Content uploaded by William C. Sullivan
Author content
All content in this area was uploaded by William C. Sullivan on Apr 09, 2021
Content may be subject to copyright.
Content uploaded by William C. Sullivan
Author content
All content in this area was uploaded by William C. Sullivan on Sep 30, 2014
Content may be subject to copyright.
Environment and Behavior
1 –24
© 2014 SAGE Publications
Reprints and permissions:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/0013916513518064
eab.sagepub.com
Article
Social Life Under Cover:
Tree Canopy and Social
Capital in Baltimore,
Maryland
Meghan T. Holtan1, Susan L. Dieterlen1, and
William C. Sullivan2
Abstract
To what extent does the density of the tree cover in a city relate to the
amount of social capital among neighbors? To address this question, we
linked social survey data (N = 361) from the Baltimore Ecosystem Study with
socioeconomic, urban form, and green space data at the census block group
level using a geographic information system. We found a systematically positive
relationship between the density of urban tree canopy at the neighborhood
block group level and the amount of social capital at the individual level (r =
.241, p < .01). Multiple regression analyses showed that tree canopy added
a 22.72% increase in explanatory power to the model for social capital.
This research adds a new variable—neighborhood tree canopy—to the
typologies of green space that affect human social connection. Trees are a
relatively inexpensive and easy intervention to enhance the strength of social
ties among neighbors.
Keywords
social capital, neighborhoods, tree canopy, ecosystem services, green space
1State University of New York, College of Environmental Science and Forestry, Syracuse, USA
2University of Illinois at Urbana Champaign, USA
Corresponding Author:
Meghan Holtan, State University of New York, College of Environmental Science and
Forestry, 441 West Fifth Avenue, Suite 202, Anchorage, Alaska, 99501, USA.
Email: holtan.meghan@gmail.com
518064EABXXX10.1177/0013916513518064Environment and BehaviorHoltan et al
research-article2014
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
2 Environment and Behavior
Introduction
More than 80% of the population in the United States lives in urban areas
(U.S. Census Bureau, 2010b). Despite our urbanism, we are more dependent
on natural ecosystems than ever before. In urban ecosystems, the urban forest
regulates climate, air, and water quality and supports the cultural life of resi-
dents through recreation and aesthetics. Yet substantially more research has
been conducted on the regulatory benefits of the urban forest or its effect on
property values than in understanding how the urban forest supports resi-
dents’ social relations (see, for example, Mansfield, Pattanayak, McDow,
McDonald, & Halpin, 2005; Mcpherson, Nowak, & Rowntree, 1994). The
Baltimore Ecosystem Study (BES) is a long-term study of human and urban
environment interactions that provided the publicly available data that com-
prises a majority of this study. In this study, we identified statistical relation-
ships between green space characteristics, including neighborhood tree
canopy, and an individual’s social capital, or the value inherent in neighbor-
hood social connections.
The Benefits of Green Space
The ambiguity of the term green space accounts in part for the diverse and at
times contradictory findings of benefits and correlations between green space
and human health and well-being (Lee & Maheswaran, 2011). Green space
has been defined broadly as “any vegetated land or water within or adjoining
an urban area” (Greenspace Scotland, 2008) and can include urban trees,
shrubs, and ground vegetation (Smardon, 1988). It has also been used as a
synonym for “public open space” (Lee & Maheswaran, 2011) and “green
neighborhood common spaces” (Kuo, Sullivan, Coley, & Brunson, 1998) and
measured through varying methods such as photo surveys (Kuo, Bacaicoa, &
Sullivan, 1998) and remote sensing (Groenewegen, van den Berg, de Vries,
& Verheij, 2006; Troy, Grove, & O’Neil-Dunne, 2012). In this study, we rec-
ognize and distinguish between the effects of personal green space such as
yards, the programmed green space of parks and the diffuse green space of
neighborhood tree canopy.
The benefits of green space for human well-being are well documented.
Views of nature have been related to decreased hospital patient recovery
times (Ulrich, 1984) and increased feelings of peace, escape from distraction,
and neighborhood satisfaction (R. Kaplan, 2001). A walk in the forest has
been shown to reduce levels of stress hormones, heart rate, and blood pres-
sure (Park, Tsunetsugu, Kasetani, Kagawa, & Miyazaki, 2010), and regulate
the effects of environmental stress (Yamaguchi, Deguchi, & Miyazaki, 2006).
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
Holtan et al 3
If people are drawn to green space for mental health benefits, they are likely
to meet other people seeking the same relaxation and restoration. Use medi-
ates the relationship between green space and social ties (Kuo, Sullivan, et
al., 1998). The resultant social interaction in and around green spaces can
also be measured using social capital indicators.
The study of the relationship between tree canopy and social capital falls
within the robust literature that establishes a connection between green space
and neighborhood social connections measured at the individual level. For
instance, a number of studies in Chicago public housing neighborhoods in
which residents were randomly assigned to apartments with varying levels of
nearby green spaces found positive relationships between green common
spaces and neighborhood social connections (Coley, Kuo, & Sullivan, 1997;
Kuo, Sullivan, et al., 1998; Sullivan, Kuo, & Depooter, 2004). These green
common spaces were also positively associated with sense of community and
social integration (Kuo, Sullivan, et al., 1998; Kweon, Sullivan, & Wiley,
1998). Because of the random assignment of residents in these neighbor-
hoods, the researchers argued that the findings were due to the increased use
of the green spaces, which then led to stronger social ties, rather than the
possibility that residents with more potential to form social ties had selected
apartments in greener areas.
In addition to the Chicago studies, two European national scale studies
found an independent positive relation between proximity to parks and col-
lective efficacy (Cohen, Inagami, & Finch, 2008) and between green space
and less loneliness and more social supports in the Netherlands (Maas, van
Dillen, Verheij, & Groenewegen, 2009).
These findings from Chicago and Europe were particularly salient for
young people, the elderly, and low-income people in urban environments,
pointing to the importance of green space for populations that are less mobile
and have less access to green space outside of their neighborhood. These
studies also demonstrate the need for a more fine-grained measurement of
green spaces over a large geographic area. In the Chicago studies, research
was restricted to vulnerable populations and green space was limited to what
little existed in the public housing developments. In the two national-level
studies, green space was operationalized as natural areas, urban green spaces,
and agricultural areas based on a 25 by 25-m land cover grid. (Maas, Verheij,
Groenewegen, de Vries, & Spreeuwenberg, 2006). This technique cannot
measure trees in yards or along the road. In contrast, the technology we lever-
aged in this study allowed us to quantify the density of greenery at the neigh-
borhood scale throughout the city.
Given that people react differently to various types of nature in neighbor-
hood settings (Coley et al., 1997; R. Kaplan & Kaplan, 1989), measuring tree
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
4 Environment and Behavior
cover may be especially relevant to outcomes such as social ties and social
capital. Tree canopy functions as a green web that has ecological, economic,
and, we would suppose, social benefits. In addition, planting street trees is a
relatively easy public intervention that individual neighbors can take respon-
sibility for and maintain (Jones, Davis, & Bradford, 2013). In contrast to
parks, which may be designed for certain demographic groups, tree canopy
benefits the neighborhood as a whole, providing shade from the summer sun,
habitat for birds, and flood mitigation during heavy rains (Mcpherson et al.,
1994).
Social Capital in the Neighborhood Context
For the purposes of this study, we defined social capital as “the shared knowl-
edge, norms, rules and networks that facilitate collective experience within a
neighborhood” (Vemuri, Grove, Wilson, & Burch, 2011, p. 6). This definition
is consistent with the design of the BES survey that provided social capital
data for this study. Similar studies have operationalized social capital by ask-
ing participants about trust, mutual cooperation, and the closeness of neigh-
borhood ties (Leyden, 2003; Putnam, 2000; Rogers, Halstead, Gardner, &
Carlson, 2010).
Despite contradictory definitions (Coleman, 1988; Hanifan, 1916; Jacobs,
1961) and critiques (Portes, 1998; Taormina, Kuok, & Wei, 2012) of its valid-
ity as a construct at times, social capital remains viable for measuring neigh-
borhood social relations. Surveys of neighborly trust reflect actual neighbor
relations, and higher social capital relates to tangible outcomes such as
reduced crime, increased physical activity, and public participation
(Lindström, 2011; Putnam, 2000). Although the environment does not create
social capital, it provides opportunities for social interaction that facilitate
strong neighborhood networks (Halpern, 2005; Sullivan et al., 2004).
Literature that tests this relationship studies the effect at the individual level
of neighborhood design characteristics. The physical form of a neighborhood
affects interactions and can increase social capital and connection (Appleyard
& Lintell, 1982; Jacobs, 1961). For example, walkable neighborhood envi-
ronments are correlated with higher levels of social capital than “sprawling,”
car-dependent neighborhoods (Leyden, 2003; Putnam, 2000; Rogers et al.,
2010). Social relationships in these neighborhoods are reflected in social
capital measured at the individual level. Therefore, neighborhood character-
istics are appropriate predictors of social capital.
This study uses a common method of measuring social capital, but adds
the novel use of tree canopy as the independent variable, in place of the more
abstract green space variables used in previous studies. Recent innovations in
remote sensing have made tree canopy data widely available, and its use as a
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
Holtan et al 5
variable has become more common in urban ecological research (Troy et al.,
2012; Zhou & Troy, 2008).
Our hypothesis grows from an understanding of how greener neighbor-
hood spaces might result in greater levels of social capital. Tree canopy cre-
ates opportunities for interaction at the street level by increasing use of the
streets as people are drawn outside their homes to enjoy the restorative ben-
efits of greenery and by making streets a more pleasant place to linger should
one encounter a neighbor. This understanding is supported by a number of
studies. Wood and Giles-Corti (2008) identified three levels of environmental
influence on social capital: the neighborhood context, the neighborhood
design, and specific neighborhood elements, one of which is access to green-
ery. Broyles, Mowen, Theall, Gustat, and Rung (2011) found parks with users
with greater levels of social capital had greater levels of use, which supports
a relationship between green space and social capital, as well as the relation-
ship’s incidental benefits (Broyles et al., 2011). Vemuri et al. (2011) found
that social capital correlated with neighborhood satisfaction and access to a
clean natural environment in Baltimore. What we do not know, however, is
the extent to which the density of nearby tree canopy is associated with the
strength of social capital reported by individuals. Until recently, it has been
difficult to accurately quantify tree canopy in the urban setting.
New Technology and Measuring Green Spaces
Increases in remote sensing technology, such as Light Detecting and Ranging
(LiDAR), have allowed researchers to study the relationship between tree
canopy and indicators of human health and well-being. Donovan, Michael,
and Butry (2011) found a correlation between tree canopy and birth out-
comes, and hypothesized stress reduction as the vehicle for better maternal
health and consequently, improved birth outcomes. Recent research from
Baltimore has used LiDAR to demonstrate a 10% increase in tree canopy to
be associated with a 12% drop in crime (Troy et al., 2012). Advances in aerial
imagery have also allowed researchers to identify correlates of tree canopy
such as urban form, and socioeconomic status (Grove, Cadenasso, et al.,
2006; Heynen & Lindsey, 2003).
Other studies have measured “green space” based on experts’ ratings of
the greenness of neighborhood common spaces (Kweon et al., 1998) or satel-
lite imagery to measure parks across an entire country (Maas et al., 2009).
LiDAR-derived data enabled us to measure a new green space typology in
social effects research—tree canopy across an entire neighborhood. The
quality and availability of data permit us to control for possible confounding
and antecedent variables like land use, socioeconomic status, neighborhood
design, and individual characteristics.
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
6 Environment and Behavior
We add to the literature that links the health of ecosystems to the health of
human inhabitants. We investigated whether neighborhood tree canopy cover
functions similarly to other types of green spaces that promote and support
neighborhood social life. Our hypothesis is that tree canopy cover at the cen-
sus block group level will be positively correlated with an individual’s social
capital, while controlling for possible confounding factors like individual
age, neighborhood aggregates of socioeconomic status, and neighborhood
urban form.
Methods
The Baltimore Ecosystem Study
We used data collected and created as part of the Baltimore Ecosystem Study
(BES), located in the Baltimore Metropolitan Region. The BES is one of two
National Science Foundation-funded urban Long Term Ecological Research
(LTER) sites in the nation.The purpose of the LTER is to study the relation-
ship between social, geophysical, and biological dynamics. Scientists in
many disciplines from universities and local, federal, and state government
agencies partner with community groups to study the complex interaction
between people and the environment (Cadenasso, Pickett, & Grove, 2006).
The research focuses on urban watersheds, patch dynamics, and building a
human ecosystem framework. The data that have been collected are open and
available for public download and use. BES data are the source of the social
survey data and land cover data we used in the study.
Study Area
The study area includes Baltimore, Maryland, which is home to more than
600,000 people in 92 square miles. Baltimore City is the cultural and finan-
cial center of the Baltimore-Townson Metropolitan Statistical Area (MSA),
which is home to almost 3 million people (U.S. Census Bureau, 2010a).
Baltimore originated as a port city and manufacturing hub for the eastern
seaboard. The decline of American manufacturing, sprawl to surrounding
Baltimore County, and disinvestment in general has negatively affected the
physical, natural and social environment within the city. At the same time, a
1960s urban growth ring, aggressive preservation plans, and community
activism have preserved and maintained high quality green spaces through-
out the city. Tree canopy within the city varies from neighborhoods with vir-
tually no tree canopy to neighborhoods that are three quarters shaded with
trees. The average tree canopy cover for the city is 24%. The city has a goal
of 40% by year 2040 (www.baltimoretreetrust.org).
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
Holtan et al 7
Data
Social survey. Data for individual level characteristics, including the measure
of social capital, came from the BES’s 2006 Greater Baltimore Recreation
and Neighborhood Questionnaire (hereafter the BES Survey; Vemuri &
Grove, 2006). The 2006 BES survey was a revised version of those from
1999, 2000, and 2003 used to measure changing perceptions, attitudes, and
behaviors about recreation, amenities, and quality of life in the Greater Balti-
more area. More detailed description of BES survey design and analysis can
be found in articles such as Vemuri et al. (2011).
The original survey was conducted with a sample of Baltimore MSA resi-
dents stratified by level of urbanization and socioeconomic status. Geographic
information and telephone numbers for each participant were purchased from
Claritas marketing firm. Target completion rates for each sampling unit
ranged from 50 to 150 participants depending on the population size of the
unit. Contact was made with the initial stratified sample of 4,880 potential
participants over the age of 18. Surveys were administered by telephone
using Computer Assisted Telephone Interviewing. Repeated callbacks and
postcard follow-ups were used to increase response rates. Researchers
obtained a response rate of 36.1%, a generally acceptable rate in social sur-
vey research. Focus groups and pretests were used to ensure the wording of
the questionnaire design (Vemuri, 2010). We obtained the survey data in an
Excel spreadsheet through the BES data coordinator; the survey instrument
and survey methodology were available for public download through the
BES website.
Secondary data sample selection. We selected cases that were in the 510 Fed-
eral Information Processing Standards (FIPS) county code to select only resi-
dents within the Baltimore City line (n = 709). This was to keep the sample
predominantly urban and eliminate correlations between tree canopy and
agricultural land uses. We included questions about social capital, personal
green space, and demographics.
Missing values are a common problem in social survey research. Of the
initial 709 cases, we removed nine because they lacked values for four or
more of the social capital variables of interest. We used a case mean method
to replace the remaining missing values in questions relating to social capital.
We rounded the mean of existing values for social capital-related questions to
the nearest whole number to replace the missing values. This is an appropri-
ate technique because the questions are highly correlated, relate to one con-
struct, and will be used to form a summative scale index for the purposes of
analysis (Roth & Switzer, 1999).
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
8 Environment and Behavior
Finally, we selected only the cases that had more than the median percent
of residential land in the neighborhood for analysis (median residential land
use per block group = 77%). This was to limit the analysis to predominantly
residential neighborhoods and avoid the confounding factors of large swaths
of undeveloped land, natural areas, or large industrial or institutional uses
that are also correlated with tree canopy and could potentially affect the mea-
surement of neighborhood connection (Heynen & Lindsey, 2003). The final
sample size was 361 individuals. These individuals were distributed across
191 block groups of the 710 block groups in the city, at an average of two
responses per block group.
Variables
Individual characteristics. We derived social capital, the dependent variable,
from the 2006 BES Survey. Residents were asked to score seven statements
about their neighborhood on a scale from 1 to 5. Of the seven questions, we
included the five questions that related most to qualities that described neigh-
borhood social connection and association, as opposed to outside group
involvement in neighborhood activities (see Table 1). The index we con-
structed for social capital closely resembled other summative scales measur-
ing social capital, including previous studies published by the BES research
team (Vemuri et al., 2011), and other studies of social capital (for example,
the American Community Social Capital Benchmark Survey; Broyles et al.,
2011; Leyden, 2003; Putnam, 2000). The Cronbach’s alpha for the index was
.835. The individual ’s age was also included in the analysis to account for
that main predictor of social capital (Putnam, 2000).
Green space characteristics. We included neighborhood tree canopy, access to
a green yard and access to parks within the neighborhood as green space
characteristics in the analysis. Tree canopy data came from a 2007 geographic
dataset made available through the BES public data access website (BES,
2007). The data were the product of high-resolution aerial imagery and light
detecting and ranging (LiDAR) technology, which is classified using light
intensity and surface heights. In addition to tree canopy, grass/shrub, water,
buildings, roads, bare earth and impervious surfaces can be classified. We
calculated neighborhood tree canopy as a percentage of the census block
group. We used bivariate responses to questions from the 2006 BES survey
asking residents whether they had a yard with grass or trees to model the
impact of personal green space.
Land use data from 2008, including parks, came from Baltimore City
Enterprise Geographic Information Services. The land use data were avail-
able at the parcel level and also included all rights of way so that the land use
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
Holtan et al 9
Table 1. Variables Used in the Analysis.
Variable name (source)
Question wording/description (response
range)
Individual characteristics
Social capital (BES Telephone
Survey)
How strongly would you agree or
disagree with the following five
statements about your neighborhood:a
people in the neighborhood are willing
to help one another; this is a close-
knit neighborhood; people in this
neighborhood can be trusted; there are
many opportunities to meet neighbors
and work on community problems;
and there is an active neighborhood
association. (1, strongly disagree to 5,
strongly agree. The summative index
had a Cronbach’s alpha reliability
coefficient of .835.)
Age (2006 BES Telephone Survey) Please stop me when I reach the category
that includes your age. Are you (under
35, 35-44, 45-54, 55-64, 65+)
Neighborhood demographic characteristicsb
College degree (U.S. Census
Bureau’s American Community
Survey 2005-2009 5-year estimate)
Percent of people over age 25 who have
received a bachelor’s degree or higher
(0%-94%)
Household income (U.S. Census
Bureau’s American Community
Survey 2005-2009 5-year estimate)
Median household income (in 2009
inflation adjusted dollars; US$8,353-
US$219,671)
Neighborhood urban form characteristicsb
Housing units per acre (U.S. Census
Bureau’s American Community
Survey 2005-2009 5-year estimate)
Housing units per acre (2.31-38.49)
Green space characteristics
Tree canopy (BES land cover
dataset)
Percent tree canopy coverage (1%-70%)
Green yard (2006 BES Telephone
Survey)
Does your residence have a yard with
grass or trees? (No, Yes)
Park presence (Baltimore City GIS
land use dataset)
Parks in block group? (No, Yes)
Note. BES = Baltimore Ecosystem Study; GIS = geographic information system.
aBES researchers defined “neighborhood” for survey participants as “block or street you live
on and several blocks or streets in each direction.”
bFor neighborhood characteristics, we defined neighborhood as the block group because it
is the closest approximation to the delimitation of neighborhood given to survey participants
(similar to Vemuri, Grove, Wilson, & Burch, 2011; Troy, Grove, & O’Neil-Dunne, 2012).
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
10 Environment and Behavior
dataset formed a complete cover of the land area of Baltimore City. Multiple
potential land use categories existed to identify green space including parks
and recreation, natural area, cemetery, undeveloped, private recreational,
public institutional, and public institutional-non-city. We selected the parks
and recreation land use for analysis because parks are specifically pro-
grammed for public use and enjoyment of nature and outdoors. We calculated
the parks as a bivariate measure of the presence of a park and recreation land
use in the block group.
We omitted public institutions, such as schools, or private recreation sites,
like golf courses, because these land uses are not accessible to all people at all
hours. In addition, schools can function as sites of access and perpetuation of
social capital (Hanifan, 1916), and we did not want to confound the impact of
schools’ green space with their social networks. Similarly, though undevel-
oped land does contain patches of green space, it is also representative of other
land use, planning, and historical mechanisms that are not part of this study.
Natural areas and cemeteries make up less than 3% of the total land area of the
city and do not occur frequently enough to include. In addition, their primary
purpose is not social meeting, whereas parks have a social function.
Neighborhood socioeconomic data. We obtained socioeconomic data for block
groups from the American Community Survey 5-year estimate for 2005-2009
(U.S. Census Bureau, 2011) to statistically control for variables that can
influence tree canopy cover. Studies have found that median income can pre-
dict vegetation richness (Martin, Warren, & Kinzig, 2004) and education
level is a correlate of tree canopy cover (Heynen & Lindsey, 2003). These
variables are also correlated with factors that could potentially affect social
capital. We constructed the education-level variable as a percentage of resi-
dents in the block group who had attained a bachelors degree or higher.
Neighborhood urban form. We included neighborhood urban form variables to
additionally control for variables that could affect the relationship between
tree canopy and social capital. Housing unit density is a measure of units per
acre to control for space available where trees can be planted (Grove, Cade-
nasso, et al., 2006). Housing unit density is also a partial indicator of walk-
ability (Cerin, Leslie, Owen, & Bauman, 2007; Leslie, Cerin, DuToit, Owen,
& Bauman, 2007) and sprawl and could control for other potential correlates
of social capital (Leyden, 2003; Putnam, 2000; Rogers et al., 2010).
We omitted the median age of the structures of the neighborhood from the
analysis because housing age has been shown to have a curvilinear relation-
ship with tree canopy (Grove, Cadenasso, et al., 2006). Trees in newer neighbor-
hoods have not had the time to attain functional maturity in terms of canopy
cover. In contrast, older neighborhoods may have had mature trees that have died
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
Holtan et al 11
and not been replaced. Further complicating this relationship is that older neigh-
borhoods in Baltimore may have smaller lot sizes, and therefore less space for
trees. We decided to omit architectural variables due to sample size and these
confounding variables, which is not unprecedented (Vemuri et al., 2011).
Case construction. Neighborhood was defined for survey participants as
including “both the block or street [they] live on and several blocks in either
direction” (Vemuri, 2010, p. 2). We operationalized neighborhood for the
biophysical, built environment, and socioeconomic variables as the census
block group (similar to Troy et al., 2012; Vemuri et al., 2011). We joined the
neighborhood variables to survey responses using a common FIPS identifier
for block group (U.S. Census Bureau MAF/TIGER database, 2009). We used
ESRI’s ArcMap 10 to calculate land use and land cover percentages within
the block group. We determined the total land area for each block group from
the land use dataset to avoid including the water area of Chesapeake Bay in
the waterfront block groups’ analyses. The land use dataset was a continuous
cover including taxable parcels and transportation rights of way, so that all
land within the city was included. The final product of this data preparation
was a set of cases (N = 361) that included individual survey responses and the
characteristics of the neighborhoods where respondents lived. Table 1
describes the variables used in the analysis.
Results
Table 2 presents the descriptive statistics for the subset of the BES survey
sample that was used for this analysis. The 2005-2009 5-year estimate from
the American Community Survey by the U.S. Census Bureau provides a
comparison. The survey sample was more white, female, wealthier, and
Table 2. Socioeconomic Characteristics of the Survey Sample and Baltimore City.
Female (%)
White
Caucasian
(%)
Household
income
> US$50,000
(%)
College-
educated
(over age 25)
(%)
Employed
(over age 16)
(%)
Married
(over age 15)
(%)
Own
home
(%)
BES survey
sample
(N = 361)
68.7 45.4 47.1 45.8 53.9 43.6 70.6
City of
Baltimorea
53.5 30.9 36.8 23.3 53.9 26.6 50.7b
Note. BES = Baltimore Ecosystem Study.
aSource. U.S. Census Bureau (2006) American Community Survey 1-year estimate.
bThis is the rate of owner occupied housing units.
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
12 Environment and Behavior
Table 3. Descriptive Statistics of Variables Used in the Analysis (N = 361).
Minimum Maximum M SD
Social capitala5 25 17.07 5.451
Ageb1 5 2.99 1.422
College degree 0% 94% 31% 26%
Household income US$8,353 US$219,671 US$50,978 US$32,251
Housing units per acre 2.31 38.49 12.0212 7.60392
Tree canopy 1% 70% 27% 16%
Green yarda0 1 0.81 0.396
Park presence 0 1 0.2 0.404
aIndicates variable measured at the individual level. The other variables were measured at the
neighborhood level.
bAge was measured at the individual level, n = 356.
.educated than the population of Baltimore City as a whole. This is typical of
other telephone surveys (Dillman, Smyth, & Christian, 2009). In addition,
survey participants included in this analysis were more likely to be married
and own their own home. The survey sample characteristic that most closely
approximates Baltimore City was the percent of employed people over the
age of 16. Both samples had an employment rate of just under 54%. However,
the survey sample employment rate might reflect a higher number of retired
participants who were more likely to be at home and willing to answer a tele-
phone survey (Dillman et al., 2009). The slightly skewed representation of
Baltimore City’s population in the survey sample affects the generalizability
of the study. These differences between the general population and survey
sample were more extreme than for previous BES studies (Vemuri et al.,
2011). This might be because this study was limited to Baltimore City, which
has fewer white residents, lower average income, and lower home ownership
rates than the Baltimore region that comprised the full BES survey sample.
Table 3 shows the descriptive statistics for the variables in the analysis.
On average, survey participants had a slightly better than average level of
social capital, as indicated by the mean score of 17.07 on the social capital
scale. An individual who answered “neutral” to all of the statements about
social capital would have received a 15 on the social capital index. Thus, the
average participant agreed at least twice or strongly agreed once to a question
about social capital. The descriptive statistics show that the sample lived in
neighborhoods with a US$50,978 median household income and the neigh-
borhood average percentage of people who have college degrees is 30%. The
socioeconomic statistics demonstrate the wide variety of neighborhoods in
Baltimore. Housing density ranged from 2 to 38 units per acre. The average
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
Holtan et al 13
neighborhood canopy cover was 27%, only slightly higher than the citywide
average of 24%. Eighty-one percent of the survey participants lived in a
house with trees or grass in the yard, and 20% of the neighborhoods where
participants lived contain a park.
Were the Major Variables Related?
The correlation matrix in Table 4 presents correlations between the variables
included in the analysis. Social capital had a statistically significant, positive
correlation with tree canopy (Pearson’s r = .241, p < .01). Social capital was also
correlated with college degree (Pearson’s r = .245, p < .01) and household
income (Pearson’s r = .201, p < .01), and inversely correlated with housing units
per acre (Pearson’s r = −.141, p < .01). Social capital was not correlated with the
other two green space variables of green yard and park presence. Contrary to
previous literature, age was not significantly correlated with social capital.
Tree canopy was also significantly correlated with neighborhood median
household income (Pearson’s r = .406, p < .01) and education levels (Pearson’s
r = .466, p < .01). There was a strong, significant negative correlation between
tree canopy and housing unit density (−.688, p < .01). This perhaps reflects
the lack of space to plant trees, the lower income of neighborhoods and his-
toric maintenance of street trees due to unequal distribution of city services..
Does Green Space Predict Social Capital After Other Factors
Are Considered?
We performed multiple linear regressions using SPSS and the enter method,
which simultaneously adds all variables to the model. We first modeled the
relationship between green space and social capital.
Table 4. Bivariate Correlations Between Variables of Interest and Social Capital
(N = 361).
Social capital Age
College
degree
Household
income
Housing
units per
acre
Tree
canopy
Green
yard
Age .080
College degree .245** −.142**
Household income .201** −.088 .655**
Housing units per acre −.141** −.026 −.107* −.268**
Tree canopy .241** .043 .466** .406** −.688**
Green yard .085 .056 .011 .107* −.474** .387**
Park presence −.025 .053 .014 −.151** −.042 .219** .058
*p < .05 level, two-tailed. **p < .01 level, two-tailed.
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
14 Environment and Behavior
Table 5. Green Space Regression Model for Social Capital (N = 361).
(Constant) .000
Tree canopy .264 .000
Green yard −.012 .825
Park presence −.082 .120
Adjusted R2 = .057
Table 5 shows that the green space linear regression model accounted for
almost 6% of the variation in participants’ reported social capital. Tree can-
opy contributed the most to the model (β = .264), and was significant at the p
< .01 level. The other two green space variables did not contribute signifi-
cantly to the model.
Next, we conducted three multiple regressions using other available social
capital predictors, neighborhood socioeconomic variables, and other vari-
ables that affect tree canopy and that might falsely affect its contribution to
social capital. Table 6 displays the results of these three models. In the first
model, we included all variables except the green space variables. The sec-
ond model included green space variables other than tree canopy. Tree can-
opy was added to the final model to see if any explanatory power was added
after accounting for the other factors.
Table 6 shows that Model 1, which did not include green space variables,
accounts for 7.3% of the variation in social capital. Participant age, neighbor-
hood education levels, and housing units per acre contributed
significantly to the model. However, when the green space variables other
than tree canopy were added to the model, the explanatory power of the
model decreased to 7.1%, though this was not a statistically significant
change. When tree canopy was added to the model, the explanatory power
increased to 7.6% of the variation in participants’ social capital, which was
greater than the other models alone. In the full model (Model 3), age, tree
canopy cover, and neighborhood education levels were the only variables that
contributed significantly to the model.
We then conducted a regression using only the significant variables from
the full models. Table 7 shows that the adjusted R2 increased from .066 to
.081 when tree canopy was added to the model using only the significant
variables from the previous models. Tree canopy added 1.5% more predictive
power to the model of social capital. All variables were significant at the p <
.01 level, except for age, which was significant at the .05 level in the second
model.
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
Holtan et al 15
Discussion
Our hypothesis was that tree canopy cover at the census block group level
would be positively correlated with social capital and that this relationship
would hold after controlling for possible confounding factors like individual
age, neighborhood aggregates of socioeconomic status and neighborhood
urban form. The bivariate correlations and regression models suggested a
significant relationship between tree canopy and social capital. This was con-
sistent with other findings on the relationship between green space and social
benefits. Green space close to an individual’s home has been shown to have
a significant relationship with social support (Maas et al., 2009). Our findings
Table 6. Full Regression Models for Social Capital (n = 356).
Model
1
(Constant) .000
Age .112 .031
College degree .225 .001
Household income .033 .637
Housing units per acre −.103 .055
2
(Constant) .000
Age .113 .030
College degree .235 .001
Household income .019 .792
Housing units per acre −.091 .132
Green yard .034 .559
Park presence −.042 .426
3
(Constant) .000
Age .102 .051
College degree .172 .029
Household income .019 .796
Housing units per acre .000 .996
Green yard .021 .726
Park presence −.070 .206
Tree canopy .154 .086
Model 1 adjusted R2 = .073; F change = .000
Model 2 adjusted R2 = .071; F change = .626
Model 3 adjusted R2 = .076; F change = .086
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
16 Environment and Behavior
Table 7. Regression Models for Social Capital, Significant Variables Only (n = 356).
Model
1
(Constant) .000
Age .117 .025
College degree .258 .000
2
(Constant) .000
Age .100 .054
College degree .187 .002
Tree canopy .147 .011
Model 1 adjusted R2 = .066; F change = .000
Model 2 adjusted R2 = .081; F change = .011
provide an additional scale of influence. Not only are closer parks better for
neighborhood social life (Maas et al., 2009), the findings here demonstrate
that the green fabric of a neighborhood created by tree canopy also facilitates
the social health that is vital to neighborhood functioning.
The purpose of the study was not to fully model social capital. Rather, we
used existing data to better understand the added effects of tree canopy as a
new measurement of green space. Using only the statistically significant vari-
ables and adding tree canopy to the final model increased the model’s explan-
atory power by 22.72%, from explaining 6.6% to 8.1% of the variation in an
individual’s social capital. The Pearson’s correlations in Table 4 show that
tree canopy has a similar strength and significance as the neighborhood
socioeconomic indicators of household income and education, two estab-
lished correlates of social capital (Putnam, 2000). Given the complexity of
modeling social capital, these positive and statistically significant results
should be welcome by researchers, community advocates and policy makers
wanting to make the case for investing in the urban tree canopy.
The mechanism by which tree canopy facilitates increased levels of social
capital is likely through driving increased use of sidewalks and outdoor
spaces with trees. The Chicago studies also suggested use of outdoor space as
a mediator between green space and social interaction (Coley et al., 1997;
Kuo, Sullivan, et al., 1998; Sullivan et al., 2004). Alternatively, perhaps
neighborhoods with greater tree canopy cover provide stress reduction.
Neighborhoods with higher levels of tree canopy create a feeling of escape
that is essential to mental restoration, or increase the sense of mystery that
draws walkers around the corner to the next block to meet their neighbors (R.
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
Holtan et al 17
Kaplan & Kaplan, 1989). Although Kuo, Sullivan, et al. (1998) did not find
that stress reduction, mental fatigue, or mood mediated the relationship
between green common spaces and social ties, this is possibly an area for
future study given the different structure of neighborhood tree canopy from
green common spaces in public housing. Finally, it is possible that tree can-
opy increases neighborhood satisfaction (R. Kaplan, 2001; Vemuri et al.,
2011), which in turn bolsters social capital.
While tree canopy was significantly correlated with having a green yard
(Pearson’s r = .387, p < .01), the green yard itself did not correlate with social
capital or add additional explanatory power to the regression models. This
lends credence to the hypothesis that the relationship between tree canopy
and social factors are the result of increased use of neighborhood common
spaces that are made more hospitable because of tree canopy. One policy
implication would be that it is important that trees are planted in shared pub-
lic spaces, like sidewalk planting strips, for the whole neighborhood to ben-
efit. The lack of relationship between a green yard and social capital runs
contrary to the yard’s status as a public–private space that facilitates interac-
tion much like a porch (Wilkerson, Carlson, Yen, & Michael, 2011).
The lack of significant relationship between park presence in the neigh-
borhood and social capital could be a response to a perceived threat of crime
or unappealing parks. Thus, green space would not facilitate the creation of
social capital through opportunities for informal interaction because fewer
people would be in these areas. Due to cuts endemic to urban parks’ budgets,
Baltimore parks might have varying levels of maintenance and use and thus
do not unilaterally facilitate social interaction and social capital as would be
suggested by previous literature on neighborhood common spaces (Kuo,
Sullivan, et al., 1998). Future research might investigate the extent to which
tree canopy has a stronger relationship to neighborhood social life where
parks and common green spaces are used less. Other research could include
multiple constructions of green space at different scales of analysis, including
tree canopy, yards, small and large parks, and natural open spaces.
Tree canopy is one part of a complex relationship between neighborhood
urban form and socioeconomic characteristics. We chose to keep housing unit
density as a variable in the models, but we acknowledge the presence of
mediating variables. Denser neighborhoods have less physical space for tree
canopy, though they can be more walkable. However, unlike the findings of
Leyden (2003) and Rogers et al. (2010), housing unit density as a partial
indicator of walkability had a small but significant negative correlation with
social capital. It is possible that denser housing in the study area is located in
lower income, higher crime neighborhoods that do not facilitate social capital
via walkability and proximity to neighbors. We added tree canopy to the
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
18 Environment and Behavior
model last to best account for the impact of tree canopy among multiple inter-
related variables. Further research should include a comparison across a
broader range of neighborhood housing densities or include a more refined
index of walkability (such as explored in Smith et al., 2008).
It is possible that higher income neighborhoods also have less crime and
thus a higher level of trust. Crime was not included in this model because
crime is also correlated with lower social capital and we wanted to avoid
issues of reverse causality (Kennedy, Kawachi, Prothrow-Stith, Lochner, &
Gupta, 1998). However, lower crime rates have been correlated with greater
tree canopy cover (Donovan & Prestemon, 2012; Kuo & Sullivan, 2001; Troy
et al., 2012). The relationship between neighborhood socioeconomic status,
neighborhood green space, crime, the perception of safety and security, and
residents’ social capital is another open area in environmental design research.
The findings also provide contrast to existing literature on social capital.
Unlike what Putnam (2000) might suggest, age was not correlated with social
capital. However, it was useful as a predictor when other variables were mod-
eled together. In all three regression models, participant age contributed a
statistically significant amount to the model’s explanatory power, consistent
with Putnam’s (2000) hypothesis that older people have more social capital.
The low level of significance of age and the low adjusted R2 of the full model
reflected that the purpose of the study was not to fully model social capital,
but rather to understand the relationship between neighborhood tree canopy
and social interaction, using social capital as a robust indicator. Social factors
are difficult to model because of their complexity; thus, the relative impor-
tance of tree canopy in this model suggests its inclusion in further studies.
The limitations of this study included generalizability, a cross-sectional
design and the limitations of secondary data collection. The sample used did
not exactly match the larger environment of the city studied. Baltimore has
vegetation patterns, an urban form and history that are particular to the
Northeastern United States. The cross-sectional design cannot rule out selec-
tion threats to internal validity, whereby residents who prefer high levels of
social capital also choose neighborhoods with more tree canopy. In addition,
having used secondary data, it was difficult to know the cause of any errors
in missing values or sampling.
As expected, we found that tree canopy was significantly and positively
correlated with social capital. When accounting for possible confounding
variables like socioeconomic status and urban form, neighborhood tree can-
opy cover explained an additional 1.5% of the variation in an individual’s
social capital, an increase of 22.72%. The relationships between the variables
suggest a complex interaction between social capital and the environment. In
spite of this complexity, these findings suggest that more street trees should
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
Holtan et al 19
be planted in neighborhood public spaces. Trees are a relatively simple and
affordable public works intervention (Mcpherson et al., 1994), provide the
quickest access to the restorative benefits of green space needed by busy
people (S. Kaplan, 1995), and have multiple additional human health and
environmental benefits (Mcpherson et al., 1994). There are few drawbacks to
planting trees in response to the desire for increased neighborhood interac-
tions (Wood & Giles-Corti, 2008). Researchers and design professionals
should continue to include neighborhood tree canopy in the network of green
space that supports the social benefits of urban ecosystems.
Acknowledgment
This research would not have been possible without the data that was shared by the
National Science Foundation funded Baltimore Ecosystem Study Long Term
Ecological Research program housed within the Cary Institute of Ecosystem Studies.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research,
authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publi-
cation of this article.
References
Appleyard, D., & Lintell, M. (1982). The environmental quality of city streets: The res-
idents’ viewpoint. In S. Kaplan & R. Kaplan (Eds.), Humanscape: Environments
for people (pp. 233-258). Ann Arbor, MI: Ulrich’s Books.
Baltimore City Enterprise Geographic Information Services. (2008).
Baltimore city land use [Data file]. Retrieved from http://www.baltimo-
recity.gov/OfficeoftheMayor/MayoralOffices/InformationTechnology/
EnterpriseGeographicInformationServices.aspx
Baltimore Ecosystem Study. (2007). Baltimore City land cover [Data file]. Available
from www.beslter.org
Broyles, S. T., Mowen, A. J., Theall, K. P., Gustat, J., & Rung, A. L. (2011).
Integrating social capital into a park-use and active-living framework. American
Journal of Preventive Medicine, 40, 522-529. doi:10.1016/j.amepre.2010.12.028
Cadenasso, M. L., Pickett, S. T. A., & Grove, M. J. (2006). Integrative approaches to
investigating human-natural systems: The Baltimore ecosystem study. Natures
Sciences Sociétés, 14, 4-14. doi:10.1051/nss:2006002
Cerin, E., Leslie, E., Owen, N., & Bauman, A. (2007). Applying GIS in physical
activity research: Community “walkability” and walking behaviors. In P. C. Lai
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
20 Environment and Behavior
& A. S. H. Mak (Eds.), GIS for health and the environment development in the
Asia Pacific region (pp. 72-89). Retrieved from http://dx.doi.org/10.1007/978-3-
540-71318-0_6
Cohen, D. A., Inagami, S., & Finch, B. (2008). The built environment and collective
efficacy. Health & Place, 14, 198-208. doi:10.1016/j.healthplace.2007.06.001
Coleman, J. S. (1988). Social capital in the creation of human capital. American Journal
of Sociology, 94 (Supplement: Organizations and Institutions: Sociological and
Economic Approaches to the Analysis of Social Structure), S95-S120. Retrieved
from http://www.jstor.org/stable/2780243
Coley, R. L., Kuo, F. E., & Sullivan, W. C. (1997). Where does community grow?
The social context created by nature in urban public housing. Environment and
Behavior, 29, 468-494. doi:10.1177/001391659702900402
Dillman, D. A., Smyth, J. D., & Christian, L. M. (2009). Internet, mail, and mixed-
mode surveys: The tailored design method. Hoboken, NJ: Wiley & Sons.
Retrieved from http://books.google.com/books?id=s1naAAAAMAAJ&pgis=1
Donovan, G., Michael, Y., & Butry, D. (2011). Urban trees and the risk of poor birth
outcomes. Health & Place, 17, 390-393. doi:10.1016/j.healthplace.2010.11.004
Donovan, G., & Prestemon, J. (2012). The effect of trees on crime in Portland,
Oregon. Environment and Behavior, 44, 3-30. doi:10.1177/0013916510383238
Greenspace Scotland. (2008). Health impact assessment of greenspace: A guide.
Stirling: Author.
Groenewegen, P. P., van den Berg, A. E., de Vries, S., & Verheij, R. A. (2006).
Vitamin G: Effects of green space on health, well-being, and social safety. BMC
public health, 6, Article 149. doi:10.1186/1471-2458-6-149
Grove, J. M., Cadenasso, M. L., Burch, W. R. Jr., Pickett, S. T. A., Schwarz, K., Neil-
dunne, J. O., . . .Boone, C. (2006). Data and methods comparing social structure
and vegetation structure of urban neighborhoods in Baltimore, Maryland. Society
& Natural Resources, 19, 117-136.
Halpern, D. (2005). Social capital. Cambridge, MA: Polity Press.
Hanifan, L. J. (1916). The rural American community center. The ANNALS of the
American Academy of Political and Social Science, 67(130), 130-138. Retrieved
from http://www.jstor.org/stable/1013498
Heynen, N. C., & Lindsey, G. (2003). Correlates of urban forest canopy cover:
Implications for local public works. Public Works Management Policy, 8(33),
33-47. doi:10.1177/1087724X03253604
Jacobs, J. (1961). Death and life of great American cities. New York, NY: Random
House.
Jones, R., Davis, K., & Bradford, J. (2013). The value of trees: Factors influencing
homeowner support for protecting local urban trees. Environment and Behavior,
45, 650-676. doi:10.1177/0013916512439409
Kaplan, R. (2001). The nature of the view from home: Psychological benefits.
Environment and Behavior, 33, 507-542. doi:10.1177/00139160121973115
Kaplan, R., & Kaplan, S. (1989). The experience of nature: A psychological perspec-
tive. New York, NY: Cambridge University Press. (Republished, 1995 in Ann
Arbor, MI: Ulrich’s Books Inc.)
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
Holtan et al 21
Kaplan, S. (1995). The urban forest as a source of psychological well-being. In G.
A. Bradley (Ed.), Urban forest landscapes: Integrating multidisciplinary per-
spectives. Seattle: University of Washington Press. Retrieved from http://books.
google.com/books?hl=en&lr=&id=PfxEX4UkTbEC&pgis=1
Kennedy, B. P., Kawachi, I., Prothrow-Stith, D., Lochner, K., & Gupta, V. (1998).
Social capital, income inequality, and firearm violent crime. Social Science
Medicine, 47, 7-17. Retrieved from http://www.sciencedirect.com/science/arti-
cle/pii/S0277953698000975
Kuo, F. E., Bacaicoa, M., & Sullivan, W. C. (1998). Transforming inner-city land-
scapes: Trees, sense of safety, and preference. Environment and Behavior, 30,
28-59.
Kuo, F. E., & Sullivan, W. (2001). Environment and crime in the inner city:
Does vegetation reduce crime? Environment and Behavior, 33, 343-367.
doi:10.1177/0013916501333002
Kuo, F. E., Sullivan, W. C., Coley, R. L., & Brunson, L. (1998). Fertile ground for
community: Inner-city neighborhood common spaces. Journal of Community
Psychology, 26, 823-851. doi:10.1023/A:1022294028903
Kweon, B. S., Sullivan, W. C., & Wiley, A. W. (1998). Green common spaces and
the social integration of inner-city older adults. Environment and Behavior, 30,
832-858. doi:10.1177/001391659803000605
Lee, A. C. K., & Maheswaran, R. (2011). The health benefits of urban green spaces:
A review of the evidence. Journal of Public Health, 33, 212-222. doi:10.1093/
pubmed/fdq068
Leslie, E., Cerin, E., DuToit, L., Owen, N., & Bauman, A. (2007). Objectively
assessing “walkability” of local communities: Using GIS to identify the relevant
environmental attributes. In P. C. Lai & A. S. H. Mak (Eds.), GIS for health
and the environment development in the Asia Pacific region (pp. 91-104).
doi:10.1007/978-3-540-71318-0_7
Leyden, K. M. (2003). Social capital and the built environment: The importance of
walkable neighborhoods. American Journal of Public Health, 93, 1546-1551.
doi:10.2105/AJPH.93.9.1546
Lindström, M. (2011). Social capital, desire to increase physical activity and leisure-
time physical activity: A population-based study. Public Health, 125, 442-447.
doi:10.1016/j.puhe.2011.01.015
Maas, J., van Dillen, S. M. E., Verheij, R., & Groenewegen, P. P. (2009). Social
contacts as a possible mechanism behind the relation between green space and
health. Health & Place, 15, 586-595. doi:10.1016/j.healthplace.2008.09.006
Maas, J., Verheij, R. A, Groenewegen, P. P., de Vries, S., & Spreeuwenberg, P.
(2006). Green space, urbanity, and health: How strong is the relation? Journal of
Epidemiology & Community Health, 60, 587-592. doi:10.1136/jech.2005.043125
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.
Journal of Forest Economics, 11, 177-199. doi:10.1016/j.jfe.2005.08.002
Martin, C. A., Warren, P. S., & Kinzig, A. P. (2004). Neighborhood socioeconomic
status is a useful predictor of perennial landscape vegetation in residential
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
22 Environment and Behavior
neighborhoods and embedded small parks of Phoenix, AZ. Landscape and Urban
Planning, 69, 355-368. doi:10.1016/j.landurbplan.2003.10.034
Mcpherson, E. G., Nowak, D. J., & Rowntree, R. A. (1994). Chicago’s urban forest
ecosystem: Results of the Chicago urban forest climate project. Retrieved from
http://www.nrs.fs.fed.us/pubs/gtr/gtr_ne186.pdf
Park, B. J., Tsunetsugu, Y., Kasetani, T., Kagawa, T., & Miyazaki, Y. (2010).
The physiological effects of Shinrin-yoku (taking in the forest atmosphere or
forest bathing): Evidence from field experiments in 24 forests across Japan.
Environmental Health and Preventive Medicine, 15, 18-26. Retrieved from
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2793346&tool=pmc
entrez&rendertype=abstract
Portes, A. (1998). Social capital: Its origins and applications in modern sociology.
Annual Review of Sociology, 24, 1-24. Retrieved from http://www.jstor.org/
stable/223472
Putnam, R. (2000). Bowling alone: The collapse and revival of American community.
New York, NY: Simon & Shuster.
Rogers, S. H., Halstead, J. M., Gardner, K. H., & Carlson, C. H. (2010). Examining
walkability and social capital as indicators of quality of life at the municipal and
neighborhood scales. Applied Research Quality of Life, 6, 201-213. doi:10.1007/
s11482-010-9132-4
Roth, P., & Switzer, F. (1999). Missing data: Instrument-level Heffalumps and item-
level Woozles (Vol. 4). Research Methods Forum. Retrieved from http://division.
aomonline.org/rm/1999_RMD_Forum_Missing_Data.htm
Smardon, R. C. (1988). Perception and aesthetics of the urban environment: Review
of the role of vegetation. Landscape and Urban Planning, 15, 85-106.
Smith, K. R., Brown, B. B., Yamada, I., Kowaleski-Jones, L., Zick, C. D., & Fan, J.
X. (2008). Walkability and body mass index: Density, design, and new diversity
measures. American. Journal of Preventive Medicine, 35, 237-244.
Sullivan, W. C., Kuo, F. E., & Depooter, S. F. (2004). The fruit of urban nature: Vital
neighborhood spaces. Environment and Behavior, 36, 678-700. doi:10.1177/01
93841X04264945
Taormina, R. J., Kuok, A. C. H., & Wei, W. (2012). Social capital as dehumaniz-
ing terminology. Advances in Applied Sociology, 2, 143-148. doi:10.4236/aas-
oci.2012.22019
Troy, A., Grove, J. M., & O’Neil-Dunne, J. (2012). The relationship between tree
canopy and crime rates across an urban–rural gradient in the greater Baltimore
region. Landscape and Urban Planning, 106, 262-270. doi:10.1016/j.landurb-
plan.2012.03.010
Ulrich, R. (1984). View through a window may influence recovery from surgery.
Science, 224, 420-421. doi:10.1126/science.6143402
U.S. Census Bureau. (2006). American Community Survey 2006 1-year estimates
[Data file]. Retrieved from http://factfinder2.census.gov/faces/nav/jsf/pages/
index.xhtml
U.S. Census Bureau. (2010a). 2010 decennial census [Data file]. Retrieved from:
http://factfinder2.census.gov/faces/nav/jsf/pages/index.xhtml
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
Holtan et al 23
U.S. Census Bureau. (2010b). 2010 urban and rural classification main page. US
Census Bureau Geography Division. Retrieved from http://www.census.gov/geo/
www/ua/2010urbanruralclass.html
U.S. Census Bureau. (2011). American Community Survey 2005 to 2009 5-year esti-
mates [Data file]. Social Explorer. Retrieved from http://www.socialexplorer.
com/pub/home/home.aspx
U.S. Census Bureau MAF/TIGER Database. (2009). TIGER/Line blockgroup shape-
files for Maryland [Data file]. Retrieved from: http://www.census.gov/cgi-bin/
geo/shapefiles2009/national-files
Vemuri, A. W. (2010). BES household telephone survey overview. Baltimore
Ecosystem Study. Retrieved from http://www.beslter.org/data_Browser.asp
Vemuri, A. W., & Grove, J. M. (2006). Baltimore ecosystem study household tele-
phone survey [Data file]. Baltimore, MD: Baltimore Ecosystem Study.
Vemuri, A. W., Grove, J. M., Wilson, M. A., & Burch, W. R. (2011). A tale of two
scales: Evaluating the relationship among life satisfaction, social capital, income,
and the natural environment at individual and neighborhood levels in metropolitan
Baltimore. Environment and Behavior, 43, 3-25. doi:10.1177/0013916509338551
Wilkerson, A., Carlson, N. E., Yen, I. H., & Michael, Y. L. (2011). Neighborhood
physical features and relationships with neighbors: Does positive physical envi-
ronment increase neighborliness? Environment and Behavior, 44, 595-615.
doi:10.1177/0013916511402058
Wood, L., & Giles-Corti, B. (2008). Is there a place for social capital in the psychol-
ogy of health and place? Journal of Environmental Psychology, 28, 154-163.
doi:10.1016/j.jenvp.2007.11.003
Yamaguchi, M., Deguchi, M., & Miyazaki, Y. (2006). The effects of exercise in forest
and urban environments on sympathetic nervous activity of normal young adults.
The Journal of International Medical Research, 34, 152-159. Retrieved from
http://www.ingentaconnect.com/content/field/jimr/2006/00000034/00000002/
art00004
Zhou, W., & Troy, A. (2008). An object oriented approach for analysing and char-
acterizing urban landscape at the parcel level. International Journal of Remote
Sensing, 29, 3119-3135. doi:10.1080/01431160701469065
Author Biographies
Meghan Holtan recently graduated from the State University of New York College
of Environmental Science and Forestry with a master’s degree in environmental sci-
ence with a concentration in environmental and community land planning. She works
as a planning analyst at Agnew::Beck Consulting, a multidisciplinary firm in
Anchorage, Alaska, where her interests include community engagement and environ-
mental planning in Alaskan communities.
Susan Dieterlen is an assistant professor of landscape architecture in the College of
Environmental Science and Forestry at the State University of New York. Dr.
Dieterlen’s research interests include sociocultural issues in the built environment and
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
24 Environment and Behavior
postindustrial communities, as well as strengthening connections between landscape
architectural research and practice.
William Sullivan is professor of landscape architecture at the University of Illinois
where he examines (a) the health benefits of having everyday contact with green
places and (b) citizen participation in environmental design. He teaches on campus
and at the Danville Correctional Center—a medium and high security prison—and is
an active member of the Education Justice Project at Illinois.
at UNIV OF ILLINOIS URBANA on February 6, 2014eab.sagepub.comDownloaded from
