collection. While variations are possible, gen-
erally surveys are classified as telephone, mail, electronic
(email or Internet) and intercept. The latter refers to recruit-
ing participants in a public setting, such as mall or other pub-
lic venue. Lysaler (27) noted that mall-intercept interview-
ing has increased substantially in use in the 1990s. While
there is no one best method, each has applications and ad-
vantages (11). Yet, several studies that use identical ques-
tions by vary data collection indicate that respondents’ de-
mographic characteristics and responses to questions are simi-
lar (8, 11, 14, 33). Several published studies used an inter-
cept technique for data collection to recruit potential respon-
dents from a variety of plant-related venues (5, 6, 10, 13, 23,
26, 30).
Between April and July of 1999, surveys were adminis-
tered in seven markets throughout the eastern and central
United States: Delaware (DE), Kentucky (KY), Louisiana
(LA), Mississippi (MS), North Carolina (NC), South Caro-
lina (SC), and Texas (TX). Responses were collected at con-
sumer home and garden shows. Responses from South Caro-
lina were collected both at local garden center and a zoo. For
each survey site, the same protocol was used for survey ma-
terials and the recruitment of participants. Multiple partici-
pants could self-administer the questionnaire simultaneously.
Researchers asked every second or third individual who
passed the booth and expressed some interest (by making
eye contact) if they would be willing to participate. This var-
ied by time of day and pedestrian traffic at each location.
Researchers estimate mall intercept surveys achieve approxi-
mately a 29% response rate, but this would be higher for
surveys in which the potential respondent has a high interest
(11). The sample size target was approximately 160 responses
from each state, determined by having at least 10 responses
per photograph included in the study. Given the anticipated
standard deviation in change in perceived home value, a
sample size of 160 was expected to provide sufficient re-
sponses to conduct the planned statistical analyses. The
sample was a convenience sample, drawn to reflect perspec-
tives of visitors to home and garden shows in seven markets
who likely had an elevated interest in enhancing the value of
their home.
Visitors were recruited to participate in the survey as they
passed the display table on which the 16 photographs were
displayed. Participants were asked to examine a photograph
of the survey home with only a lawn and a straight poured
cement walk and driveway (Fig. 1). They were told the home
value, as estimated by local realtors in each market, and also
the county in which the home was hypothetically located.
Researchers also stipulated in writing that the home was in a
subdivision with similar new homes, and that the home was
a 4-bedroom, 2-1/2 bathroom two-story structure located on
a half-acre lot (approximately 100 ft by 200 ft). Participants
were asked to look at the 16 additional photographs. Consid-
ering the price of the home assigned by realtors and the land-
scaping and features around the homes, they were asked to
assign a value to each home. The second part of the survey
asked respondents to provide demographic information about
themselves, their family, home, landscape and landscape ser-
vice usage. All questionnaires were identical in format.
Statistical methods. The overall conjoint analysis results
were defined as the mean of the conjoint results generated
for each respondent. Since this method of analysis allows for
calculation of variance, typical tests of significance were
employed using SPSS 10.0 (34).
Results and Discussion
Demographic profile of respondents. Female respondents
outnumbered male respondents in all states except North
Carolina (Table 1). North Carolina had the youngest respon-
dents with a mean average age of 40.4 years while Missis-
Fig. 1. Base home with no landscape.
130 J. Environ. Hort. 23(3):127–133. September 2005
sippi and Kentucky had the oldest at a mean age of 48.4 years
(Table 1). Although over 99% of respondents in the Dela-
ware sample stated they had finished at least 12 years of for-
mal education, they were the least educated group, with only
14.8% of the respondents completing formal education be-
yond high school. Mean household size varied from 3.3
people in Louisiana to 2.6 people in Kentucky. Average per
capita income ranged from $28,198 to $34,361 with no dif-
ferences among states.
The 2003 National Gardening Survey reported 84 million
households participated in lawn and gardening activities (9).
Those more likely to participate in lawn and gardening ac-
tivities were married, ages 35–54 years, college educated,
have children, employed full-time with an annual income of
>$35,000. This sample averaged 57% female respondents,
whereas the National Gardening Survey sample had 52%
female respondents. The average age was 45 years in this
study which was the midpoint of the age range from the Na-
tional Gardening Survey. The average level of formal educa-
tion completed by the respondents in this sample was nearly
16 years, equivalent to a college education. Forty-seven per-
cent of the National Gardening Survey sample had completed
some college while 43% had earned a college diploma. There
were similar parallels in household size and income. Statisti-
cal tests between the two samples were not possible without
the standard deviation in the National Gardening sample, but
many similarities were evident among them.
Time and money expenditures on lawn and garden. The
mean number of hours respondents stated they spent on their
lawn and yard during a typical summer week in 1998 ranged
from 3.9 hours in Delaware to 9.8 hours in Mississippi (Table
1). In several individual cases in each state, respondents stated
that they work over 30 hours per week on their lawn and
garden. With these extremes, the median hours per week was
a better indicator of the average for each sample. The me-
dian hours ranged from 2.0 hours in Delaware to 9.8 hours in
Mississippi. For the remaining states (KY, LA, NC, SC, TX)
the median time spent on lawn and garden ranged from 5 to
6 hours.
The National Gardening Association (9) reported average
lawn and garden expenditures in 2003 were $457, up slightly
from $452 in 1998. Respondents to the study spent from $445
(Mississippi) to $2327 (Kentucky) with a broad range of re-
sponses, yet there were no statistically significant differences
among them.
Model fit. Because the base price assigned to each home
reflected the local housing market, a direct comparison of
the dollar values predicted for various combinations of land-
scapes was inappropriate. We limited analysis to relative
importance of factors and to the percentage increases in value
for various factor levels over the base home value.
Each of the individual conjoint models for each of the seven
states explained at least 94% of the variance in respondent
answers. In the context presented, plant size, design sophis-
tication and plant material type were good indicators of the
perceived change in home value. Thus, across all seven states,
the experimental stimuli yielded a similar range of respon-
dent answers.
Relative importance of attributes. For each of the seven
states, relative importance of factors increased from plant
material type to plant size to design sophistication (Table 2).
Design sophistication was the most important landscape fac-
tor in these seven states, accounting for 40 to 45% of the
value added to the home. Paired sample t-tests indicated that
for each state, there was one significant difference between
the relative importance of the three factors in each location.
Louisiana respondents valued sophistication less than North
Table 1. Demographic characteristics showing means ± standard deviation of survey participants from seven states evaluating landscaped homes.
Survey Percent Age Education Persons in Per capita Hours spent Gardening
location (state) nfemale (years) (years) household income ($) in the garden expenditures ($)
Delaware 174 52.6 42.8 ± 12.4 14.7 ± 2.21 3.1 ± 1.4 28,198 ± 17,156 3.9 ± 6.8 771 ± 1,960
Kentucky 238 50.6 48.4 ± 11.0 16.0 ± 2.63 2.6 ± 1.1 34,361 ± 19,301 8.8 ± 10.3 2,327 ± 5,625
Louisiana 191 63.2 41.7 ± 7.7 16.0 ± 2.57 3.3 ± 1.4 28,231 ± 19,513 8.0 ± 12.8 800 ± ,997
Mississippi 199 65.7 48.4 ± 11.3 15.6 ± 2.58 2.7 ± 1.3 32,895 ± 18,632 9.8 ± 14.6 445 ± ,984
North Carolina 182 46.2 40.4 ± 11.4 15.7 ± 2.18 2.7 ± 1.1 29,480 ± 21,314 4.3 ± 5.7 1,508 ± 2,760
South Carolina 164 63.4 47.0 ± 12.8 15.8 ± 2.41 2.9 ± 1.3 29,280 ± 16,328 5.6 ± 5.5 1,861 ± 8,683
Texas 175 50.3 45.5 ± 13.0 15.9 ± 2.67 2.7 ± 1.1 28,906 ± 17,248 5.2 ± 5.2 902 ± 1,233
Total 1323 57.3 45.8 ± 12.1 15.8 ± 2.51 2.8 ± 1.2 31,903 ± 19,586 6.9 ± 9.2 1,409 ± 4,640
Table 2. Relative importance percentages of three landscape attributes for participants from seven states and the average percent increase over base
home value for the highest level of each factor in the landscape design.
Plant Plant Design Average percent increase
State material size sophistication over base home value
Delaware 24.8 30.6 44.6 6.79%
Kentucky 20.8 36.4 42.8 8.74%
Louisiana 23.4 32.9 43.7 5.54%
Mississippi 23.9 34.1 42.0 10.76%
North Carolina 24.4 34.5 41.2 7.06%
South Carolina 23.3 34.1 42.6 11.36%
Texas 21.0 39.0 40.1 10.16%
Average (mean) 22.4 35.9 41.7 9.0%
131J. Environ. Hort. 23(3):127–133. September 2005
Carolina (p = 0.003). There were no significant differences
between any other pairs of states for design sophistication.
Relative importance of plant size was of intermediate im-
portance between design sophistication and plant material,
and ranged from 39% in Texas to 30.6% in the Delaware
sample. In paired t-tests, North Carolina, South Carolina, and
Mississippi were similar (NC to SC p = 0. 191; NC to MS
p = 0. 523; MS to SC p = 0.455) with res pect to the relative
importance of plant size. Additionally, Louisiana was simi-
lar to Mississippi (p = 0.184) and North Carolina (p = 0.567).
Texas was similar to Kentucky (p = 0.188). Kentucky was
similar to South Carolina (p = 0.134). Generally, plant size
importance appeared to decrease as plant growing zone num-
ber increased.
For respondents from all states, the diversity of plant ma-
terial type installed contributed the least to the value added
to the home landscape. In all states, plant material type con-
tributed 16 to 22% less to the added home value than did
design sophistication. The relative importance of plant ma-
terial in Delaware and North Carolina was similar. Louisi-
ana was different than North Carolina (p = 0.042). The re-
maining comparisons of Delaware and Louisiana to any other
states showed significant differences (p < 0. 01). Generally,
plant material type was perceived as having half the impor-
tance of design sophistication and less important than plant
size.
Changes in perceived home value. Across all states, smaller
plant sizes reduced perceived home values (as indicated by
utility scores), while larger plant sizes increased perceived
home values (Table 3). The medium-sized plants produced
virtually no change to existing home values. The greatest
increase in perceived value due to plant size was seen in North
Carolina, where the base home value was $220,000. When
the smallest plants were used, perceived home values de-
creased by 2.3%. When the largest plant sizes were used,
perceived home value increased by 2.8%. Louisiana showed
the smallest difference in perceived home value due to plant
size. The estimated home value for Louisiana was $176,000.
When small plants were used, perceived home value de-
creased 1.2%, while the use of the largest plant sizes increased
values by 0.6%. In northern climates, plant size will increase
at a slower rate than in more southern climates.
For respondents across all states, the simplest design (foun-
dation-only plantings) decreased perceived home value by
an average of 2.1%, while the most sophisticated landscape
design increased home values by an average of 1.9% (Table
3). Island plantings added to foundation-only beds had virtu-
ally no effect on perceived home value. Texas showed the
greatest range in perceived home value due to design sophis-
tication. The estimated base home value for Texas was
$125,000. Foundation-only plantings decreased perceived
home value by 3.3%, while sophisticated designs increased
perceived home values by 2.6%. Louisiana had the smallest
variation between foundation and sophisticated plantings.
Foundation plantings decreased home values by 0.9%, while
sophisticated plantings increased home values by 0.8%.
Generally, the additional diversity of plant material in-
creased perceived home value (Table 3). Data from respon-
dents across all states found that material levels 1 and 2 de-
creased home values, while material level 3 increased per-
ceived home values somewhat. The use of materials described
in level 4 increased perceived home values most. Missis-
sippi, with a base home value of $150,000, showed the larg-
est fluctuation between material levels 1 and 2 when com-
pared to designs containing all four material levels. Material
levels 1 and 2 decreased home values by 1.0%, while levels
3 and 4 increased perceived home values by 1.7%. Louisi-
Table 3. Utility score and percentage change over base home values for three levels of landscape plant size, design sophistication, and four levels of
diversity of landscape material in seven locations.
Delaware Kentucky Louisiana Mississippi North Carolina South Carolina Texas
Utility %Utility %Utility %Utility %Utility %Utility %Utility %
Base home value $180,000 $140,000 $176,000 $150,000 $220,000 $150,000 $125,000
Plant sizez
Small –1971.7 –1.1 –2210.2 –1.6 –2094.8 –1.2 –2738.4 –1.8 –2698.6 –1.2 –2785.9 –1.9 –2442.6 –2.0
Medium 68.3 0.0 26.5 0.0 1097.9 0.6 –530.4 –0.4 –433.5 –0.2 –187.0 –0.1 615.2 0.5
Large 1903.4 1.1 2183.8 1.6 996.9 0.6 3268.8 2.2 3132.1 1.4 2973.0 2.0 1827.4 1.5
Design sophisticationy
Foundation –2792.1 –1.6 –2732.3 –2.0 –1561.3 –0.9 –3872.1 –2.6 –3655.5 –1.7 –3901.0 –2.6 –4105.0 –3.3
Island –57.6 –0.0 295.7 0.2 98.2 0.1 345.4 0.2 348.0 0.2 598.3 0.4 856.5 0.7
Sophisticated 2849.7 1.6 2436.6 1.7 1463.0 0.8 3526.7 2.4 3307.5 1.5 3302.8 2.2 3248.6 2.6
Diversity of landscape materialx
Evergreen –1199.1 –0.7 –1101.6 –0.8 –1340.5 –0.8 –1752.7 –1.2 –1663.4 –0.8 –1371.6 –0.9 –229.8 –0.2
+ Deciduous –850.3 –0.5 –722.6 –0.5 –1186.8 –0.7 –1449.6 –1.0 –1421.0 –0.7 –1562.3 –1.0 –2171.7 –1.7
+ 20% Color 115.8 0.1 414.5 0.3 1666.1 1.0 728.4 0.5 728.4 0.3 555.1 0.4 1816.3 1.5
+ Hardscape 1933.6 1.1 1409.6 1.0 861.1 0.1 2473.9 1.7 2473.9 1.1 2378.8 1.6 585.3 0.5
zPlant size was defined as: (1) small = smallest plant size commercially available for the product, (2) medium = intermediate plant size between small and large
which was commercially available and (3) large = largest plant size commercially available.
yDesign sophistication was defined as (1) Foundation = foundation planting only, (2) Island = foundation planting with one large, oblong island planting and one
or two single specimen or shade trees in the lawn and (3) Sophisticated = a foundation planting with adjoining beds and two or three large island plantings, all
incorporating curved bedlines.
xThe plant material types were defined as : (1) Evergreen = evergreen only, (2) + Deciduous = evergreen and deciduous plants, (3) + 20% Color = evergreen and
deciduous plants with 20% of the visual area of the landscape beds planted in annual or perennial color, and (4) + Hardscape = evergreen and deciduous plants,
20% annual or perennial color, and the addition of a colored brick sidewalk entrance.
132 J. Environ. Hort. 23(3):127–133. September 2005
ana showed the smallest variation in perceived home value;
material levels 1 and 2 decreased home value by 0.7%, while
the addition of all four levels increased perceived home value
by 0.1%. For Louisiana and Texas, the addition of material
levels 3 and 4 to existing levels 1 and 2 increased perceived
value by 1.0 and 1.5% respectively. In these locations, the
addition of material level 4 decreased home values in com-
parison to the landscape containing only material levels 1, 2
and 3. However, the addition of material level 4 to existing
material levels 1–3 in the remaining states (DE, KY, MS,
NC and SC) increased base home values by an average of
1.2%.
Overall preference. All states shared the same most pre-
ferred landscape: a sophisticated design incorporating large
deciduous, evergreen, and annual color plants and colored
hardscape. The percent increase in home value from the least
valued to the most valued varied among states from 5.5% in
South Carolina to 11.4% in Mississippi (Table 2). In all states,
the ranking from least percentage added to most percentage
added followed the same pattern for percent increase from
the least favored to most favored, percent increase of the most
favored over the predicted base and percent decrease of the
least favored over the predicted base. Additionally, the order
of the state rankings of the percent increase from the least
favored to the most favored bore no resemblance to the rank
order of the base price of the house. In other words, large
percent increases in home value were not associated with
larger base prices.
Why were these results somewhat different than Hardy
(13) found, using identical methodology? The first differ-
ence was the venue. Michigan participants were the oldest
and had the highest income when compared to other respon-
dents. Yet they were not dissimilar in age and income to par-
ticipants in all states (data not shown). Michigan respondents
were similar to respondents from other states in terms of
number of persons in the household, and hours and dollars
spent in the garden. Michigan respondents did place a higher
relative value on plant size than respondents from all other
markets. A flower show venue was used to collect data in
Michigan, not a home and garden show. Participants to the
former venue may have added even more value for having a
colorful landscape than the participants from other venues.
The second reason for some discrepancy may be USDA har-
diness zone. The authors speculate that perhaps in colder
hardiness zones, where plants grow more slowly, gardeners
value plant size more and perceive that more colorful plants
add greater value. Respondents from warmer hardiness zones,
where plant material has a longer growing season and grows
more quickly, may value landscape design sophistication
more than plant size. Design sophistication was the second
most important attribute to Michigan participants. Respon-
dents from all venues agreed that plant material used in the
landscape was relatively the least important factor.
We found that in all markets, consumers preferred the larg-
est, most sophisticated, and colorful landscape design. The
sophisticated planting category consisting of a foundation
planting with adjoining beds and two or three large island
plantings, all incorporating curved bedlines, increased home
values by an average of 1.8%. This indicates that consumers
in the states tested could increase home values by $2,375–
$3,648 depending upon the initial base home value and cost
of materials and installation of the plants.
Results of this study indicate that a ‘good’ landscape adds,
depending on region of the country, anywhere from 6 to 11%
to the base value of the home. The landscape attributes that
contributed most to the increase in perceived home value
were, in order, design sophistication, plant size, and plant
material type. Clearly, the investment in a good landscape
can be recovered and increase the perceived value of a home.
The minimalist landscapes, with small plant size and little
sophistication, even detracted from the perceived value of
the home. The landscape company manager now has con-
crete data to show that a good landscape adds to the value of
a home, and is a home improvement that will increase per-
ceived home value and, unlike most home improvements,
appreciate over time.
Literature Cited
1. Adobe Systems, Inc. Adobe PhotoShop version 5.0. Adobe Systems,
Inc. San Jose, CA.
2. American Nursery and Landscape Association. 2002. Understanding
the Landscape Consumer. American Nursery and Landscape Association.
Washington, DC.
3. Anonymous. 1986. The Value of Landscaping. Weyerhaeuser Nursery
Products Division. Tacoma, WA.
4. Barton, S., J. Brooker, C. Hall, and S. Turner. 1999. Review of
customer preference research in the nursery and landscape industry. J.
Environ. Hort. 16:118–124.
5. Behe, Bridget K., Elizabeth C. Moore, Arthur C. Cameron, and Forest
S. Carter. 2003. Consumer perceptions for and uses and perceptions of
selected flowering perennial plants. HortScience 38:460–464.
6. Behe, B., R. Nelson, S. Barton, C. Hall, C. Safley, and S. Turner.
1999. Consumer preferences for geranium flower color, leaf variegation and
price. HortScience 34:740–742.
7. Bretton-Clark. 1992. Conjoint Designer version 3. Bretton-Clark,
Morristown, NJ.
8. Bush, A.J. and A. Parasuraman. 1985. Mall intercept versus telephone-
interviewing environment. J. Advertising Res. 25:36–44.
9. Butterfield, B.W. 2004. National Gardening Survey 2003. The
National Gardening Association, Inc. Burlington, VT.
10. Census of Retail Trade, 1997. http://www.census.gov/epcd/www/
econ97.html accessed April 18, 2005.
11. Churchill, G.A. and D. Iacobucci. 2005. Marketing Research:
Methodological Foundations. Ninth Edition. South-Western, Mason, OH.
12. Cory, J. 2003. 2002 Cost vs. Value Report. Remodeling Online.
Accessed September 23, 2003. http://remodelingonline.yellowbrix.com/
pages/remodelingonline/Story.nsp?story_id=1000027497.
13. DeBossu, A. 1988. What do people want to buy? Amer. Nurseryman.
May 1, 1988. 91–96.
14. Denstadli, J.M. 2000. Analyzing Air Travel: A Comparison of different
survey methods and data collection procedures. J. Travel Res. 39:4–11.
15. Frank, C., E. Simonne, B. Behe, and A. Simonne. July 2001.
Consumer preferences for color, price, and Vitamin C content of bell peppers.
HortScience 36:795–800.
16. Gaasbeck, A. and V. Bouwman. 1991. Conjoint analysis in market
research for horticultural products. Acta Hort.: Hort. Economics and Mktg.
295:121–125.
17. Gineo, M. 1988. Nursery marketing can be improved. J. Environ Hort.
6:72–75.
18. Hardy, J., B. Behe, S. Barton, T. Page, R. Schutzki, K. Muzii, R.T.
Fernandez, M.T. Haque, J. Brooker, C. Hall, R. Hinson, P. Knight, R. McNiel,
D.B. Rowe, and C. Safley. 2000. Consumers preferences for plant size, type
of plant material and design sophistication in residential landscaping. J.
Environ. Hort. 18:224–230.
19. Hartigan, J.A. 1975. Clustering Algorithms. John Wiley and Sons.
New York, NY.
133J. Environ. Hort. 23(3):127–133. September 2005
20. Henry, M. 1999. Landscape quality and the price of single family
houses: further evidence from home sales in Greenville, South Carolina. J.
Environ. Hort. 17:25–30.
21. Henry, M. 1994. The contribution of landscaping to the price of single
family homes: A study of home sales in Greenville, South Carolina. J.
Environ. Hort. 12:65–70.
22. Kalmbach, K. and J. Kielbaso. 1979. Resident attitudes toward
selected characteristics of street tree plantings. J. Arboriculture 5:124–129.
23. Kelley, Kathleen M., Bridget K. Behe, John A. Biernbaum, and
Kenneth L. Poff. 2002. Combinations of colors and species of containerized
edible flowers: Effect on consumer preferences. HortScience 37:218–221.
24. Kelley, Kathleen M., Bridget K. Behe, John A. Biernbaum, and
Kenneth L. Poff. 2001. Consumer preference for edible flower color,
container size, and price. HortScience 36:801–804.
25. Khatamian, H. and A. Stevens. 1994. Consumer marketing preferences
for nursery stock. J. Environ. Hort. 12:47–50.
26. Klingeman, W.E., D.B. Eastwood, J.R. Brooker, C.R. Hall, B.K. Behe,
and P.R. Knight. 2004. Consumer survey identifies plant management
awareness and added value of dogwood powdery mildew resistance.
HortTechnology 14:275–282.
27. Lysaker, R.L. 1989. Data collection methods in the U.S. J. Market
Research Soc. 31:477–489.
28. Morales, D. 1980. The contribution of trees to residential property
value. J. Arboriculture 6:305–308.
29. Nasar, J. 1983. Adult viewers’ preference in residential scenes: a study
of the relationship of environmental attributes to preference. Environ. and
Behavior 15:589–614.
30. Oppenheim, P. 2000. Segmentation and target marketing in a floral
market. Acta Hort. (ISHS) 536:529–536.
31. Orland, B., J. Vinning, and A. Ebreo. 1992. The effect of shade trees
on perceived value of residential property. Environ. and Behavior 24:298–
325.
32. Sommer, R. 1997. Further cross-national studies of tree form
preference. Ecological Psychology 9:153–160.
33. Spooner, C. and B. Flaherty. 1993. Comparison of three data collection
methodologies for the study of young illicit drug users. Australian J. of Public
Health 17:195–203.
34. SPSS Inc. 1998. SPSS Base 8.0 for Windows user’s guide. SPSS Inc.
Chicago, IL. SPSS Inc. 1997. SPSS Conjoint 8.0. SPSS Inc. Chicago, IL.
