societies
Article
Social Capital Accumulation among Puerto Rican
Mothers in Urban Neighborhoods
Phillip J. Granberry 1,2, * and Maria Idalí Torres 3
1
2
3
*
Department of Economics and Gastón Institute, University of Massachusetts Boston, Boston, MA 02125, USA
Research Division, Boston Planning and Development Agency, Boston, MA 02201, USA
Department of Anthropology and Gastón Institute, University of Massachusetts Boston, Boston, MA 02125,
USA; [email protected]
Correspondence: [email protected]; Tel.: +1-617-331-6942
Academic Editor: Gregor Wolbring
Received: 1 January 2017; Accepted: 27 February 2017; Published: 3 March 2017
Abstract: Social capital provides access to material and personal resources through participation in
social networks and other social structures. Social capital may not function equally for all populations,
especially those living in residentially segregated urban neighborhoods with increased levels of
poverty. This is because inequalities exist in social capital accumulation and are found where
disadvantaged socioeconomic groups cluster. Using probabilistic household survey data consisting
of 205 Puerto Rican mothers in Springfield, Massachusetts in 2013, this research tests hypotheses
regarding the association of social capital accumulation with Puerto Rican mothers’ individual,
neighborhood, and social network characteristics. Logistic regression results suggested that Puerto
Rican mothers who were employed and lived in neighborhoods with other Latinos were more likely
to accumulate social capital. In addition, mothers who participated in activities of their children also
had increased social capital accumulation. This neighborhood effect on social capital accumulation
may promote bonding social capital but not bridging social capital among these Puerto Rican mothers.
Keywords: bridging social capital; bonding social capital; civic participation; social capital formation;
social cohesion; collective efficacy; trust; residential segregation; isolation
1. Introduction
Social capital is a mechanism for people to access otherwise unavailable information and
resources through their participation in social networks and other social structures [1]. Social capital
is important for Latinos as they develop an increased presence in the United States, so that they can
access non-pecuniary resources to strengthen their socioeconomic position. The effects of poverty on
Puerto Rican social capital have been studied [2], but a gap in the literature exists as to how Puerto
Ricans accumulate social capital. Other research has identified that Mexicans, the largest Latino
population in the United States, accumulate social capital through their employment and from living
in neighborhoods with increased homeownership [3]. Increased residential segregation, especially in
cities with a newly arriving population, could weaken the mechanisms through which social capital is
accumulated [4]. Puerto Rican neighborhoods provide opportunities to accumulate social capital, but
a combination of individual- and neighborhood-level factors could influence the type of social capital
accumulated. Puerto Rican mothers’ facility in accumulating bonding social capital developed through
relationships with others who are primarily like themselves (e.g., ethnicity, family, age, educational
attainment), in place of bridging social capital developed through relationships with others across
pronounced social divisions (e.g., race, class, or religion) may be problematic.
A challenge in conducting social capital research is addressing its heterogeneity and multiple
avenues of creation. Social capital implicitly embodies several dimensions: reciprocal social network
Societies 2017, 7, 3; doi:10.3390/soc7010003
www.mdpi.com/journal/societies
Societies 2017, 7, 3
2 of 13
exchange, civic engagement, and neighborhood social cohesion. This research focuses on how social
capital is accumulated through one dimension: reciprocal exchange. Even though Portes (1998)
highlighted the need for research into mechanisms through which social capital is accumulated, to date
a limited literature exists on individual-level determinants of social capital accumulation [5–10]. We are
motivated to study Puerto Rican women’s social capital accumulation (SKA) for two reasons. First,
similar to other Latinas with increased rates of poverty, Puerto Rican women are a socioeconomically
vulnerable segment of the population [11] that could be highly dependent on their accumulated social
capital, but poorer Latino families tend to depend on kin networks that offer fewer opportunities for
exchange [12] and have already lower resilience to added burden [13]. Second, Puerto Rican women
play an important role in their families [14] and in matricententric families they are the primary brokers
for formal institutional support systems and social networks [15]. Regardless of their family structure,
they are also the primary activator of their multigenerational and multinational family networks [16].
This position places them in contact with institutional support necessary for SKA [17]. The current
research tests hypotheses regarding the association of individual- and neighborhood-level factors
with SKA. These factors include individual demographic characteristics (e.g., age, marital status,
employment, income) and neighborhood and social network characteristics (e.g., collective efficacy,
trust, civic participation, Latino population density).
1.1. Theoretical Underpinnings of Social Capital
Putnam’s (2000) initial social capital research highlighted a dimension of community integration
where dense reciprocal networks develop through civic engagement that facilitated resource
transfers [1]. By social networks, we mean a set of interdependent relations among family, friends, or
neighbors who exchange some form of support on a regular basis that influence social behavior [12].
We use the concept of resources when referring to supportive behaviors that demonstrate solidarity,
trust, and other resilience-enhancing support; provide tangible aid and services to meet an immediate
personal need for instrumental support; and/or share knowledge and information needed to facilitate
problem-solving [18]. According to Putnam's theory, when relationships are reciprocal, trust develops
among social network members. Thus, the combination of trust and exchange allows for the potential
development and accumulation of more resources. Glaeser, Laibson, and Sacerdote made the distinction
that this trust must be shared because by itself, trust is a necessary but not sufficient condition for
reciprocity to develop [5]. Another dimension of social capital, social cohesion, arises when bonds
link social network members to the larger group process. Sampson and colleagues further nuanced
social cohesion with their research on collective efficacy: a process of activating social ties among
neighborhood residents in order to achieve collective goals, such as public order or control of crime [10].
By developing collective efficacy, individuals engage in social networks, building trusting relationships
that activate mechanisms for the common goal of social change. Underlying both trusting reciprocal
relationships and social cohesion is a shared lived experience, where social networks develop among
individuals with shared interests and identities. Thus, through a principle known as homophily,
individuals are more likely to develop social capital among individuals with whom they share a
cultural identity [19].
Two types of social capital have been identified: bonding and bridging [1]. Both strong ties
(e.g., relatives and close relationships) and weak ties (e.g., acquaintances or friends of a friend)
provide valued resources for SKA. Weak ties are crucial for bridging resources outside of one’s
neighborhood [3,20]. In fact, the strength of weak ties, as developed by Granovetter (1973) in his
influential paper, depends on these relationships’ ability to provide access to new information [21].
While strong ties found in dense networks can mobilize help more quickly and intensely, they often
circulate redundant information [22]. Additionally, people with homogeneous networks may find
themselves subject to the normative expectations and practices that must be met to ensure continued
inclusion in a network, and this occurs at the expense of developing weak ties [5]. Weak ties, on the
other hand, add a diversity of ideas, attitudes, backgrounds, goods, and services to a person’s network,
Societies 2017, 7, 3
3 of 13
facilitating access to non-redundant information and resources. Even though bridging and bonding
social capital are distinct, at the neighborhood level the transformation of bonding social capital into
bridging social capital is highly dependent upon residents’ social status [23].
1.2. Social Capital in a Puerto Rican Neighborhood
Small’s research on an urban Puerto Rican neighborhood in Boston demonstrated how social
capital exists in a unique context of people’s experiences in their environment [2]. This research
highlighted the mechanisms behind social organization and social isolation that generate social ties
necessary for SKA to occur. An important ecological artifact, neighborhood loyalty, can entice either
attachment or dissociation, and this enticement may limit a resident’s motivation to develop important
weak ties with individuals outside of his or her neighborhood. Villa Victoria, the study’s neighborhood,
had public plazas that were actively used and were located only a short distance from individuals’
homes, providing a resource-rich Spanish-friendly environment, but personal characteristics such as
being employed influenced a resident’s ability to generate external networks important for developing
bridging social capital.
1.3. Individual and Demographic Characteristics and Social Capital Accumulation
The gendered provision of social support is a key mechanism by which social networks affect a
variety of outcomes [7,24]. According to both the gender and social support literatures, women are
more likely than men to seek and provide social support [25]. In general, women tend to incorporate
a larger number of kin into their networks [26]. Fuhrer and Stansfeld found that women not only
reported more close relationships than men but were also less likely to identify their partner as their
closest relationship [27]. Martin and colleagues’ research [28] identified a nuanced process highlighting
how a historical dependency on strong ties hindered Latina women in developing and using weak
ties after entering undergraduate engineering programs. Once in a program, school personnel offered
support, but delayed recognition or identification of available resources slowed these women’s access
and activation of resources, leading to difficult university transitions. Once activated, these resources
provided sources of social capital to assist in their academic success. In addition, other demographic
and individual characteristics can shape SKA. Social capital was positively correlated with employment
and financial capital, which provided important labor market information to Puerto Rican women that
they used to acquire higher-paying jobs [29]. However, compared to Mexican men, employed Mexican
women were less likely to accumulate social capital [3].
1.4. Neighborhood Effects on Social Capital Accumulation
Putnam’s research suggested that social capital was strongest in areas with less diversity. Because
homophily was associated with trust, it was a strong correlate with social capital, especially bonding
social capital [30]. This does not mean that diversity was necessarily harmful to SKA. In fact, diversity
may assist in developing bridging social capital. As contextual or city-level diversity increased in
Canada, the level of participation in organizations and trust in others increased [31]. However,
increasing Latino isolation in the United States [32] may limit opportunities to develop weak ties
necessary for developing social capital. Residential segregation is problematic because it concentrates
certain ethno-racial groups in a smaller geographic area. If poverty increases in segregated areas, poorly
distributed resources become even scarcer because of a perceived helplessness and political apathy that
can lead to a public abandonment of a segregated geographic area and the people living there [33–35].
A positive association exists for Latino residential segregation in metro areas with increased Puerto
Rican populations, and Springfield has the fourth highest segregation index in U. S. metropolitan
areas with high concentrations of Puerto Ricans [36]. Increased isolation of Latinos in segregated
neighborhoods could thus be associated with a decline in bridging social capital. In contrast, both
increased levels of neighborhood homeownership and population density in Latino neighborhoods
Societies 2017, 7, 3
4 of 13
have been found to be positively associated with SKA [3]. These neighborhoods provided abundant
opportunities for developing strong ties necessary for bonding social capital to be accumulated.
1.5. Civic Participation
Civic participation has garnered much attention in the social capital literature, especially among
political scientists. Social capital is a dimension of community integration that is an artifact of
people’s ties to their communities and their institutions. Civic participation increases generalized social
trust [6,31], and increasing participation in community institutions is generally viewed as an effective
strategy for developing strong communities and democracy. Community participation and social
capital have been found to be significantly higher for adults in households with children. These adults
reported having increased reciprocity and trust in relationships [37]. This participation does not have to
be active, but simply being affiliated with several organizations increases SKA through formal avenues
of information sharing [9]. For example, church membership and (not necessarily regular attendance)
is integral for the development of generalized social trust [8]. Participating in large civic organizations
is related to developing homophilous personal networks, and participating in heterogeneous civic
organizations provides more opportunities for individuals to form diverse personal networks [38].
Civic participation in organizations with individuals from diverse occupational backgrounds helps
develop relations to high-status contacts, even for individuals with lower socioeconomic status [39].
Civic participation also appears to be related to the development of negation skills, and the creation of
networks of mutual obligation that facilitate SKA [40]. When people are civically active, they encounter
people whom they might not meet through their regular social network participation, especially if this
participation changes over time.
2. Methods
Data used to test hypotheses about SKA are from an intervention study to evaluate the
effectiveness of a theory-based culturally appropriate Spanish media campaign to improve Puerto
Rican mother–child communication about sexual health [41]. Social capital was hypothesized to be a
predictor of maternal communication, and questions about social capital were included in the survey
instrument. A mother was not necessarily a biological mother: any woman in a full-time caregiving
role for a child could be included in the study. These roles included grandmothers, aunts, sisters,
and foster mothers. We merged these data with 2009–2013 American Community Survey block-level
data. Our research was reviewed and approved by the Institutional Review Board and was conducted
in collaboration with the Puerto Rican Cultural Center, Inc., a community-based organization that
provides educational and cultural programs to support Puerto Rican residents in Springfield.
2.1. Sample Frame and Selection
Springfield, the third largest city in Massachusetts, was home to an estimated 61,586 Latinos as of
the 2010 Census, and they made up 40.3% of the city’s population. Puerto Ricans were the dominant
Latino subpopulation, comprising 76.8 percent of the Latino population in the city [42]. Hampden
County, where Springfield is located, ranked second in 2010 among U.S. counties where Puerto Rican
residents had lived in Puerto Rico one year prior [43].
The sampling framework for the study began by identifying 10 census tracts of the 39 in
Springfield, in which at least seven percent of households had a Puerto Rican mother and at least one
child between the ages of 10 to 19 years. We randomly selected 100 census blocks from 298 populated
blocks in the 10 tracts. Figure 1 highlights the sampling methodology used to collect the data for
this research. Six teams containing an undergraduate student and a community resident enumerated
100 blocks and identified 4828 eligible households during the fall of 2012 [44]. Second, we randomly
assigned approximately half (2252) of the enumerated households to this survey. Third, we screened
the households assigned to the study and found 265 (11.7%) that met our study criteria. That is, living
in the household was a self-identified Puerto Rican mother, with at least one child between the ages of
Societies 2017, 7, 3
Societies 2017, 7, 3
5 of 13
5 of 13
of 10
to years,
19 years,
to carry
a conversation
in Spanish.
Of265
theeligible
265 eligible
mothers,
60 refused
10
to 19
andand
ableable
to carry
a conversation
in Spanish.
Of the
mothers,
60 refused
to be
to
be interviewed,
yielding
65.2 percent
AAPOR
#3 response
rate [45].
interviewed,
yielding
a 65.2apercent
AAPOR
#3 response
rate [45].
Figure 1. Sample Framework.
The
interviewing
waswas
conducted
fromfrom
July July
2013 2013
to January
2014.
The fieldwork
fieldwork for
forscreening
screeningand
and
interviewing
conducted
to January
All
interviews
were
conducted
at
a
mother’s
home
by
one
of
twelve
teams
consisting
of
Spanish2014. All interviews were conducted at a mother’s home by one of twelve teams consisting of
speaking
undergraduate
students students
and community
residents.residents.
Interviews
were conducted
in Spanish
Spanish-speaking
undergraduate
and community
Interviews
were conducted
in
and
averaged
56
min.
All
participants
received
US$25
upon
the
interview’s
completion
[44].
Spanish and averaged 56 min. All participants received US$25 upon the interview’s completion [44].
2.2. Variable Construction and Model Specification
The final data set contained a sample of 205 Puerto Rican mothers of children between the ages
of 10 to 19 years.
years. Survey
Survey questions
questions used
used for
for variable
variable construction
construction are
are included
included in
in Table
Table 1. Of this
sample, three mothers did not provide complete social network information, and these observations
were dropped from this analysis. Similar to previous research, the dependent variable social capital
was generated through two questions addressing social network reciprocity [3,20]. A mother was
prompted to identify up to five people she had spoken
spoken with
with regarding
regarding problems
problems over
over the
the last
last year,
year,
and these people were considered to be her identified social network. A mother was asked how many
times she received help from each identified person in her network, and how many times she gave
help to each person. The dependent variable, SKA, used responses to these questions and was coded
to receive aa value
support
to to
anan
individual
in her
social
network
and and
if she
valueof
ofone
one(1)
(1)ififa amother
mothergave
gave
support
individual
in her
social
network
if
relied
on aon
person
for similar
support.
Affirmative
responses
to both
questions
metmet
thethe
definition
of
she relied
a person
for similar
support.
Affirmative
responses
to both
questions
definition
network
reciprocity
necessary
forforaccumulating
of network
reciprocity
necessary
accumulatingsocial
socialcapital.
capital.The
Thedependent
dependentvariable
variable received
received a
gave or
or received
received support,
support, or
or gave
gave and
and received
received no
no support.
support.
value of zero (0) if a mother gave
The model was specified to estimate five individual and demographic characteristics:
characteristics: a mother’s
age, marital status, educational attainment, employment, and income. Age was a continuous variable
that ranged from 20 to 73
73 years.
years. Marriage was a dichotomous variable defined as being currently
married and coded to receive a value of one (1) and received a value of zero (0) if a mother was
widowed, separated,
separated, divorced,
divorced, or
or never
never married.
married. A mother’s educational attainment was categorized
widowed,
as (1) less
oror
itsits
equivalent;
(3)(3)
some
college;
or
less than
than aahigh
highschool
schooldiploma;
diploma;(2)
(2)high
highschool
schooldiploma
diploma
equivalent;
some
college;
(4)
a
bachelor
degree
or
higher.
Those
with
less
than
a
high
school
education
are
the
comparison
or (4) a bachelor degree or higher. Those with less than a high school education are
group. The model also controlled for labor force characteristics: a mother being employed and her
income. Employed was a dichotomous variable receiving
receiving a value of one (1) if a mother was employed
and a zero (0) if she was unemployed or not in the labor force. Income
Income was
was measured
measured subjectively.
subjectively.
Mothers
wereasked
asked“in
“ingeneral,
general,would
would
you
family
living
you) more
have
Mothers were
you
saysay
thatthat
youyou
(and(and
youryour
family
living
with with
you) have
more
money
than
you
need,
just
the
amount
of
money
you
need,
or
not
enough
money
to
cover
your
money than you need, just the amount of money you need, or not enough money to cover your needs?”
needs?”
If
their
income
was
more
than
needed
or
just
the
amount
needed,
the
income
variable
was
If their income was more than needed or just the amount needed, the income variable was coded
to
coded
haveofaone
value
oneif(1),
if thewas
income
was nottoenough
to coverexpenses,
needed expenses,
it was
have a to
value
(1),ofand
theand
income
not enough
cover needed
it was coded
to
coded
to
have
a
value
of
zero
(0).
Five
mothers
refused
to
answer
this
income
question,
and
they
were
have a value of zero (0). Five mothers refused to answer this income question, and they were dropped
dropped
this analysis.
from this from
analysis.
Societies 2017, 7, 3
6 of 13
Table 1. Survey Questions Used for Variable Construction.
Variables
Survey Questions
Social Capital
Accumulation
(SKA = Reciprocity)
Received: In the last 12 months, how many times have you helped [SOCIAL
NETWORK MEMBER NAME] with problems of transportation, family, finances,
health, home, or with some other kind of problem?
Provided: How many times has [SOCIAL NETWORK MEMBER NAME] helped
you with problems of transportation, family, finances, health, home, or with some
other kind of problem?
Individual
Demographics
Age: How old are you?
Marital Status: Are you married, divorced, separated, widowed, or never married?
Education: What grade did you reach in school?
Employment: Are you currently employed and receiving a salary, either full-time
or part-time?
Income: In general, would you say that you (and your family living with you) have
more money than you need, just the amount of money you need, or not enough
money to cover your needs?
Neighborhood
and Network
Attends Child’s Activities: In the last 12 months, have you attended a sports event,
concert, play, or art exhibition outside of school for your child?
Length of time in neighborhood: How many years, more or less, have you lived in
this neighborhood?
Race of SN member: What is the race or ethnic origin of [social network member] is
he/she Hispanic or Latino/a, white, black, Asian, or something else?
Perceived Collective
Efficacy
Perceived Collective Efficacy: If a group of neighborhood children misbehave, skip
school and hang out on a street corner, how many of your close neighbors do you
think would do something about it—more than half, a few, less than half, or none?
The model estimated coefficients for seven neighborhood and social network characteristics:
perceived collective efficacy, trust, civic engagement, race of social network members, time spent,
and percentage of Latinos in the neighborhood. Collective efficacy was defined as the perception of
people doing something about a problem in their neighborhood. Mothers were asked, “If a group of
neighborhood children misbehave, skip school and hang out on a street corner, how many of your
close neighbors do you think would do something about it—more than half, a few, less than half, or
none?” The collective efficacy variable was coded to receive a value of one (1) if a mother responded
with more than half or a few and received a zero (0) if a mother responded with half or none. Three
mothers failed to answer this question, and they were dropped from the analysis. Trust addressed
mothers trusting their neighbors. A mother was asked, “how much do you trust your neighbors—very
much, some, a little, or not at all.” The trust variable was coded to receive a value of one (1) if a
mother responded with very much or some, and received a zero (0) if she responded with a little or
not at all. Civic engagement was defined by participation in activities outside of the home, and this
model controlled for a mother’s participation in her child’s activities. Mothers were asked, “In the
last 12 months, had you attended activities like a sports event, a concert, a play, or an art exhibition
for your child?” This civic engagement variable was coded to receive a value of one (1) if a mother
responded yes, and a zero (0) if she had not. Mothers were asked how long they had lived in their
current neighborhood. This time-in-neighborhood variable is a continuous variable measuring years
and ranges from 1 to 36 years. Two mothers did not answer this question, and they were dropped from
the analysis. To measure for Latino concentration in a mother’s neighborhood, 2009–2013 American
Community Survey data were used to estimate the percentage of a mother’s census block who were
Latino. This percentage is a continuous variable ranging from 25 to 100 percent. The social network
variable measured a social network member’s race. Mothers were asked if an identified social network
Societies 2017, 7, 3
Societies 2017, 7, 3
7 of 13
7 of 13
network variable measured a social network member’s race. Mothers were asked if an identified
social network
member
Hispanic
or black,
Latino/a,
white,
black, Asian,
something
else,received
and this
member
was Hispanic
or was
Latino/a,
white,
Asian,
or something
else,orand
this variable
variable
received
a
value
of
one
(1)
if
any
social
network
member
was
white
and
was
coded
as
a zero
a value of one (1) if any social network member was white and was coded as a zero (0) if all social
(0) if all social
network
were
Hispanic
or Latino/a,
Asian,else.
or something else.
network
members
were members
Hispanic or
Latino/a,
black,
Asian, orblack,
something
A
logistic
regression
model
was
specified
and
results
using
Stata
12
were generated
generated to
to test
test
A logistic regression model was specified and results using Stata 12 were
hypothesis
regarding
a
mother’s
social
capital
accumulation
with
individual,
neighborhood,
and
hypothesis regarding a mother’s social capital accumulation with individual, neighborhood, and social
social network
characteristics.
The logistic
regression
equation
is below.
network
characteristics.
The logistic
regression
equation
is below.
b3(educational attainment)
= b0 + b0 b1(age)
+ b2(married)
) + e+
b3(educational attainment)
SKA = e + b4(employed)
eb1(age) + eb2(married
b5(income)
b6(child’sactivity)
+ +eb4(employed+
) + eb5(income) + e+
b6(child0s activity)
b8(percentageLatino)
) + eb8(percentage Latino)
+ b7(time)++eb7
(time
b9(percentagehomeowner)
b10(sn_race)
b9
(
percentage
homeowner)+
(sn_race)
+
+e
+ eb10
)
b12(trust)
+eb11(collective efficacy
eb12(trust) +
+ eέ
+ b11(collective
+ )+
(1)
Due to
tomissing
missingdata,
data,the
thefinal
finaldata
dataset
setfor
forthis
thisanalysis
analysiscontained
contained192
192observations.
observations.Inversed
Inversed
Due
probability
weighting
produced
population-based
estimates.
Robust
standard
errors
were
generated
probability weighting produced population-based estimates. Robust standard errors were generated
toaccount
accountfor
forrespondents’
respondents’clustering
clusteringwithin
withincensus
censusblocks.
blocks.The
Theoverall
overallrate
rateof
ofcorrect
correctclassification
classification
to
was
estimated
to
be
65.8%.
was estimated to be 65.8%.
Results
3.3.Results
This research
research tested
tested hypotheses
hypotheses regarding
regarding the
the association
association of
of SKA
SKA with
with demographic
demographic and
and
This
individualcharacteristics
characteristicsalong
alongwith
withneighborhood
neighborhoodand
andsocial
socialnetwork
networkcharacteristics
characteristicsamong
amongPuerto
Puerto
individual
Rican mothers
mothers living
living in neighborhoods with
lived
in
Rican
with high
high concentrations
concentrationsof
ofLatinos.
Latinos.These
Thesemothers
mothers
lived
census
blocks
where
71.6
percent
(SD(SD
18.6)
of their
neighbors
were
Latino.
Nearly
two-thirds
(64.9
in
census
blocks
where
71.6
percent
18.6)
of their
neighbors
were
Latino.
Nearly
two-thirds
percent)
of the
Rican
mothers
hadhad
reciprocal
social
network
relationships
that
were
used
in
(64.9
percent)
of Puerto
the Puerto
Rican
mothers
reciprocal
social
network
relationships
that
were
used
creating
Table 22 suggested
suggested
in
creatingthe
thedependent
dependentvariable
variableSKA.
SKA.Descriptive
Descriptivestatistics
statisticsand
and bivariate
bivariate results in Table
that attending
attending activities
activities of
of their
their children ((Ϫ 2 =2 =7.242,
7.242,p p==0.007)
0.007)and
andliving
livingininneighborhoods
neighborhoodswith
with
that
highershares
sharesof
ofLatinos
Latinos(t(t==2.1898,
2.1898,pp== 0.0298)
0.0298) were
werestatistically
statisticallysignificantly
significantlyassociated
associatedwith
withPuerto
Puerto
higher
Ricanmothers
mothersin
inSpringfield
Springfieldhaving
havingincreased
increasedSKA.
SKA.
Rican
After controlling for vectors of five individual variables and of seven neighborhood and social
Table
2. Descriptive
Statistics
of Explanatory
Variablesevidence
Used in Logistic
Regressionneighborhood,
of Puerto Rican and
network
variables,
logistic
regression
results provided
that individual,
Mothers’ Socialcharacteristics
Capital Accumulation.
civic participation
were associated with SKA. In addition to the bivariate results of
attending events for their children and living in neighborhoods with
higher
Mean
High concentrations
Mean Low of Latinos,
Variable
Definition
H0
Test Statistic
Socialpositively
Capital
Social
Capital with SKA.
this model also suggested that employed Puerto Rican mothers were
correlated
Dependent Variable
Tests for interaction effects between employment and having a non-Latino white social network
SKA
Reciprocity in social networks
0.649
0.351
memberand
were
conducted,
but results were not strong enough (p = 0.084) to be statistically significant at
Individual
Demographic
Characteristics
the
Age0.05 level.
Years since birth
+
40.5
40.8
t = 0.9741
2 = 0.4485
Married
Married
at
the
time
of
the
survey
+
0.287
0.370
Ϫ
Assuming we have estimated causal rather than associational relationships, what were
the
Education
+
Ϫ2 = 1.4359
probabilistic effects of these three statistically significant variables on social capital? Figure 2 converts
Less than High
Up to high school but no diploma
0.440
0.427
these
coefficients
(Column High
2 of school
Tablediploma
3) intoorprobabilities
and reports 0.290
them on a bar0.365
chart. Specifically,
High
School
equivalent
Some
College
Attended
college but
Bachelor’s
degree
0.130
being
employed
or attending
activities
ofno
their
children
was estimated 0.214
to have independently
increased
Degree that a Puerto Rican
Bachelor’s
degree
of higher
0.056
0.078
theCollege
probability
mother
had
accumulated social capital
by about 13.7
and 14.0 percent
Employed
Employed at the time of the survey
+
0.479
0.343
Ϫ2 = 2.1541
respectively.
In
other
words,
Puerto
Rican
mothers
with
a
job
were
13.7
percent
more
likely
than
Income
More or right amount of money to cover needs
+
0.479
0.599
unemployed
Rican mothers
to have accumulated social capital, and Puerto Rican mothers who
Neighborhood
andPuerto
Social Network
Characteristics
Attended a sports event, concert, play, or art
attended
activities
than
those who did
any *
Attends
Child’s
Activities of their children were 14 percent more likely
+
0.673
0.458not attend
Ϫ2 = 7.242
exhibition for your child
activities to have accumulated social capital. Furthermore, one standard deviation increase in the
Time in Neighborhood
Years Living in Neighborhood
+
7.40
6.07
t = −1.8813
percentage
of Latinos in a mother’s
(e.g., an 18.6+percent0.738
rise) augmented
Percentage Latino
Percentage census
Latino in block
census block
0.672the probability
t = −2.1898 *
Percentage
Homeowner
Percentage
homeowners
in census block
+
0.116
0.120
t = −0.7182
of accumulating
social capital
byof12.3
percent.
Race—Social Network
Member
Perceived Collective
Efficacy
Trust
* p < 0.05
Has Non-Latino white in social network
member
Percentage of close neighbors who would
intervene in a specific neighborhood problem
Trusts neighbors some or very much
Unweighted N = 192
+
0.260
0.216
Ϫ2 = 0.830
+
0.479
0.432
Ϫ2 = 0.4206
+
0.417
N = 3590
0.364
N = 1920
Ϫ2 = 2.1623
network variable measured a social
network member’s
race. Mothers
w
b9(percentagehomeowner)
A logistic regression model
was specified
and results using
Stata
+ b10(sn_
+ b7(time)
+ b8(percentageLatino)
social network member was Hispanic or Latino/a, white, black, Asian, o
b11(collective
)
b12(trust)
b9(percentagehomeowner)
b10(sn_
hypothesis regarding a mother’s
social capital accumulation
indiv
+ (1)
+
+with
variable received a value of one
if any social network+
member
was whi
social network characteristics. The
logistic regression equation
is below.
b11(collective
)
b12(trust)
Societies
3network
+ final
+ black,
+1
(0)
if Due
all2017,
social
members
were
Hispanic
Latino/a,
Asian,
to7,missing
data,
the
data
set for or
this
analysis
contained
b0
b1(age)
b2(married)
b3(educatio
A
logistic
regression
was
specified
results
using Stata
= data,
+model
final
+set for and
probability
produced
population-based
stand
Due
to7,weighting
missing
the
data
this
1
Societies 2017, 7, 3
8estimates.
of analysis
13 + Robust
Societies
2017,
3
network
variable
measured
a social
network
member’s
race.contained
Mothers
w
b4(employed)
b5(income)
b6(ch
hypothesis
regarding
a
mother’s
social
capital
accumulation
with
indiv
to
account
for
respondents’
clustering
within
census
blocks.
The
overall
r
+
+
+
probability
weighting
produced
population-based
Robust
stand
socialnetwork
network
member
was Hispanic
or regression
Latino/a,estimates.
white,
black,
Asian,
o
social
characteristics.
logistic
equation
is
below.
Societies
2017,
7, 3 to
b7(time)
b8(percentageLatino)
was
estimated
be
65.8%. +
network
variable
measured
aThe
social
network
member’s
race.
Mothers
w
to
account
for
respondents’
clustering
within
census
blocks.
The
overall
r
+
variable
received
a
value
of
one
(1)
if
any
social
network
member
was
wh
Table 2. Descriptive Statistics of Explanatory Variables Used in Logistic Regression of Puerto Rican Mothers’ Social Capital Accumulation.
social
networkto
member
or
white,+black,
Asian,
o
b9(percentagehomeowner)
b10(sn_
b0was
b2(married)
was
be 65.8%.
b1(age)
+b3(educatio
=
++
aHispanic
+ Latino/a,
(0)Results
ifestimated
all social
network
members
were
Hispanic
or Latino/a,
black,
Asianw
network
variable
measured
social
network
member’s
race.
Mothers
3.
variable
received a value of one (1)
if
any
social
network
member
was
whi
b11(collective
)resultsb12(trust)
b4(employed)
b5(income)
b6(ch
Anetwork
logistic
regression
model
was or
specified
using
Stata
+ black,
+o
Mean
High
Mean
+Hispanic
Low
+ and
+Asian,
member
was
Latino/a,
white,
Test
Statistic
Variable
Definition
H0social
3. Results
(0)
if This
all
social
network
members
were
Hispanic
or
Latino/a,
black,
Asian,
research
tested
hypotheses
regarding
the
association
of
SKA
Social
Capital
Social
Capital
b7(time)
b8(percentageLatino)
hypothesis
regarding
a
mother’s
social
capital
accumulation
with
indiv
variable
received
a value
one
(1) ifdata
any social
member
was whi1
+
final
+ fornetwork
Due
tocharacteristics
missing
data,of
the
set
this
analysis
contained
A
logistic
regression
model
was
specified
andand
results
Stata
individual
along
with
neighborhood
socialusing
network
ch
This
research
tested
hypotheses
regarding
the
association
of SKA
social
network
characteristics.
The
logistic
regression
equation
is below.
Dependent Variable
b9(percentagehomeowner)
b10(sn_
(0)
if
all
social
network
members
were
Hispanic
or
Latino/a,
black,
Asian,
+
+
probability
weighting
produced
population-based
estimates.
Robust
stand
hypothesis
regarding
a
mother’s
social
capital
accumulation
with
indiv
Rican mothers
living in neighborhoods
with high concentrations
of Latin
individual
characteristics
along
with
neighborhood
and
social
network
ch
A logistic
regression
model
was within
specified
and blocks.
results
using
Stata
SKA
Reciprocity in social networks
0.649
0.351
b11(collective
)
b12(trust)
b0 +
b1(age)
b2(married)
b3(educati
to
account
for where
respondents’
clustering
census
The
overall
r
social
network
characteristics.
The
logistic
regression
equation
is
below.
+
+
+
=
+
+
census
blocks
71.6
percent
(SD
18.6)
of
their
neighbors
were
Latin
Rican mothers
living ina neighborhoods
with
high
concentrations
of Latin
hypothesis
regarding
mother’s
social
capital
accumulation
with
indiv
Individual and Demographic Characteristics
b4(employed)
b5(income)
b6(c
was
estimated
to
be 65.8%.
percent)
of
Puerto
Rican
mothers
had
reciprocal
social network
relati
final(SD
+thisneighbors
+
b0
percent
b1(age)
b2(married)
b3(educatio
census
blocks
where
71.6
18.6)
of
were
Latin1
Due
tothe
missing
data
fortheir
analysis
social
network
characteristics.
The
logistic
regression
equation
is
below.
= data,
+the
+
+set
+ contained
Age
Years since birth
+ creating the40.5
40.8
t
=
0.9741
b7(time)
b8(percentageLatino)
dependent
variable
SKA.
Descriptive
statistics
and
bivariate
r
percent)
of
the
Puerto
Rican
mothers
had
reciprocal
social
network
relati
+
+
probability
weighting produced
population-based
Robust
stand
b4(employed) + estimates.
b5(income)
b6(ch
+
b1(age)
+and
3.
Results
b0of
their
b3(educatio
2 b2(married)
2 0.4485
Married
Married at the time of the survey
+ to
0.287
0.370
that
attending
activities
children
(Ϫ
=
7.242,
p
=
0.007)
livi
=
=
+
+
+
b9(percentagehomeowner)
b10(sn
creating
thefor
dependent
variable
Descriptive
statistics
account
respondents’
clustering
within census
blocks.and
The
overall r
+ SKA.
+bivariate
b7(time)
b8(percentageLatino)
+
+2=2 were
b4(employed)
b5(income)
higher
shares
of
Latinos
(t =hypotheses
2.1898,
p = 0.0298)
statistically
significan
This
research
regarding
the
ofb6(ch
SKA
Education
+ was
that
attending
activities
of
their
children
(Ϫ
=
7.242,
p
=
0.007)
and
livi
1.4359
+
+
+
b11(collective
)association
b12(trust)
estimated
to
be tested
65.8%.
+
+
+
b9(percentagehomeowner)
b10(sn_
+
+
Rican mothers
inLatinos
Springfield
having
increased
SKA.
b7(time)
b8(percentageLatino)
individual
characteristics
along
with
neighborhood
and
social
network
ch
higher
shares
of
(t
=
2.1898,
p
=
0.0298)
were
statistically
significan
Less than High
Up to high school but no diploma
0.440
0.427
+
b11(collective
+
)
Due
to missing
data,
the
data
set high
for
this
analysis
contained
Rican
mothers
living
in
neighborhoods
with
concentrations
of
Latin
+
final
+ b12(trust)
+
3.
Results
in
Springfield
having
increased
SKA.
b9(percentagehomeowner)
b10(sn_
High School
High school diploma or equivalent
0.290
0.365
+
+
Table
2.
Descriptive
Statistics
of
Explanatory
Variables
Used
in
Logistic
Reg
probability
weighting
produced
population-based
estimates.
Robust
stan
census blocks where 71.6 percent
(SD 18.6) of their neighbors
were Latin
b11(collective
)association
b12(trust)
Due to
missing
data, Accumulation.
the
data
set for this
contained
This
research
tested
hypotheses
regarding
theanalysis
of SKA
Some College
Attended college but no Bachelor’s degree
0.130
+
final
+ network
+1r
Mothers’
Social
Capital
to account
respondents’
clustering
within
census social
blocks.
The
overall
Table
2.for
Descriptive
Statistics
of Explanatory
Variables
Used
in
Logistic
Reg
percent)
of0.214
the
Puerto
Rican
mothers
had
reciprocal
relati
probability0.056
weighting produced
population-based
stand
individual
characteristics
along
with
neighborhoodestimates.
and socialRobust
network
ch
College Degree
Bachelor’s degree of higher
0.078
was Mothers’
estimated
to beCapital
65.8%.
Social
Accumulation.
creating
the
variable
SKA.data
Descriptive
statistics
andcontained
bivariate
r1
Mean High
Due
to dependent
missing
data,
the
final
set for this
analysis
Variable
Definition
H0The overall
to account
for respondents’
clustering
within
census
blocks.
r
Rican
mothers
living in neighborhoods
with2 high
concentrations
of
Latin
Social
Capit
2 2.1541
Employed
Employed at the time of the survey
+ that
0.479
0.343
attending
activities
of
their
children
(Ϫ
=
7.242,
p
=
0.007)
and
livi
=
probability weighting produced population-based estimates. Robust
Meanstand
High
was
estimated
to
be
65.8%.
census
blocks
where
71.6
percent
(SD
18.6)
of
their
neighbors
were
Latin
Dependent
Variable
Variable
Definition
H0
3. account
Results
higher
shares
Latinos (t = clustering
2.1898, p =within
0.0298)census
were statistically
significan
Capitr
to
for of
respondents’
blocks. TheSocial
overall
Income
More or right amount of money to cover needs
+SKA
Reciprocityhad
in social
networks social network0.649
percent)Variable
of0.479
the Puerto Rican0.599
mothers
reciprocal
relati
Dependent
Rican
mothers
in
Springfield
having
increased
SKA.
was
estimated
to
be
65.8%.
Individual
and Demographic
Characteristics
This
tested
hypotheses
regardingstatistics
the association
of SK
Neighborhood and Social Network Characteristics
3. Results
creating
theresearch
dependent
variable
SKA. Descriptive
and bivariate
SKA
Reciprocity
in social networks
0.649 r
Age
Years since birth
+
40.5
individual
characteristics
along
with
neighborhood
and
social
network
ch
2
and 0.673
Demographic
Characteristics
Attends Child’s Activities
Attended a sports event, concert, play, or art exhibition for your child
+Individual
0.458of
that
attending
activities
of Married
their
children
(Ϫthe=2survey
= 7.242,
p = 0.007)
and
livi
7.242
* association
Married
at
the
time
of
+
0.287
This
research
tested
hypotheses
regarding
the
of
SKA
Table
2.
Descriptive
Statistics
Explanatory
Variables
Used
in
Logistic
Reg
3.
Results
Age
Years
since
birth
+
40.5
Rican
mothers
in (t
neighborhoods
with high
concentrations
of Latin
Education
+
shares
ofliving
Latinos
= 2.1898,
p = 0.0298)
were
statistically
significan
Time in Neighborhood
Years Living in Neighborhood
+ higher
7.40
6.07with
= −survey
1.8813and
Mothers’
Social
Capital
Accumulation.
individual
characteristics
along
social
Married
Married
at theneighborhood
time oft the
+ network
0.287ch
census
blocks
where
71.6
percent
(SD
18.6)
of
their
neighbors
were
Latin
Less
than
High
Up
to
high
school
but
no
diploma
0.440
This
research
tested
hypotheses
regarding
the
association
of
SKA
Rican
mothers
in
Springfield
having
increased
SKA.
Education
+
with
concentrations
of0.290
Latin
Percentage Latino
Percentage Latino in census block
+ Rican
0.738 living in neighborhoods
0.672
tor=equivalent
−high
2.1898
*
High mothers
School
High
school diploma
Mean
High
percent)
of
the
Puerto
Rican
mothers
had
reciprocal
social
network
relat
individual
characteristics along
with
neighborhood
Definition
H0 network
Less Variable
than High
Up to high
school
but no diploma and social
0.440ch
blocks
where 71.6
percent
(SD
of−their
neighbors
were
Latin
Social
Capit
Some College
Attended
college
but18.6)
no Bachelor’s
degree
0.214
Percentage Homeowner
Percentage of homeowners in census block
+ census
0.116
0.120
t =equivalent
0.7182
creating
the
dependent
variable
SKA.
Descriptive
statistics
and
bivariate
High
School
High
school
diploma
or
0.290
Table
2.
Descriptive
Statistics
of
Explanatory
Variables
Used
in
Logistic
Regr
Rican
mothers
living
in
neighborhoods
with
high
concentrations
of
Latin
CollegeVariable
Degree
Bachelor’s
degree
of
higher
0.056
Dependent
percent)
of
the
Puerto
Rican
mothers
had
reciprocal
social
network
relati
2
2
Some
College
Attended
college
but
no
Bachelor’s
degree
0.214
Race—Social Network Member
Has Non-Latino white in social network member
+SKA
0.260
0.216
that
attending
activities
their
children
(Ϫ
= 7.242,
p = +0.007)
and
liv
=their
0.830
Mothers’
Social
Capital
Accumulation.
census
blocks
where
71.6of
percent
(SD
18.6)
ofnetworks
neighbors
were
Latin
Employed
Employed
at
the time
of the
survey
0.479
Reciprocity
in social
0.649
creating
the dependent variable
SKA.degree
Descriptive
College Degree
Bachelor’s
of higher statistics and bivariate
0.056 r
Income
More
or
right
amount
of
money
to
cover
needs
+
0.479
Individual
and
Demographic
Characteristics
higher
shares
of
Latinos
(t
=
2.1898,
p
=
0.0298)
were
statistically
significa
percent)
of the Puerto Rican
mothers
reciprocal
social network
relati
Percentage of close neighbors who would intervene in a specific
2the
Employed
Employed
at thehad
time(Ϫ
of
+
0.479
Mean
High
Perceived Collective Efficacy
+Neighborhood
0.479
0.432children
that
activities
their
= 7.242, p = H0.007)
and
=2 survey
0.4206
and
Social
Networkof
Characteristics
Age attending
Years
since birth
+0
40.5livi
Variable
Definition
neighborhood problem
Rican
in Springfield
having
Social
Capitr
creating
the dependent
variable
SKA.increased
Descriptive
statistics
bivariate
Income mothers
More
or right
amount
of money to SKA.
cover
needs and
+
0.479
Attended
a sports
concert,
play, orstatistically
art
Married shares of Latinos
at the
of the survey
+
0.287
higher
(t Married
= 2.1898,
pevent,
=time
0.0298)
were
significan
Attends
Child’s
Activities
+
0.673
Variable
Neighborhood
and
Social
Networkof
Characteristics
Trust
Trusts neighbors some or very much
+Dependent
0.417
0.364
that
attending
activities
their
children
(Ϫ 2child
= 7.242, p = 0.007)
and
livi
=2 2.1623
exhibition
for
your
Education
+
RicanTable
mothers
in
Springfield
having
increased
SKA.
SKA
Reciprocity
in
social
networks
0.649
Attended
a
sports
event,
concert,
play,
or
art
2.
Descriptive
Statistics
of
Explanatory
Variables
Used
in
Logistic
Re
Time
in
Neighborhood
Years
Living
in
Neighborhood
+
7.40
Attends
Child’s
Activities
0.673
Less
than
High
Up
to
high
school
but
no
diploma
0.440
higher shares
of Latinos
(t N
= 2.1898,
0.0298) were statistically significan
* p < 0.05
Unweighted N = 192
N Demographic
= 3590
=exhibition
1920 p =
for your child
Individual and
Characteristics
Mothers’
Accumulation.
Percentage
Latino Social CapitalHigh
Percentage
Latino
in or
census
block
+
0.738
High
School
school
diploma
equivalent
0.290
Rican
mothers
in Springfield
having
increased
SKA. Used+ in Logistic
Time
Neighborhood
Years
Living
in Neighborhood
7.40Reg
Agein Table
Years
since
birth Variables
40.5
2. Descriptive
Statistics
of
Explanatory
Percentage
Homeowner
Percentage
of homeowners
in censusdegree
block
+
0.116
Some College
Attended college
but no Bachelor’s
0.214
Percentage
Percentage
Latino
block
0.738
Married Latino
Married at the
timeinofcensus
the survey
++
0.287
Mean
Hig
Mothers’
Social Capital
Race—Social
Network
HasAccumulation.
Non-Latino
white
network
College
Degree
Bachelor’s
degreeinofsocial
higher
0.056
Variable
Definition
H0
0.260
Percentage
Homeowner
Percentage
of
homeowners
in
census
block
+
0.116
Social Capi
Education
Table
2.
Descriptive
Statistics
of
Explanatory
Variables
Used
in
Logistic
Member
member
Employed
Employed at the
time of the survey
+
0.479Reg
Race—Social
Network
HasUp
Non-Latino
whitebut
in no
social
network
Dependent
Variable
Less than
High
to high school
diploma
0.440
Mean
High
Perceived
Collective
Percentage
of
close
neighbors
who
would
Mothers’
Social
Capital
Accumulation.
+
0.260
Income
More
or
right
amount
of
money
to
cover
needs
0.479
Variable
Definition
H
+0
0.479
Member
member
SKA
Reciprocity
in social
networks
0.649
High
School
High
school
diploma
or
equivalent
0.290
Social
Capit
Efficacy
intervene
in a specific neighborhood problem
Neighborhood
and Social Network
Characteristics
Perceived
Collective
Percentage
of close
who degree
would
Individual
and Demographic Characteristics
Some College
Attended
college
butneighbors
no Bachelor’s
0.214
Dependent
Variable
Mean
High
Trust
Trustsaneighbors
some
or veryplay,
much
+0
0.417
0.479
Attended
sportsDefinition
event,
concert,
or art
Variable
H
Efficacy
interveneBachelor’s
in
a specific
neighborhood
problem
Attends
Child’s
Activities
++
0.673
Age
Yearsdegree
since
birth
40.5
College
Degree
of
higher
0.056
SKA
Reciprocity
in
social
networks
Social
* p < 0.05
Unweighted
N = 192
N0.649
= Capit
3590
exhibition
for your
child
Trust
Trusts
neighbors
orthe
very
much
++
0.417
Marriedand
Married
at
time
survey
0.287
Employed
Employed
at the
the some
time of
of
the
survey
0.479
Individual
Demographic Characteristics
Dependent
Variable
Time in Neighborhood
Years Living in Neighborhood
+
7.40
* pIncome
<Education
0.05
Unweighted
Nbirth
= 192
N0.479
= 3590
More or right
amount
ofinmoney
to cover needs
++
Age
Years since
40.5
SKA
Reciprocity
social
networks
0.649
Percentage Latino
Percentage Latino in census block
+
0.738
Less than
Up to high
school
no diploma
0.440
Neighborhood
and
Social Network
Characteristics
Married
Married
at the
timebut
of the
survey
+
0.287
Individual
andHigh
Demographic
Characteristics
Percentage Homeowner
Percentage of homeowners in census block
+
0.116
High School
Higha school
diploma
or equivalent
0.290
Attended
sports
event,
play, or art
Education
+
Age
Years
sinceconcert,
birth
40.5
Race—Social
Network
Has Non-Latino white in social network
Attends
Child’s
Activities
+
0.673
Societies 2017, 7, 3
9 of 13
Table 3. Logistic Results of Puerto Rican Mothers’ Social Capital Accumulation.
Variable
Coefficient
95% Confidence Interval
Individual and Demographic Characteristics
Age
Married
Education
Less than High
High School
Some College
College Degree
Employed
Income
−0.0092882
−0.5562095
−0.0381712–0.0195947
−1.416847–0.3044283
−0.3528094
0.0301
−0.840606
0.6047695 **
−0.7210817
−0.979637–0.2740181
−0.7620988–0.8222988
−2.318145–0.6369326
0.216603–0.992936
−1.455572–0.0134082
Neighborhood and Social Network Characteristics
Attends Child’s Activities
Years Living in Neighborhood
Percentage Latino
Percentage Homeowner
Race of Social Network Member
Perceived Collective Efficacy
Trust
Constant
0.615501 *
0.0159491
2.86277 **
1.643363
0.055652
0.097684
0.3114251
0.0866646–1.144337
−0.042258–0.0741562
0.8713682–4.854173
−0.4618607–3.748587
−1.074326–1.18563
−0.7300011–0.9253692
−0.5597921–1.182642
−1.543374
−3.525166–0.4384189
* p < 0.05; ** p < 0.01.
Societies 2017, 7, 3
9 of 13
15%
13.7%
14.0%
Employed
Attended Child's
Activities
12.3%
10%
5%
0%
Percentage Latino in
Census Block
Figure 2. The Probability of Having Engaged in Social Capital Accumulation.
Figure 2. The Probability of Having Engaged in Social Capital Accumulation.
4. Discussion
4. Discussion
This research helped fill a gap in the literature by identifying mechanisms through which
This research helped fill a gap in the literature by identifying mechanisms through which Puerto
Puerto Rican mothers accumulated social capital. This research also supports existing literature
Rican mothers accumulated social capital. This research also supports existing literature that
that highlights how residentially segregated neighborhoods and the social organization shaped by
highlights how residentially segregated neighborhoods and the social organization shaped by them
them influenced the development of social ties necessary for Puerto Rican mothers to accumulate social
influenced the development of social ties necessary for Puerto Rican mothers to accumulate social
capital in Springfield, MA in 2013 [2,4,34]. Interaction with individuals from a higher socioeconomic
capital in Springfield, MA in 2013 [2,4,34]. Interaction with individuals from a higher socioeconomic
status can facilitate the transfer of more advantageous economic, political, and social resources and
status can facilitate the transfer of more advantageous economic, political, and social resources and
information [46,47]. SKA’s positive association with percentages of Latinos in a mother’s neighborhood
information [46,47]. SKA’s positive association with percentages of Latinos in a mother’s
suggested strong ties were being developed. Homophilous relationships shaped not only where a
neighborhood suggested strong ties were being developed. Homophilous relationships shaped not
only where a mother lived but also with whom she interacted; that is she socialized in residentially
segregated neighborhoods with people who were socioeconomically like herself [48]. Puerto Ricans
were the largest Latino subpopulation in Springfield, and these mothers accumulated social capital
in neighborhoods with more concentrated Latino populations. This segregation effect suggested that
Societies 2017, 7, 3
10 of 13
mother lived but also with whom she interacted; that is she socialized in residentially segregated
neighborhoods with people who were socioeconomically like herself [48]. Puerto Ricans were
the largest Latino subpopulation in Springfield, and these mothers accumulated social capital in
neighborhoods with more concentrated Latino populations. This segregation effect suggested that
information and resources transmitted through social capital were shaped by the social organization in
these segregated neighborhoods. As Small’s research highlighted, this organization did not keep social
capital from being accumulated; it just shaped the resources and information available to Puerto Rican
mothers and subsequently the type of social capital accumulated [2,49]. This research also identified
that having a non-Latino white social network tie was not associated with SKA, and these relationships
could have been important for developing bridging social capital. Latinos’ median household income
was significantly lower (US$20,874) than non-Latino whites’ (US$48,420) [50]. Based on this income
distribution, information and resources transferred to Puerto Rican mothers by their accumulated
social capital can enhance their ability to manage their daily responsibilities but may not enhance their
ability to improve their socioeconomic standing [22,49]. Puerto Rican mothers developed reciprocal
relationships with non-Latinos, even with their lower socioeconomic standing. Further analysis of
our data showed that 23 mothers had non-Latino white social network members, and 87% (n = 20)
received support and 70% (n = 16) gave support. This suggested that any limitations in developing
cross ethno-racial social capital was more related to exposure to another population than to the lack of
the ability to develop reciprocity due to income or other resource disparities [51].
As has been found for Mexican migrants, employment was an important domain for accumulating
social capital [3]. Puerto Ricans’ unemployment rate in Springfield in 2010 was nearly 20 percent
compared to 13 percent for non-Latino whites [42]. Thus, labor market conditions could have a
detrimental effect on these mothers’ ability to accumulate social capital, but broader economic trends
could also be shaping this outcome. For example, unemployment and the decreased access to SKA
that it entails could also be related to a lack of child care when a mother’s children were younger [46].
The absence of child care could have limited a mother’s previous employment and development
of job skills, making her a less qualified job candidate despite her now having less dependent care
responsibility and more available time to work. To compensate, some mothers reported providing
services such as childcare and event food preparation in the informal economy. These mothers used
their social capital to develop jobs in the informal labor market, but this supplemental work could be
limited in its ability to provide access to weak ties similar to those developed in the formal labor market
that are important for developing bridging social capital [52]. Structural efforts to increase Puerto Rican
mothers’ participation in the labor market could provide both pecuniary and non-pecuniary benefits.
One strategy to overcome the effect of living in residentially segregated neighborhoods that
Puerto Rican mothers experienced was through civic participation [39,49]. Similar to other research,
these results suggested that mothers who attended events for their children were more likely to
develop reciprocity and trust and thus enhance their SKA [37]. This supported Small’s argument
that institutions (not only or nor primarily people’s own independent efforts) determine the extent
and nature of social capital that people develop. A mother's engagement with her child's activities
suggested that she benefitted from an array of institutional actions, arrangements, and activities
regardless of her individual characteristics [12]. When mothers attended events for their children, these
institutions provided access to opportunities. One of these opportunities was to interact with others
from outside of their regular social networks and possibly help strengthen relations to high-status
contacts and augment their social capital [37,39]. However, accumulated social capital can differ
based on the civic organization attended, as people in socially advantageous positions facilitate the
transfer of more desirable resources [47]. Community efforts to develop and strengthen institutions
for sustainable engagement should be a priority for planners and policy makers in Puerto Rican
neighborhoods [53].
Societies 2017, 7, 3
11 of 13
5. Conclusions
This research addresses a component of social capital whose process of accumulation is not well
understood. Using a sample of Puerto Rican mothers, this research tests hypotheses about these
mothers’ individual, demographic, neighborhood, and social network characteristics and found that
mothers who were employed, attended activities of their children, and lived in census blocks with
increased percentages of Latinos were more likely to accumulate social capital. Social capital can play
an important role in Puerto Rican mothers’ ability to access non-pecuniary resources to strengthen their
socioeconomic position. This research used cross-sectional data, and although these weighted results
allow for generalization to the Puerto Rican mothers of children between the ages of 10 to 19 years,
their generalization to Puerto Ricans in other cities should be done with caution. Future research is
needed to gather data on how these women perceive their SKA and to explore the information and
resources that are transferred through this SKA.
Acknowledgments: The research, Por Ahí Dicen, was funded by a 5-year NIH P60 award for the Center of Health
Equity Intervention Research (CHEIR) from the National Institute of Minority Health and Health Disparities
(NIMHD) #P60MD006912. The authors would like to acknowledge reviews and comments by Rosalyn Negrón
and Jim O’Brien along with anonymous reviewers that contributed to improve this manuscript.
Author Contributions: Maria Idalí Torres was the Principal Investigator and designed this study. Phillip Granberry
was an Investigator on the project. Both contributed to the fieldwork to obtain the data. Phillip Granberry
conducted the statistical analyses and drafted the manuscript. Both authors contributed to the interpretation of
data, commented on multiple drafts of the manuscript, and read and approved the final manuscript.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
Putnam, R. Bowling Alone: The Collapse and Revival of American Community; Simon & Shuster: New York, NY,
USA, 2000.
Small, M. Villa Victoria: The Transformation of Social Capital in a Boston Barrio; University of Chicago Press:
Chicago, IL, USA, 2004.
Granberry, P.; Marcelli, E. In the Hood and on the Job: Social Capital Accumulation among Legal and
Unauthorized Mexican Immigrants. Sociol. Perspect. 2007, 50, 579–595. [CrossRef]
Lichter, D.; Parisi, D.; Taquino, M.; Grice, S. Residential Segregation in New Hispanic Destinations: Cities,
Suburbs and Rural Areas Compared. Soc. Sci. Res. 2010, 39, 215–230. [CrossRef]
Portes, A. Social Capital: Its Origins and Applications for Modern Sociology. Am. Sociol. Rev. 1998, 24, 1–24.
[CrossRef]
Glaeser, E.; Laibson, D.; Sacerdote, B. An Economic Approach to Social Capital. Econ. J. 2002, 112, F437–F458.
[CrossRef]
Kumra, S.; Vinnicombe, S. Impressing for Success: A Gendered Analysis of Key Social Capital Accumulation
Strategy. Gender Work Organ. 2010, 17, 521–546. [CrossRef]
Stromsnes, K. The Importance of Church Attendance and Membership of Religious Voluntary Organizations
for the Formation of Social Capital. Soc. Compass 2008, 55, 478–496. [CrossRef]
Wollebaek, D.; Selle, P. Participation and Social Capital Formation: Norway in a Comparative Perspective.
Scand. Political Stud. 2003, 26, 67–91. [CrossRef]
Sampson, R.; Raudenbush, S.; Earls, F. Neighborhoods and Violent Crime: A Multi-Level Study of Collective
Efficacy. Science 1997, 277, 918–924. [CrossRef] [PubMed]
Menjívar, C. Fragmented Ties: Salvadoran Immigrant Networks in America; University of California Press:
Berkeley, CA, USA, 2000.
Bermúdez-Millán, A.; Damio, G.; Cruz, J.; D’Angelo, K.; Segura-Pérez, S.; Hromi-Fiedler, A.;
Pérez-Escamilla, R. Stress and the Social Determinants of Maternal Health among Puerto Rican Women: A
CBPR Approach. J. Health Care Poor Underserv. 2011, 22, 1315. [CrossRef] [PubMed]
Domínguez, S. Getting Ahead: Social Mobility, Public Housing, and Immigrant Networks; NYU Press: New York,
NY, USA, 2011.
Societies 2017, 7, 3
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
12 of 13
Torres, M.; Marquez, D.X.; Carbone, E.T.; Stacciarini, J.M.R.; Foster, J.W. Culturally Responsive Health
Promotion in Puerto Rican Communities: A Structuralist Approach. Health Promot. Pract. 2008, 9, 149–158.
[CrossRef] [PubMed]
Chavira-Prado, A. Latina Experience and Latina Identity. In Handbook of Hispanic Cultures in the United States:
Anthropology; Weaver, T., Ed.; Arte Público Press: Houston, TX, USA, 1994; pp. 244–269.
Alicea, M. A Chambered Nautilus: The Contradictory Nature of Puerto Rican Women’s Role in the Social
Construction of a Transnational Community. Gender Soc. 1997, 11, 597–626. [CrossRef]
Small, M. Unanticipated Gains; Oxford University Press: New York, NY, USA, 2009.
Heaney, C.A.; Israel, B.A. Social Networks and Social Support. Health Behav. Health Educ. Theory Res. Pract.
2008, 4, 189–210.
McPhererson, M.; Smith-Lovin, L.; Cook, J. Birds of a feather: Homophily in Social Networks. Annu. Rev. Sociol.
2001, 27, 415–444. [CrossRef]
Granberry, P. More Social than Capital: Social Capital Accumulation through Social Network Exchange
among Legal and Unauthorized Mexican Migrants. Soc. Sci. J. 2014, 51, 590–597. [CrossRef]
Granovetter, M. The Strength of Weak Ties. Am. J. Sociol. 1973, 78, 1360–1380. [CrossRef]
Stack, C. All Our Kin: Strategies for Survival in a Black Community; Harper: New York, NY, USA, 1975.
Larsen, L.; Harlan, S.; Bolin, B.; Hackett, E.; Hope, D.; Kirby, A.; Nelson, A.; Rex, T.; Wolf, S. Bonding and
Bridging: Understanding the Relationship between Social Capital and Civic Action. J. Plan. Educ. Res. 2004,
24, 64–75. [CrossRef]
Granberry, P.; Marcelli, E. Social Capital is Associated with Earnings among Foreign-born Mexican Men but
not Women in Los Angeles County. Int. Migr. 2011, 49, 113–128. [CrossRef]
Taylor, S. Tending Instinct: Women, Men, and the Biology of Relationships; Henry Holt: New York, NY, USA, 2002.
Moore, G. Structural Determinants of Men’s and Women’s Personal Networks. Am. Sociol. Rev. 1990, 55,
726–735. [CrossRef]
Fuhrer, R.; Stansfeld, S. How gender affects patterns of social relations and their impact on health: A
comparison of one or multiple sources of support from ‘close persons’. Soc. Sci. Med. 2002, 54, 811–825.
[CrossRef]
Martin, J.; Simmons, D.; Lu, S. The Role of Social Capital in the Experiences of Hispanic Women Engineering
Majors. J. Eng. Educ. 2013, 102, 227–243. [CrossRef]
Aguilera, M. The Impact of Social Capital on the Earnings of Puerto Rican Migrants. Sociol. Q. 2005, 46,
569–592. [CrossRef]
Putnam, R. E Pluribus Unum: Diversity and Community in the Twenty-first Century The 2006 Johan Skytte
Prize Lecture. Scand. Political Stud. 2007, 30, 137–174. [CrossRef]
Pendakur, R.; Mata, F. Social Capital Formation and Diversity: Impacts of Individual and Place-Related
Characteristics. J. Ethn. Migr. Stud. 2012, 38, 1491–1511. [CrossRef]
Logan, J. The Persistence of Segregation in the 21st Century Metropolis. City Community 2013, 13, 1–10.
[CrossRef] [PubMed]
Massey, D.; Denton, N. American Apartheid: Segregation and the Making of the Underclass; Harvard University
Press: Cambridge, MA, USA, 1993.
Wacquant, L.; Wilson, W. The Cost of Racial and Class Exclusion in the Inner City. Ann. Am. Acad. Political
Soc. Sci. 1989, 501, 8–26. [CrossRef]
Wilson, W. When Work Disappears: The World of the New Urban Poor; Knopf: New York, NY, USA, 1996.
Sacks, M. The Puerto Rican Effect on Hispanic Residential Segregation: A Study of the Hartford and
Springfield Metro Areas in National Perspective. Latino Stud. 2011, 9, 87–105. [CrossRef]
Wood, L.; Giles-Corti, B.; Zubrick, S.; Bulsara, M. Through the Kids . . . We Connected With Our Community:
Children as Catalysts of Social Capital. Environ. Behav. 2013, 45, 344–368. [CrossRef]
McPhererson, M.; Smith-Lovin, L. Homophily in Voluntary Organizations: Status Distance and the
Composition of Face-to-Face Groups. Annu. Sociol. Rev. 1987, 52, 370–379.
Benton, R. Uniters or Dividers? Voluntary Organizations and Social Capital Acquisition. Soc. Netw. 2016, 44,
209–218.
Green, M.; Brock, T. Organizational Membership versus Informal Interaction: Contributions to Skills and
Perceptions that Build Social Capital. Political Psychol. 2005, 26, 1–25. [CrossRef]
Societies 2017, 7, 3
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
13 of 13
Torres, M.I.; Granberry, P.; Person, S.; Allison, J.; Rosal, M.; Rustan, S. Influential Factors of Puerto Rican
Mother–Child Communication About Sexual Health Topics. Matern. Child Health J. 2016, 20, 2280–2290.
[CrossRef] [PubMed]
Granberry, P.; Rustan, S.; Karp, F. Latinos in Massachusetts Selected Areas: Springfield; Paper 176; Gastón Institute
Publications: Boston, MA, USA, 2013; Available online: http://scholarworks.umb.edu/cgi/viewcontent.
cgi?article=1177&context=gaston_pubs (accessed on 27 February 2017).
Silver, P.; Vargas-Ramos, P. Demographic Transitions; Centro RB2012-03; Center for Puerto Rican Studies:
New York, NY, USA, 2012; Available online: http://www.ahaa.org/Portals/0/Research/The%20Hispanic%
20Consumer/Demographics/Puerto%20Rican%20Demographic%20Transitions%202012.pdf (accessed on
27 February 2017).
Granberry, P.; Torres, M.I.; Allison, J.; Rosal, M.; Rustan, S.; Colón, M.; Fontes, M.; Cruz, I.; Puerto
Rican Cultural Center Research Team. Developing Research and Community Literacies to Recruit Latino
Researchers and Practitioners to Address Health Disparities. J. Ethn. Racial Health Dispar. 2016, 3, 138–144.
[CrossRef] [PubMed]
Granberry, P.; Torres, M.I. Developing Social Capital among Research Team Members to Enhance Survey
Response Rates. In Presented at the Annual Meeting of the Southern Demographic Association, Memphis,
TN, USA, 15–17 October 2015.
Dominguez, S.; Watkins, C. Creating Networks for Survival and Mobility: Social Capital among
African-American and Latin-American Low-Income Mothers. Soc. Probl. 2003, 50, 111–135. [CrossRef]
Fukui, H.; Menjívar, C. Bound by Inequality: The Social Capital of Older Asians and Latinos in Phoenix,
Arizona. Ethnography 2015, 16, 416–437. [CrossRef]
Liebow, E.; McGrady, G.; Branch, K.; Vera, M.; Klovdahl, A.; Lovely, R. Eliciting Social Network Data and
Ecological Model-Building: Focus on Choice of Name Generators and Administration of Random-Walk
Study Procedures. Soc. Netw. 1995, 17, 257–272. [CrossRef]
Small, M. Racial Differences in Networks: Do Neighborhood Conditions Matter? Soc. Sci. Q. 2007, 88,
320–343. [CrossRef]
United States Census Bureau, American FactFinder. S1903: Median Household Income in the Past 12 Months;
2009–2013 American Community Survey; U.S. Census Bureau: Washington, DC, USA, 2016.
Pearson-Merkowitz, S. The Limits of the Homogeneity Model Segregation and Civic Engagement in Latino
Communities. Am. Politics Res. 2012, 40, 701–736. [CrossRef]
Smith, S. Mobilizing Social Resources: Race, Ethnic, and Gender Differences in Social Capital and Persisting
Wage Inequalities. Sociol. Q. 2000, 41, 509–537. [CrossRef]
Muñoz-Erickson, T.; Cutts, B. Structural Dimensions of Knowledge-Action Networks for Sustainability.
Curr. Opin. Environ. Sustain. 2016, 18, 56–64. [CrossRef]
© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).