Unanticipated Gains 1-50

Unanticipated Gains
This page intentionally left blank

Unanticipated Gains
Origins of Network Inequality in
Everyday Life
Mario Luis Small
1
2009
3
Oxford University Press, Inc., publishes works that further
Oxford University's objective of excellence
in research, scholarship, and education.
Oxford New York

Auckland Cape Town Dar es Salaam Hong Kong Karachi
Kuala Lumpur Madrid Melbourne Mexico City Nairobi
New Delhi Shanghai Taipei Toronto
With offices in
Argentina Austria Brazil Chile Czech Republic France Greece
Guatemala Hungary Italy Japan Poland Portugal Singapore
South Korea Switzerland Thailand Turkey Ukraine Vietnam
Copyright # 2009 by Oxford University Press, Inc.
Published by Oxford University Press, Inc.

198 Madison Avenue, New York, New York 10016
www.oup.com
Oxford is a registered trademark of Oxford University Press.
All rights reserved. No part of this publication may be reproduced,
stored in a retrieval system, or transmitted, in any form or by any means,
electronic, mechanical, photocopying, recording, or otherwise,
without the prior permission of Oxford University Press.
Library of Congress Cataloging-in-Publication Data

Small, Mario Luis.
Unanticipated gains : origins of network inequality in everyday life / Mario Luis Small.
p. cm.
Includes bibliographical references and index.
ISBN 978-0-19-538435-2
1. Social networks. 2. Social capital (Sociology) 3. Day care centers—Case studies.
I. Title.
HM741.S585 2009
302.3082—dc22 2008041902
9 8 7 6 5 4 3 2 1
Printed in the United States of America
on acid-free paper
Preface
How well people do depends on the range and quality of their connections.
The thousands of books and articles spawned by social capital theory have
probably convinced even the toughest skeptics that better connected people
enjoy better health, faster access to information, stronger social support, and
greater ease in dealing with crises or everyday problems. It has convinced
many that to understand inequality in well-being we must understand
something about the structure of people's connections. And it has inspired
hundreds to study the formal structure of these networks, to uncover the
patterns in the systems of nodes (people) and ties (relations) that constitute
an actor’s network.
But networks do not arise out of thin air. People’s networks emerge over
the course of their routine activities, in the everyday organizations where
those activities take place. Every day, women and men drop off their
children in childcare centers, head to work in office buildings, eat lunch in
cafeterias, pick up food in grocery stores, get manicures in beauty salons, and
attend PTA meetings in schools. They kneel down to pray in churches
among other believers, play sports in gyms with other sports fans, and
discuss politics in neighborhood associations with other concerned residents.
Networks do not exist in a vacuum; they are formed and sustained in offices,
schools, churches, country clubs, barber shops, gyms, community centers,
universities, political clubs, YMCAs, childcare centers, and countless other
everyday organizations where people encounter others.
This book’s point of departure is the proposition that these everyday
organizations matter to not merely the size but also the nature, quality,
and usefulness of people’s networks. Routine organizations are not merely
places, sites where clusters of nodes and ties happen to exist. Instead, they
constitute sets of institutional rules, norms, and practices that to lesser or
greater extent affect how their members or participants interact, form
personal connections, think of one another, build trust, develop obligations,
and share information and other resources. In some organizations, patrons
routinely encounter many other people; in others, they meet few. In some,
they are subject to formal and informal rules that affect the obligations they
vi Preface
feel toward members; in others, they are not. These and many other dynam-
ics affect the nature of the ensuing social relations. That is, the organizations
shape, in varying degrees, their patrons’ social capital.
This book introduces a perspective on inequality in well-being that con-
siders how people’s social capital responds to organizational conditions.
Rather than conceiving of networks primarily as nodes and the ties between
them, it conceives them mostly as sets of context-dependent relations result-
ing from routine processes in organizational contexts. Prioritizing context
over structure, the book proposes that how much people gain from their
networks depends fundamentally on the organizations in which those net-
works are embedded. It also proposes that individuals receive distinct advan-
tages from being embedded in effective brokers—organizations that, both
intentionally and unintentionally, connect people to other people, organiza-
tions, and their resources. This book, following the tradition of studying the
unanticipated consequences of social action, documents the network advan-
tages that people may gain from doing little more than participating in the
organizations that structure their day-to-day lives. In so doing, it identifies
many of the often hidden mechanisms that sustain social inequality.
The book illustrates and develops this model through what might appear
to be an unlikely case: a study of the experiences of scores of mothers whose
children were enrolled in New York City childcare centers. These mothers
varied in race, class, education, and lifestyle; most of them worked, but they
had little else in common. The book documents that, because of the condi-
tions of their centers, many of these mothers expanded both the size of their
networks and the resources available through them. At the same time, the
book reveals that how much, if anything, mothers gained depended on
the institutional practices of their respective centers. And it shows that the
practices of the centers often resulted from larger factors such as policies of
the state, something far removed from the mothers’ everyday lives. To help
assess these mothers’ experiences, the book also analyzes a national survey
of 3,500 urban mothers and a New York City survey of nearly 300 centers. To
help assess whether the experiences of these mothers were unique, the book
compares its findings to published studies on how colleges, churches, beauty
salons, and many other organizations affect various aspects of the networks
of their members. The results make clear that the experiences of the mothers
are not unique, suggesting that differences in well-being arise, in part,
because of differences in the organizations in which people’s networks are
embedded.
Writing this book would have been impossible without the help of many
people and institutions. I received invaluable course relief from Princeton
University and the University of Chicago during the research and writing of
the book. Semester-long leaves taken at Columbia University and at New
Preface vii
York University were instrumental. I thank Irwin Garfinkel at Columbia and

Dalton Conley at New York University for making the leaves possible. Parts
of this study were funded by grants from Princeton University and from
National Institute of Child Health and Human Development (grant 3 R01
HD039135–03S1; Christina Paxson, principal investigator). Sara McLanahan
provided invaluable help early on, including the opportunity to add ques-
tions to her national survey of mothers of newborns. Christina Paxson
allowed me to join one of her grants, and to make this possible did more
than anyone could ask. Without Sally Waltman’s expert assistance, the
survey of New York City childcare centers would not have been successful.
Jean Knab and Kevin Bradway provided expertise assistance on the intrica-
cies of the Fragile Families data set. Emily Art, Martha Biondi, Kate Cagney,
Cathy Cohen, Michael Dawson, Edward Laumann, Jennifer Lee, Devah Pager,
Nicole Marwell, Omar McRoberts, Yasmina McCarty, Jamila Michener,
Sabrina Placeres, Sandra Smith, Celeste Watkins-Hayes, and Chris Winship
read several early chapters, and in some cases entire drafts. Several anony-
mous reviewers provided invaluable feedback. Several research assistants
were instrumental to the completion of the work. Laura Stark worked dili-
gently on this project when it was nothing more than a hunch. Erin Jacobs and
Rebekah Massengill were hard-working, astute, and thoughtful researchers.
My original plan, when I still entertained the fantasy that the book would be
finished in two years, was to coauthor the book with the two of them. The
changes in the book and the shifts in their interests did not make this possible.
However, I am happy that at least one coauthored paper resulted from our
collaboration. That paper was published in the September 2008 issue of Social
Forces and is reproduced (with some changes) in Chapter 7 with permission.
At Oxford University Press, James Cook was a plasure to work with, and
Stephanie Attia was a patient and constructive production editor. Finally,
and most important, I thank the many women and men who agreed to give us
their time for this project. I gained much more from their openness than
I ever anticipated.
Contents
PART I: PERSONAL TIES IN ORGANIZATIONAL SETTINGS

1 Social Capital and Organizational Embeddedness, 3
2 Childcare Centers and Mothers’ Well-Being:
Whether Mothers Did Better When Their
Children Were in Daycare, 28
PART II: SOCIAL TIES

3 Opportunities and Inducements:
Why Mothers So Often Made Friends in Centers, 51
4 Weak and Strong Ties:
Whether Mothers Made Close Friends,
Acquaintances, or Something Else, 84
5 Trust and Obligations:
Why Some Mothers’ Support Networks Were
Larger Than Their Friendship Networks, 107
PART III: ORGANIZATIONAL TIES

6 Ties to Other Organizations:
Why Mothers’ Most Useful Ties Were Not Always Social, 129
7 Organizational Ties and Neighborhood Effects:
How Mothers’ Nonsocial Ties Were Affected by Location, 157
PART IV: BEYOND CHILDCARE CENTERS

8 Extensions and Implications, 177
x Contents
APPENDICES: A MULTIMETHOD CASE STUDY

Appendix A The Process, 201
Appendix B Quantitative Data, 206
Appendix C Qualitative Data, 221
Notes, 229
References, 263
Index, 281
Part I
PERSONAL TIES IN ORGANIZATIONAL
SETTINGS
1
Social Capital and Organizational

Embeddedness
Imagine two women, Jane and Victoria, both single mothers of a young child
who earn $18,000 a year as cashiers in a department store. While Jane’s
education ended after high school, Victoria’s has continued, and she is
currently a sophomore in college. What is Victoria’s advantage over Jane?
An economist might expect Victoria to have better prospects because of
her investment in human capital, the skills and education she is accumulating
in college.1 This human capital would include not only the knowledge of
specific subject matters such as English literature or engineering, but also
the general skills that students tend to acquire through their classes, such as
how to write clearly, how to investigate a topic using libraries and the
Internet, and how to make convincing arguments. This human capital
would improve Victoria’s future prospects and also her current circum-
stances, since her research and communication skills would probably help
her when meeting immediate needs such as finding a doctor, a babysitter,
or subsidized health care.
A sociologist might expect Victoria to reap additional gains from her
investment in social capital, the resources inherent in the social networks
she is acquiring in college.2 These resources, which include the information
these networks provide and the informal and reciprocal obligations their
members may feel toward her, would motivate her to invest in meeting
others, such as students and professors. Some students would provide social
support; others would connect her to people beyond her socioeconomic
background; and her professors would likely guide and advise her, inform
her about job and educational opportunities, and recommend Victoria to
potential employers. Social capital theory would alert us to the fact that the
college is a place not merely to acquire skills but also to make connections.
The theory ignores, however, that the magnitude of Victoria’s social
capital advantage would depend substantially on the conditions of the
college. A college is not merely a place; it is a formal organization with
norms, rules, and practices that, by guiding the behavior and interactions
of its participants, inevitably shapes their networks. Victoria’s network
3
4 Personal Ties in Organizational Settings
advantage, I suggest, might be either much greater or greatly diminished,

depending on conditions of this organization.
First, these conditions would strongly affect the size and composition
of her new network. Social capital theorists such as Pierre Bourdieu and
Nan Lin, when explaining why people made connections, used the term
“investment” to emphasize that people made ties because they recognized,
or believed, that connections offer advantages. (This helps explain the term
“social capital.”) In doing so, however, they implicitly prioritized, in the
process of tie formation, the role of people’s choices over that of their
environments. While some people certainly make connections because
they know it may help them later on, others are shy, reluctant to approach
strangers, or uncomfortable with thinking of relations in instrumental terms.
The less of a social investor Victoria is, the more her new friendships will
depend on how the college structures her interactions with others—for
example, on its professors’ rules with respect to group projects, on its
deans’ support of sports teams or student clubs, and on its available cafés,
lounges, socials, festivals, and the like. In fact, Victoria may be least likely
to make friends with some of the students who, from an investment perspec-
tive, might be the most useful. Seniors, being more experienced, knowledge-
able, and locally connected, are for many purposes more useful social ties
than are underclassmen and women. But Victoria, if she resembles most
sophomores, will tend to make friends among sophomores, if for no other
reason than colleges tend to structure courses, and therefore opportunities
to meet, around cohorts—with seminars reserved for juniors and seniors
and introductory lectures intended for first-years and sophomores. Victoria’s
new social ties would be made not merely in the college, but in many
ways by it.
Second, the organizational conditions of the college would shape the
nature of these social networks, including the obligations they carry. James
Coleman and other social capital theorists emphasized that the obligations
that people feel toward each other tend to emerge informally from within
their networks, a process easy to imagine in Victoria’s situation.3 For exam-
ple, if Victoria resembles most students, then over time she will probably
develop informal relationships with some of her professors, who will in turn
feel inclined to provide some form of guidance. But not all obligations
bubble up naturally from interpersonal relations. Depending on her college,
many professors who feel no personal inclination to help Victoria will
provide some guidance anyway, because many colleges institute formal
obligations under which professors are supposed to guide students.
Under these obligations, faculty must become available to students for
meetings, advise direction in the course of study, provide insight into a
future career, write letters of recommendation, and generally provide access
Social Capital and Organizational Embeddedness 5
to the resources characteristic of their roles. If her college strongly enforced

such obligations, then Victoria’s advantage over Jane would be significantly
greater.
Third, the conditions of the college would determine whether, in addition
to ties to other people, Victoria also formed ties to other organizations.
Colleges can maintain ties to many organizations, and many types of re-
sources may travel across them. Sometimes, these connections might pro-
duce little more than access to information from other organizations. The
college might make Victoria’s job search much easier than Jane’s if it provided
a “career services” office with a small library with file folders listing contact
information for area employers and graduate programs. This targeted, ready-
collected, presorted information would be available to Victoria as needed.
Other times, the connections might be more engaged, such that the college
effectively mobilized these connections for its students. If the college coop-
erated with those management and finance firms that visit campuses each
spring to recruit graduating seniors, Victoria’s advantage over Jane in finding
a well-paying job would be immeasurably greater. Finally, some connections
to other organizations would be not so much mobilized as institutionalized,
providing access to resources useful to her day-to-day well-being. Victoria
might now be indirectly associated with a loose network of organizations that
target resources toward those who share membership in organizations such
as hers—that is, toward formally enrolled college students. By virtue of her
student identification card, Victoria could receive automatic discounts on
museum, theater, musical entertainment, and even train and airplane tickets.
In short, Victoria’s network advantage over Jane would depend strongly
on her college, which might, intentionally or not, either increase it dramati-
cally or diminish it substantially. Colleges are unique organizations, ones
that can be as close to a “total institution” as an average citizen is likely
to encounter in contemporary industrialized societies.4 But with respect
to their impact on networks, colleges are often little more than especially
effective brokers, organizations that, through multiple mechanisms, tie peo-
ple to other people, to other organizations, and to the resources of both.
This book argues that people’s social capital depends fundamentally on
the organizations in which they participate routinely, and that, through
multiple mechanisms, organizations can create and reproduce network ad-
vantages in ways their members may not expect or even have to work for.
Some organizations are more effective than others, and others are not
effective at all. But understanding people’s connections—and how much
connections generate social inequality—requires understanding the organi-
zations in which those connections are embedded. It requires conceiving of
people as organizationally embedded actors, as actors whose social and
organizational ties—and the resources both available and mobilized through
them—respond to institutional constraints, imperatives, and opportunities.

The book proposes a model to understand these processes and illustrates this
model by applying it to the case of mothers whose children are enrolled
in childcare centers. I believe the childcare center represents an ideal place
to examine these questions, because centers tend to be effective brokers
while nonetheless differing dramatically in their effectiveness, and because
they exhibit a remarkably wide array of mechanisms by which both social
and organizational ties are brokered.
The present chapter makes the case that the mechanisms by which
organizations broker social and organizational ties can be studied systemati-
cally, and that childcare centers represent an ideal site in which to identify
many of these mechanisms. The chapter proceeds in four parts. First,
it briefly reviews and critiques the social capital perspective, identifying
the questions the theory has failed to answer and explaining the conse-
quences of this neglect. Second, it identifies the three basic assumptions
on which this book’s alternative, the organizational embeddedness perspec-
tive, rests.5 Third, it briefly summarizes the implications of these assump-
tions, specifying what aspects of people’s networks are affected by
organizational conditions, how they are affected, and why. Finally, the
chapter explains why childcare centers, and the experience of mothers
within them, constitute an ideal case to examine these questions.
SOCIAL CAPITAL THEORY
The Theory
Social capital theory argues that people do better when they are connected
to others because of the goods inherent in social relationships. These goods—
the social capital—include the obligations that people who are connected
may feel toward each other, the sense of solidarity they may call upon, the
information they are willing to share, and the services they are willing to
perform. People who are socially connected therefore have recourse to a
stock of “capital” they can employ when needed. The term “social capital”
was first used in this sense by economist Glenn Loury. However, the intel-
lectual roots of the theory lie more firmly in the works of Bourdieu and
Coleman. In recent years, Lin has worked harder than anyone to develop a
formal theory of social capital. These three authors, Bourdieu, Coleman,
and, Lin, define social capital in somewhat different ways, but they all
conceive it as the resources that inhere in social relations.6
Bourdieu defined social capital as “the aggregate of the actual or potential
resources which are linked to possession of a durable network of more or less
institutionalized relationships of mutual acquaintance and recognition”—

that is, as the resources one derives from belonging to a network.7 Bourdieu
believed that people possessed varying amounts of different kinds of “capi-
tal,” such as cultural capital (knowledge of high art) and economic capital
(possession of financial resources). As a result, he argued that social capital
included any type of resource available through one’s social ties: the “volume
of the social capital possessed by a given agent thus depends on the
size of the network of connections he can effectively mobilize and on the
volume of the capital (economic, cultural, or symbolic) possessed in his own
right by each of those to whom he is connected.”8
Coleman defined social capital as the obligations, norms, and informa-
tion available to a person from her or his network. To understand Coleman’s
definition, it helps to know that his intellectual mission was to develop a
model of social behavior that was both sociologically compelling and rooted
in the idea that actors are rational.9 Thus, he explained, “If we begin with a
theory of rational actor, in which each actor has control over certain
resources and interests in certain resources and events, then social capital
constitutes a particular kind of resource available to an actor.”10 The first of
these resources was the set of obligations a relation might feel: “If A does
something for B and trusts B to reciprocate in the future, this establishes an
expectation in A and an obligation on the part of B.”11 This obligation
becomes a resource that actor A can employ in the future. A second resource
was the presence of norms that encourage people to help each other:
“A prescriptive norm within a collectivity that constitutes an especially
important form of social capital is the norm that one should forgo self-interest
and act in the interests of the collectivity.”12 If a group has a norm that people
should forgo their self-interest, then an actor within it can reliably turn to
others in the group for help when needed. A third resource was information,
the knowledge that people to whom an actor is connected possess.
Lin defined social capital as the “resources embedded in a social structure
that are accessed and/or mobilized in purposive actions.”13 Lin’s objective
was both to synthesize the work of earlier theorists and to fit the theory
formally into social network analysis, something Bourdieu never did and
Coleman only began to do. In fulfilling this objective, Lin assumed that
actions are not only purposive but also rational. Lin, following Coleman
and Bourdieu, argued that four types of resources constitute social capital:
information, the influence that networks have over people, the social cre-
dentials that networks can impart, and the personal reinforcements, essen-
tial for mental health, that networks provide actors. Lin has contributed to a
large body of work showing that people who have access to these resources
have better mental health and reach higher positions in the occupational
ladder.14
In sum, while the authors differed in the specific resources they included
under the rubric of social capital, they all agreed that social capital referred
to resources people derived directly from their social ties.
The Missing Question: How Do People Make Social Ties?

These researchers have spawned a massive literature on the benefits of social
capital.15 But the researchers devoted little time to an important question:
how do people make the social ties that provide access to these resources?
Coleman sidestepped this question; Bourdieu devoted but a few paragraphs
to it; and Lin, who wrote perhaps more, ultimately did not answer it.16
Lin did not explain how people made social ties because his model,
in which networks result from investments, made the question irrelevant.
Lin believed that people act rationally, motivated by both instrumental and
expressive needs. For either reason, actors invest in social relations with an
eye to the returns: “[The] theory . . . suggests that actors . . . are motivated by
instrumental or expressive needs to engage other actors in order to access
these other actors’ resources for the purpose of gaining better outcomes.”17
That is, the model explicitly proposed that people make connections because
of the gains they anticipate. However, the model did not show this to be
the case—the proposition followed naturally from Lin’s assumptions that
actors tend to behave purposely and that social capital is an effective
investment. From this perspective, asking why people make ties inspires as
much curiosity as asking why they make money; people do it because they
know it is good for them, which is to say that rational actors do what is
rational. And since the why question was answered in rational terms, the how
question became trivial, resulting in a programmatic neglect of the processes
by which people form useful ties.
Bourdieu also wrote about how people form ties and stated that networks
result from investments. However, what he meant by “investments” was
ambiguous, so why he did not say more about how people form ties remains
unclear. In his most extensive comments on social capital, Bourdieu argued
that the “existence of a network of connections is not a natural given, or
even a social given.”18 Instead, it results from people’s deliberate efforts.
“In other words,” he explained, “the network of relationships is the product
of investment strategies, individual or collective, consciously or unconsciously
aimed at establishing or reproducing social relationships that are directly
usable in the short or long term.”19 By one reading, Bourdieu is arguing that
people strategically enter into social relationships because these will be
useful in the future. This is not only consistent with Lin’s argument; it
is Lin’s reading of Bourdieu. In fact, Lin argues that an “investment” per-
spective is the common thread across all major social capital theorists.20 By a
different reading, Bourdieu is arguing that, regardless of how they are

formed, new ties effectively end up being an investment. This second
reading, which is considerably more generous to the author, can be justified
by noting that in other works Bourdieu has argued that many actions tend
to be habitual, rather than purposive.21 Nevertheless, I do not believe
Bourdieu resolved this tension.22 So, whether his model made the question
irrelevant or he simply never came to address it, Bourdieu did not articulate
the mechanisms by which people make connections. At most, he explained
that it requires “unceasing effort.”23
The assumption that social ties result from investments was abandoned
by many later social capital researchers, especially those whose research
did not involve formal mathematical modeling. Today, authors differ in
how much emphasis they place on this assumption, as Charles Kadushin
noted when evaluating the idea of investment in social capital theory.24
Nonetheless, the theoretical lacunae it generated persist in most of the
later works: in their devotion to studying the consequences of social ties,
many researchers have taken for granted the processes from which ties arise.
As a result, recent reviews of research on social capital have found little
to report about how actors form ties (despite the fact that recent research by
formal network analysts and experimental psychologists provides tools with
which to answer this question).25
Why the Missing Question Matters

Not asking how people form ties creates important problems. First, it
leaves unanswered a critical question in the study of network in-
equality: what mechanisms account for it? A theory of network inequality
cannot be content with demonstrating that social ties are useful: if some do
better because they have more ties than others, then it certainly seems
important to understand why they have more ties. For example, recent
studies have tried to determine, using advanced statistical techniques, wheth-
er having connections helps people find jobs.26 Researchers increasingly
acknowledge, to their credit, that to properly answer this question
they must take into account unobserved differences among people
that determine who is well connected in the first place. As a result, research-
ers have examined ways of statistically controlling for these differences.
This solution, however, addresses only half the problem: how people
make ties is not merely a statistical nuisance to “control away”; it is a
substantive process to understand. Knowledge comes about not only
from demonstrating associations but, more important, as Peter Hedstrom
and Richard Swedberg maintain, from explaining the mechanisms that give
rise to them.27
Consider that, without knowing how network inequality arises, developing

the means to reduce it becomes impossible. Some believe that social capital
theory “blames the victims” for their problems, since it could lead one to the
conclusion that if only people bothered to network more effectively, then
social inequality might be lowered. More important, critics maintain that
the theory’s focus on personal networks comes at the expense of the study of
larger structural forces. In fact, the theory can be faulted for leaving practi-
tioners at a loss, since one remains unclear about how to decrease differences
in the number of useful connections to which people have access.28
Second, while this book’s concern is the social capital of individuals,
this question also informs our knowledge of the social capital of neighbor-
hoods and of nations. At the neighborhood level, Robert Sampson and his
colleagues have argued that collective efficacy, the willingness of neighbors
to get involved in local problems on behalf of the common good, reduces
neighborhood delinquency.29 When neighbors are willing to call the police
if they see something suspicious or scold young people they witness vandal-
izing street corners, they make it difficult for people to perpetrate street
crimes. Sampson and colleagues have found that collective efficacy rises
when neighbors tend to know one another. How do neighbors come to
know one another? Why do people know more of their neighbors in some
areas than others? Unpacking this process is critical to understand what gives
rise to collective efficacy. A similar concern motivates this question at the
national level. Political scientist Robert Putnam has argued that over the
last several decades the United States witnessed a decline in civic participa-
tion and collective social capital, as people spend less time in the company
others, even when conducting recreational activities such as bowling.
To know why people are more likely to bowl alone, it is important to
know how people normally make friends with whom to bowl, and why
some have more such friends than others.30
Third, as I discuss in this book, how a person forms and sustains a tie can
affect the social capital to which she has access. That is, many of the
obligations people feel and the resources they feel willing to provide others
derive from the contexts that gave rise to and sustain their relationships.
Of these contexts, none is more important than the organizational context.31
ASSUMPTIONS UNDERLYING THIS BOOK
This book proposes an alternative model that first asks how people make
connections and argues that a major part of the answer lies in those organi-
zations in which people participate routinely.32 The book examines how
the resulting social capital is affected by these organizational contexts. And
it argues that participating in organizations that effectively broker social

capital improves people’s well-being. The model rests on three assumptions,
each of which I discuss with some care: first, that actors may form ties either
purposely or nonpurposely; second, that forming ties either purposely or
nonpurposely depends on the context of social interaction; and third, that
the context of interaction can be shaped significantly by organizations.33
Actors May Form Ties Either Purposely or Nonpurposely

Earlier, I described the proposition by social capital theorists that actors
invest in networks with an eye to their gains. That proposition suggests that
people make ties as a result of purposive actions geared toward the benefits
of acquiring social capital. However, people may form new ties through at
least three other types of action. To understand what these are, we must
distinguish purposive from nonpurposive acts, and global from local action.
Theorists such as Robert Merton refer to an act as purposive if the
perpetrator was motivated by an objective, such as requesting an application
(act) in order to apply for a job (purpose). Often, the objective motivating an
act is rational; in fact, rational choice theory assumes that actors tend to act
in pursuit of rational objectives.34 However, the objective may be rational or
irrational, self-interested or altruistic, individual or collective, psychological
or cultural.35 Consider, for example, that people can deliberately act in ways
that harm them. All that is required is some purpose. By contrast, a nonpur-
posive act has no conscious objective. An example is the act of laughing.36
An act may be defined either globally or locally, depending on how
narrowly we draw the boundaries around it. A global purposive act, for
example, is to attend college with the objective of obtaining a diploma;
a very local one is to sign the registration card containing a semester’s list
of courses. Global acts constitute bundles of local acts; consequently, they
may result from multiple or even contradictory motivations, even when
they are guided by one overarching purpose. The overarching purpose
of college attendance may be to obtain a diploma, but other, subsidiary
purposes may include escaping from home, feeling good about oneself,
yearning for new experiences, or finding someone to marry.
Depending on their purpose or lack thereof, people may form ties as
a result of at least four types of action. First, an actor may form a tie when
the purpose of her action is to make a tie. That is, an individual may
introduce herself to another to either have more connections or gain access
to his or her resources. This idea, the core of the investment perspective,
reflects the common usage of “network” as a verb.37
Second, an actor may form a tie when the purpose of her action is to
accomplish some other objective. That is, making a tie may be a by-product
of the pursuit of another aim. Consider a local example: in a grocery store a

man waiting in a long checkout line asks the one standing before him for
the time. The other responds and starts a conversation that finishes with
an exchange of phone numbers and an invitation to have coffee. In this
case, the time-seeker formed a tie even though “networking” was not his
purpose; the relationship was triggered by an act whose (local) purpose was
to learn the time.38 Consider a global example, which might illustrate more
common circumstances: a churchgoer volunteers at a summer block party
that serves as a fund-raiser, and consequently meets several of her neigh-
bors. Her (global) purpose was to raise funds for her church; meeting her
neighbors was a by-product.
Third, an actor may form a tie when her action had no purpose other than
itself. Sociologists have defined an act as expressive when it is perpetrated
for its intrinsic value, when conducted for its own sake rather than in pursuit
of an objective. Many of the actions that we consider emotional, such
as laughing, sighing, or crying—when they have no ulterior purpose—are
expressive. And they can lead to new ties in informal social situations.
Consider an example: a teenager awaiting his turn at a barbershop overhears
a barber crack a joke, to the client whose hair he is trimming, about the
presidential election; the teenager laughs, prompting a reply from the barber
that leads to a conversation and an informal relationship. The teenager’s
laughter had no purpose; it was strictly an expression that triggered a
conversation and relationship.39 While this teenager’s act was locally expres-
sive, globally expressive acts also often form and sustain social relations,
according to sociologist Georg Simmel. Simmel believed that much of what
makes us human is our practice of “sociability,” the state of associating
with others for the sake of association, a state in which “talk is a legitimate
end in itself.”40 Just as we cry for the sake of crying, Simmel would argue,
we often talk for the sake of talking, not in pursuit of some other objective
or even for the sake of making ties.
Fourth, an actor may form a tie when her action had neither a true
purpose nor even itself (expressively) as a purpose. In these actions, purpose
plays no role whatsoever, because they result from preexisting dispositions.
Such an act may be defined as habitual. While Bourdieu’s writings on social
capital relied on a model of purposive action, his books on cultural capital
examined what he called the “habitus,” the set of cultural dispositions to
act, accumulated over an actor’s lifetime, that embody both a person’s own
history and that of the group, class, or society of which she formed part.
In this model, people act as a result of their customary predispositions.41
These often operate locally. Consider an example: a woman at a bus stop
sees a man bring his forearm to his face and sneeze. She instinctively blurts,
“Bless you”; he thanks her, and a conversation ensues. Her act did not,
properly speaking, have a purpose (even if later, when prompted, she might
explain that her blessing was the polite reaction). Nor was it expressive,
the way laughing or crying are intended to express a feeling. Instead, her
“Bless you” was blurted out of habit(us). Predispositions also (and perhaps
more commonly) operate globally, since global actions that result from
habitual tendencies may either place or not place people in situations
where meeting others is likely. For example, in the crowded grocery store,
suppose that two shoppers see an opportunity to cut the long checkout lines,
as one cashier is about to open a register. One cuts; the other does not even
think about it—cutting was not his predisposition. Only the second man
is likely to meet other shoppers, as they commiserate over the perils of
rush-hour shopping.42 When acts are habitual, expressive, and perpetrated
for other purposes, forming ties is often unexpected, resulting from social
interactions in the presence of strangers. In these three types of action,
forming ties, by definition, is a by-product.
In sum, people can make ties when it was their purpose, when they had
a purpose other than making ties, when their purpose was nothing but
the act itself, and when they had no purpose at all at the time of social
interaction. These four circumstances must form part of any model of tie
formation that considers the motivations of the actor.43
Forming Social Ties Either Purposely or Nonpurposely Depends

on the Context of Social Interaction
The fact that people may form ties when doing so was not necessarily
their intention makes clear that understanding how their actions lead to
new connections requires knowing something about the context of social
interaction. And the less purposely actors pursue social connections, the
more we must know about their context. Specifically, we must know wheth-
er, how, and under what conditions they interact with others.44
First, forming social ties depends on whether actors interact at all—that is,
on the availability of opportunities to interact with strangers. Among the
first researchers to make this point explicitly were Paul Lazarsfeld and
Robert Merton, who argued that the root of friendship formation lay in
the opportunities people had to interact.45 Since then, this idea has been
proposed and demonstrated many times, perhaps most systematically by
Peter Blau and Joseph Schwartz, who relied on it to ground their network
theory of contemporary society.46 For example, sociologist Ray Oldenburg
argued that people will find it difficult to meet more than just a few of their
neighbors in the absence of informal gathering places such as cafés and
neighborhood bars. And Maureen Hallinan has found that elementary
and junior high school students tend to have more friends if they are
enrolled in larger classes. In fact, opportunities to interact may be critical

even for purposive social investors. For example, William J. Wilson argued
that people in poor neighborhoods looking for jobs will find it difficult to
develop middle-class networks because they lack opportunities to interact
with them informally, an argument consistent with the evidence.47
Second, making ties depends on how actors interact with others: how
long they interact, how frequently, how intensely, and while performing
what activity. Two of these modes of interaction have been shown to be
especially important: how frequently and while performing what activity.
The consequences of frequent interaction have been the foundation of
several lines of research, especially those inspired by Richard Emerson’s
exchange theory and by George Homans’s theory of the group. Emerson
believed that repeated exchange between people reduced their mutual
uncertainty, while Homans believed that repeated interaction between two
parties heightened their mutual affection.48 In support of these ideas, social
psychologists have shown in controlled experiments that the more frequent-
ly two people interact, the closer they become and the more they trust each
other.49 I suggest that when strangers, for whatever reason, encounter each
other repeatedly, they become increasingly likely to develop a friendship.
For example, if two neighborhood residents repeatedly see each other at
a local diner, with each successive encounter they are more and more likely
to become friends, as sociologist Mitchell Duneier chronicles in his study
of men’s friendships in a Chicago restaurant.50
Equally important is the activity being performed. People may form
ties when engaged in many activities, as the earlier examples of the shoppers
and the passengers at the bus stop illustrate; but not all activities produce
new ties in equal measure. Sociologist Scott Feld has defended the signifi-
cance of “focus,” which he defined as “any social, psychological, or physical
entity around which joint activities of individuals are organized.”51 A “focus
of activity” may result from a common concern, such as when two mothers
in a childcare center start a conversation about their children’s progress, or
two black men waiting at a barbershop begin to debate the best means
to avoid razor bumps. The common focus—childcare or grooming—provided
each pair a topic around which to start a conversation, a natural entry point
to the relationship. Relationships often arise around a common object of
attention.52
Third, forming social ties also depends on the conditions under which
people interact. I refer specifically to the degree of competitiveness and the
degree of cooperation. When interaction is competitive, the parties are
struggling over a particular good or resource, such that one will acquire it,
or acquire most of it, or acquire it first. Competition makes opponents out
of actors, undermining trust and the formation of friendships. As the stakes
increase, the chances of tie formation decrease quickly. Consider two job
applicants waiting at a small lobby for their respective job interviews. While
they might strike up a friendship as they wait to be summoned, the proba-
bility that they will do so becomes lower if each of them knows why the other
waits; even lower if the firm will offer only one position; and lower still if, in
the current job market, no other openings have emerged in many months.53
The dynamics of cooperation differ substantially. In cooperative interac-
tion, the parties work jointly to accomplish a collective goal. Cooperating
with strangers, I suggest, tends to produce friendships. In experimental
settings, social psychologist Edward Lawler has shown that when people
successfully accomplish joint tasks, the cohesion of the group increases.54
For example, when two new mothers at a childcare center are asked to
collaborate on a fund-raiser, their need to coordinate and find a way to
work together should increase the chances that they become friends.
In sum, independent of their own intentions, people are more likely to
form ties when they have opportunities to interact, when they do so fre-
quently, when they are focused on some activity, when they are not compet-
itors, and when they have reason to cooperate.
The Context of Interaction Is Shaped Significantly by Organizations

Finally, I assume that these elements of interaction—whether, how, and
under what conditions people interact—react to the organizations in which
people participate.55 This book defines an organization as a loosely coupled
set of people and institutional practices, organized around a global purpose,
and connected, both formally and informally, to other organizations.56
It considers routine organizations in which people, whether patrons or
employees, have opportunities to interact with others, such as childcare
centers, barbershops, diners, Internet cafés, colleges, firms, synagogues,
YWCAs, bowling alleys, and recreation centers. Three issues in this defini-
tion demand attention.
First, an organization refers to both the actors who compose it and the
institutional practices that organize their behavior. A childcare center is
composed of teachers, directors, parents, children, and janitors and institu-
tional practices such as teaching, pickup and drop-off, play time, after-lunch
napping, and PTA meetings. An organization’s actors influence tie formation
to the extent that they determine how people interact, such as when a center
director asks parents to introduce themselves to one another at a PTA
meeting.57 An organization’s institutional practices also influence tie forma-
tion to the extent that they shape social interaction. Practices, however,
may be “institutional” in two different senses, one normative and one
cognitive.
An institution in the normative sense may be defined, following Victor

Nee and Paul Ingram, as “a web of interrelated norms—formal and
informal—governing social relationships.”58 That is, an organization may
enforce rules and norms that its actors feel compelled to follow, such as a
requirement that all new college students participate in first-year orientation
or the norm that congregation members should fund-raise for the church.
Since organizations can enforce compliance under the threat of lost mem-
bership, they can motivate participants effectively. An institution in the
cognitive sense may be defined, following John Meyer and Brian Rowan,
as “classifications built into society as reciprocated typifications and inter-
pretations.”59 These institutions are not mandates but categories, generally
taken for granted, through which actors interpret their world and social
interactions. Whereas normative institutions tell actors how they ought to
behave, cognitive institutions shape their perception of their circumstances.
Such cognitive understandings shape, for example, whether coworkers per-
ceive one another primarily as competitors or, instead, as members of the
same team. For example, while some clothing retailers do not pay their
workers on commission, others do, implicitly encouraging competition
even though there is no “rule” or “norm” mandating competitiveness. Either
normatively or cognitively, organizations may institutionally encourage so-
cial interaction that is limited or frequent, competitive or cooperative.60
Second, while organizations may have a global purpose, the people who
compose them may be motivated by separate or additional objectives and
beliefs, resulting in a collection of actors with multiple purposes. In this
sense, the various actors and activities of an organization constitute a loosely
coupled entity.61 Consequently, understanding what motivates people in an
organization is more complex than simply knowing its purported objective
or function. A childcare center may be more than a place for childcare;
a mosque, more than a site for prayer. For example, while a beauty salon
might pursue the global objective of earning profits by setting hair and
decorating fingernails, a given beautician might be motivated, on a daily
basis, by the wish to spend time in the company of others or to sustain an
informal support group among neighborhood women.62
Third, organizations may be tied to other organizations through multiple
arrangements that vary in many elements, such as their complexity, stabili-
ty, authority, and formality.63 I assume that these organizational ties will
affect the attitudes and behavior of actors within the organization. Particu-
larly, external organizations that possess strong authority may motivate
people within an organization to institute practices that build or sustain
social ties among its members. For example, a philanthropic agency may
donate a large sum to a community college provided that part of it is used
to build a gathering place for students.64 In this respect, the community
college was externally motivated to establish a place likely to build social

capital.
In sum, the context under which people interact can be shaped by both
the actors and the institutional practices that constitute an organization,
which may be motivated by internal and external factors.
OTHER ASPECTS OF PERSONAL TIES
The preceding assumptions make clear why organizations may affect wheth-
er and how people form ties. But they also provide a way to understand how
and why organizations may affect other aspects of social capital. First, people
may not just form but also sustain, strengthen, or weaken their relations
to others either purposely or nonpurposely. For example, globally habitual
actions that repeatedly place people in the company of others may sustain
their relations even when the actors fail to make “unceasing efforts” to do so.
People sustain friendships at gyms with others whom they only see at gyms,
because they patronize gyms repeatedly. Second, the context of social
interaction—whether it is frequent or infrequent, focused or unfocused,
competitive or noncompetitive, and cooperative or noncooperative—is likely
to affect the quality of the ensuing relations: how strong or weak ties are,
how much the parties trust each other, and what resources, services, infor-
mation, or support the parties are willing to provide. For example, people
may more willingly trust and help others when their interaction takes place
in noncompetitive contexts. Third, both the actors and institutions that
compose an organization may not only regulate activity but also impose
obligations, enforce pro-social norms, and encourage organizational mem-
bers to share resources. For example, if a church encourages cooperative
interaction, it helps build pro-social norms (a form of social capital), as
researchers have documented in recent years.65
Fourth, what is true about organizations’ motivations to form social ties is
likely true about their motivation to form organizational ties. That is, just as
a center director might be motivated to form connections among mothers,
she might be motivated to form connections between mothers and external
organizations. Such ties rarely form part of social capital analyses, but they
may constitute, I suggest, a major source of goods and information.
In short, I argue that (1) organizational contexts affect most aspects of
social capital, including whether a person makes ties, what kind of ties she
makes, whether the goods in those ties are available to the person, and how
those goods are acquired; (2) organizations may affect social capital either
purposely or nonpurposely, and through the influence of either actors or
institutional practices; and (3) organizations, or their members, may be
motivated to affect social capital by either internal or external pressures. The

three arguments represent answers to three questions on the impact of
organizational embeddedness on people’s ties: what is affected, how, and
why. I briefly summarize these implications, before turning to the case study,
the childcare center.
What Organizational Contexts Affect
Social and Organizational Ties

By shaping their participants’ interactions and activities, organizations can
shape the extent to which they form either social or organizational ties.
Organizational ties, such as Victoria’s ties to external libraries and employment
firms, have formed little part of social capital theory, but they represent an
important advantage potentially available through an organization. In addi-
tion, organizations can shape the nature of the ensuing relationships, including
their strength or weakness, and the resources available through them.66
Resources, Access, and Mobilization

We have seen that social capital can refer to different kinds of resources
embedded in social networks. Bourdieu, Coleman, Lin, and others included
the following: information, services, material goods, trust, obligations, and pro-
social norms. Organizational contexts may affect not only ties but also these
resources—specifically, whether a person has access to them and whether she
makes use of them, which scholars have called, respectively, access to and
mobilization of social capital.67 Organizations affect the former by influencing
whether people form ties to the people or organizations that possess the goods;
they contribute to the latter to the extent that they enact, enforce, or encourage
trust, pro-social norms, supportive services, information sharing, the provision
of services, and the distribution of material goods. One must not neglect the
importance of people’s agency: some actors mobilize connections more effec-
tively than others, and some will take greater advantage of the network
resources available in a given organization. However, I suggest that mobiliza-
tion does not depend solely on how willing a person is to use her ties; mobiliza-
tion is mediated, and sometimes perpetrated, by organizations.68
Organizational context may also affect what scholars have called negative
social capital.69 For example, scholars have noted that group solidarity may
restrict individual freedoms or place stringent demands on members, such
as when religious institutions forbid their members from marrying outsiders.
Through mechanisms such as the threat of lost membership, organizations
may enforce norms or practices that result in such consequences.
How They Do So
Brokerage
To capture the many ways organizations can shape social ties, an effective
umbrella term is brokerage. The notion of brokerage enjoys a distinguished
lineage in sociology; a broker is generally defined as the actor who brings
two previously unconnected actors together. In recent years, Ronald Burt
has demonstrated convincingly the importance of brokers in organizational
contexts.70 In this book, brokerage is the general process by which an organi-
zation connects an individual to another individual, to another organization, or
to the resources they contain. Since here the broker is an organization, and
since brokerage involves connecting people both to other people and
organizations and to their resources, the process of brokerage may be
significantly more complex than in person-to-person situations. Yet the
term succinctly captures the heterogeneous set of practices by which
organizations function as connectors. Organizations broker connections in
many different ways; however, these may be categorized as either actor
driven or institution driven and either purposive or nonpurposive.71
Actor-Driven and Institution-Driven Brokerage

Actor-driven brokerage is the process by which a person in the organization
connects people to other people, to other organizations, or to the resources of
either; institution-driven brokerage is the process by which an institution, in the
normative or cognitive sense, brokers any of these connections. For example,
when a college professor (actor) teaching a seminar requests that students
introduce themselves to one another, she is connecting persons to other per-
sons; when a librarian (actor) obtains for a student a difficult-to-find book from
another library, he is connecting a person to the resources of another organiza-
tion. These constitute forms of actor-driven brokerage. But when an STD clinic
automatically refers a client testing positive to a social support agency, it is
institutionally connecting a person to another organization.
Purposive and Nonpurposive Brokerage

The act of brokerage may be purposive or nonpurposive, depending
on whether the broker intended to connect people or to attain some other
objective. The professor who asked students to introduce themselves cer-
tainly acted purposely. Other instances are neighborhood association meet-
ings where participants, sitting in a circle, are asked to introduce themselves
to the group, or office parties where employees are asked to affix name tags
to their clothing. While such strategies vary in their effectiveness, they

do not differ in their purpose: to form connections among participants.
I emphasize that personal motivations may operate independently of global
organizational purposes and that brokerage may involve other types of
connections. While a Dominican botanical drugstore might pursue the global
objective of earning profits by dispensing ethnically specific medication,
a given pharmacist might be motivated, on a daily basis, to help new
immigrants transition to American society by tying them to resource agen-
cies—that is, to connect people to the resources of another organization.72
Much of this book explores the interesting and important nature of
nonpurposive brokerage, whereby an organization had no intention of form-
ing ties among participants or patrons. In the sociology of complex organiza-
tions, several researchers have studied related ideas, including Charles
Perrow, whose concept of “normal accidents” suggests that organizational
systems may unwittingly be designed such that accidents are inevitable, and
Diane Vaughan, whose idea of “normalization of deviance” captures how
small, daily acts over time result in mistakes an external observer might find
to be rationally avoidable.73 Both authors believe that institutionalized
practices produce unexpected consequences. In the present context, non-
purposive brokerage may occur in multiple ways. For example, many neigh-
borhood laundries, to ensure that washers and driers are continuously
available, do not permit their patrons to leave clothing unattended, forcing
them to spend hours sitting and waiting in the company of others. The
laundry’s purpose was efficiency; an unintended consequence, socialization
among neighbors. In a study of dormitories for returning World War II
veterans at MIT, researchers found that simple architectural design deci-
sions, such as how many apartments to place on one floor or how far apart
to situate the staircases, affected friendship formation, with the strongest
ties forming among those living close enough that unplanned encounters
were frequent.74 Many circumstances or practices common across organiza-
tions, such as waiting lobbies or collective tasks, create similar possibilities of
unintentionally brokered connections.75
In sum, organizations may broker ties through mechanisms that may be
either actor or institution driven, and either purposive or nonpurposive.
Why They Do So
The discussion to this point should make clear that organizations may broker
connections for multiple reasons, depending on who does the brokering
and whether it is purposive. This discussion nonetheless benefits from
distinguishing motivations internal to the organizations from those external
to it. Internal motivations include the personal goals of the members, as in
the case of the Dominican pharmacist, and the institutional imperatives

necessary for organizational survival, such as a struggling church’s thirst
for cash from fund-raisers. External motivations may derive from funders,
professional associations, and the state, which may encourage or mandate
practices that in the end contribute to local social capital. For example,
a philanthropic organization may grant several hundred thousand to a not-
for-profit childcare center, provided it institutes a parent association or
collaborates with a neighborhood agency aimed at helping women with
domestic abuse.
One important implication is that few external entities are able to
exert stronger influence than the state, which can impose regulations
with the threat of fines, penalties, decertification, or incarceration. As a
result, it is possible that state pressures may be so powerful that they
increase social capital in measurable ways. This book suggests that such
pressures do, in fact, produce some forms of social capital under some
circumstances.
EFFECTIVE BROKERS
We have seen that when people participate in organizations, they encounter

a set of actors and institutions that, through varying mechanisms, may alter
their social capital in ways that could be beneficial to their well-being. The
sheer number of factors at play would seem to suggest a rather haphazard
process, one in which the accumulation of social capital remains an unwieldy
amalgam of practices, institutions, and motivations impossible to study
systematically. On the contrary, I submit that organizations exhibit regular
patterns that would lead us to expect some of them to be systematically
effective brokers—so effective that, through their effects on social capital,
they can measurably improve the well-being of their participants.
What such patterns are depends on the types of ties being brokered,
whether social or organizational. With respect to social ties, effective brokers
are likely to exhibit, among other traits, (a) many opportunities for
(b) regular and (c) long-lasting interaction, (d) minimally competitive and
(e) maximally cooperative institutional environments, and both (f) internal
and (g) external motivations to maintain those opportunities and sustain
those environments—particularly such practices that would likely contribute
to organizational survival. With respect to organizational ties, effective
brokers will likely demonstrate (a) resource rich and (b) diverse organiza-
tional networks in which (c) transferring resources fulfills the objectives of
multiple constituencies. Among effective brokers, these newly acquired ties
and the resources gained therein would observably, even if unexpectedly,
benefit participants. One organization that, on average, is a remarkably

effective broker is the childcare center.
CASE STUDY: MOTHERS AND CHILDCARE CENTERS
Childcare centers are often thought of as the impersonal alternative to

family care, as the place where formal rules overwhelm informal relations.
Yet for a study of how people make and use the connections that matter to
their well-being, few cases are more appropriate than that of mothers of
children in daycare centers. Childcare centers are strategic study sites, ones
that, because of their uniqueness, allow us to observe processes that would
be difficult to examine in other settings.76 They are ideal for theoretical,
substantive, and methodological reasons.
Why Centers?
First, childcare centers, as described throughout this book, tend to be re-
markably effective brokers for the mothers whose children they service.
They broker both social and organizational ties, and their brokerage is
associated with greater material and mental well-being, bringing to light
the concrete implications of organizationally embedded networks.
Second, childcare centers provide an important methodological advan-
tage, the analytical leverage to unravel why some organizations are more
effective brokers than others. Even though the average center is a rather
effective broker, the variance is high, and many centers are terribly ineffec-
tive at connecting mothers either socially or organizationally. This wide
range is due in part to the diverse array of organizational forms that centers
can take: childcare centers may be for-profit or nonprofit, privately or
publicly funded, corporate or family run, and secular or religious, yielding
a rich variety of configurations and institutional practices. This organization-
al heterogeneity is matched by few other organizations in which one might
conduct such a study. Churches, probably the most commonly studied
organization among urban sociologists, are always religious nonprofit enti-
ties; firms, the most commonly studied entity among organizational sociol-
ogists, are always profit-oriented businesses. Childcare centers allow us
the analytical leverage to observe the role of multiple sets of pressures
while holding the type of organization constant. A different type of meth-
odological advantage is that centers provide a unique opportunity to study
tie formation as it happens. Since a new cohort of families enrolls in child-
care centers every year, it is possible to study the formation of new networks
just as it is happening, something difficult to accomplish in other settings.
People in churches, for example, have often known each other for years,
making it difficult to examine how participants think about people they are
just beginning to meet.
Third, mothers of young children remain one of the most important
populations in the study of well-being and networks, as evidenced by classic
studies such as Elizabeth Bott’s Family and Social Network, Carol Stack’s
All Our Kin, and Kathryn Edin and Laura Lein’s Making Ends Meet.77
Mothers of young children have more to gain from supportive or resourceful
ties than do people at almost any other point in the life cycle, since having a
new child undercuts parents’ free time, increases their household costs,
and introduces the many unpredictable events, crises, illnesses, and accidents
to which young children are prone. While fathers are increasingly affected, it
still remains the case that mothers bear most of the burden of caring for
young children. At the same time, most mothers of young children today
either work or actively are looking for work, as I document in chapter 2.
Consequently, at the moment when new friends might be most useful,
mothers have the least time to make them.
In fact, the childcare center has become an increasingly important orga-
nization because births to unmarried mothers have reached historic highs, at
36% in 2004 for all unmarried mothers. Among Latinas the rate is 46%;
among African-Americans, 69%.78 It has become especially important to
low-income mothers since the Personal Responsibility and Work Opportuni-
ty Reconciliation Act of 1996, which eliminated the old welfare system,
instituted a work requirement, forcing mothers to find childcare arrange-
ments. Among organizations relevant to unmarried mothers, the childcare
center is one of the most crucial.
Data Sources
This project is based primarily on four data sources (see table 1.1). This
multilevel, multimethod project employed both qualitative and quantitative
data sources collected from both individuals and organizations. The
integrated nature of the design provides the means to untangle the mecha-
Table 1.1. Principal data sources for the study
Quantitative Qualitative
Individual Fragile Families and child well- In-depth interviews (67

level being study (n ~ 3,500 mothers) mothers and some fathers)
Organizational Childcare Centers and Families Center case studies (23
level Survey (n ~ 300 centers) centers)
nisms through which centers broker mothers’ social and organizational ties
while identifying how prevalent and how distributed at least some of these
mechanisms are. I describe each data source briefly below. For details, see
appendices B and C.
The quantitative, individual-level data set is a national panel survey of
approximately 3,500 mothers of children born between 1998 and 2000 in 20
large U.S. cities. The survey, named the Fragile Families and Child Wellbeing
Study, or Fragile Families, is representative of all mothers of newborns in
U.S. cities larger than 200,000. Mothers were interviewed at the time of their
child’s birth and when the child was one, three, and five years old; the study
described in this book employs the last two waves of the survey. Mothers
were asked questions regarding their demographic characteristics, their use
of formal childcare centers or alternatives, and their friendships in and out
of childcare centers. The Fragile Families survey allows us to compare basic
network characteristics of mothers who use childcare centers and those who
do not and to test a few of the ideas developed from the qualitative studies
about the factors affecting tie formation.79
The quantitative, organization-level data set is my Childcare Centers and
Families Survey, an original study of approximately 300 centers randomly
sampled in New York City in 2004.80 The survey, which I commissioned after
conducting preliminary fieldwork, contains data on the organizational struc-
ture of childcare centers, basic institutional practices, opportunities avail-
able for parents to network, services provided other than childcare, referrals
to other organizations, ties to other organizations providing resources, the
characteristics of those organizations, and other organizational variables.
The survey, which I will refer to as the Childcare Centers survey, allows
me to paint a basic picture of the average characteristics of childcare centers
in a major city, to observe the level of heterogeneity, and to test several of
the hypotheses developed from the qualitative studies.
The qualitative data consist of case studies. For the qualitative, organization-
level data, two research assistants and I interviewed the directors or other
supervising personnel of 23 childcare centers in New York City. The centers
were selected to exhibit a range of income and racial characteristics, particularly
among blacks, whites, and Latinos. In 11 of the centers, a plurality of children
was white; in five, it was black; in seven, it was Latino. Nine of the 23 centers were
publicly funded by either Head Start or New York City’s Administration for
Children’s Services. We interviewed directors and other center staff on their
motivations for establishing interorganizational ties, the nature of those
ties, opportunities available to network, basic institutional practices, and the
general resources available to parents. We also observed parent meetings, field
trips, and other practices and center characteristics. The case studies allow
us to detail the particular mechanisms by which centers broker connections,
adding subtlety to the basic pictures painted by the quantitative data sets and
generating the core hypotheses about the nature of organizational brokerage.
The qualitative, individual-level data set consists of 67 in-depth inter-
views conducted by my research assistants and me with parents, most of
whom were mothers, in a subset of the case study centers. Our objective was
to understand, from the parents’ perspective, whether and how they had
formed ties in the center, how they understood these relations, and under
what circumstances they mobilized these ties. The interviews yielded the
richest data of all, providing a clear window into how ties are formed, what
types of ties these are, how trust and obligations operate, and how useful the
connections truly are for those purported to benefit from them.
This book stands with the sociological tradition in which the key to
understanding social processes is uncovering the mechanisms that give rise
to them—in this case, the mechanisms producing and reproducing inequality
in personal networks.81 I use the quantitative survey of mothers to frame the
findings in New York City and to empirically assess whether childcare center
participation is associated with larger networks and greater well-being. The
three remaining data sources I use to examine the mechanisms that give rise
to this association. To the extent it uncovers how brokerage operates in
centers, the book will have met its empirical objectives.
WHAT FOLLOWS
The chapters of the book are arranged in four parts. Chapter 2 concludes
part I by making the case, using the national survey of urban mothers
(Fragile Families survey), that childcare centers tend to be effective brokers
for the average mother, generating personal connections strongly associated
with greater well-being. Parts II and III, which constitute the core of the
book, answer the core empirical question: how did childcare centers shape
the mothers’ social and organizational ties?
The chapters in part II examine how centers affected mothers’ social ties.
Chapter 3, “Opportunities and Inducements,” asks why mothers so frequent-
ly made new friends in their childcare centers, given the demands on their
time that most of them reported. Chapter 4, “Weak and Strong Ties,” asks
whether the friends mothers made in centers were typically close friends,
acquaintances, or something else. Chapter 5, “Trust and Obligations,” ex-
plains why some mothers’ support networks in the center were, in fact, larger
than their friendship networks. It shows that participating in the center built
trust even among mothers who did not know each other personally, reveal-
ing how trust and obligations respond to formal imperatives sustained in
organizational contexts such as centers.
The chapters in part III examine how centers shaped mothers’ organiza-
tional ties. Chapter 6, “Ties to Other Organizations,” examines why mothers’
most useful ties were not always social. Shifting the focus from the mother to
the center and its relationships, the chapter reveals that many centers were
formally tied to other organizations—nonprofit entities, businesses, and
government agencies—that provided material and nonmaterial resources to
center parents. Chapter 7, “Organizational Ties and Neighborhood Effects,”
asks why centers bothered to establish these relationships with other orga-
nizations and what role location played in this process. Examining one
aspect of the persistent “neighborhood effects” question, the chapter exam-
ines whether childcare centers in poor neighborhoods were less well
connected than those in nonpoor neighborhoods, as would be expected by
standard theories about organizational capacity in poor areas.
Part IV extends the discussion beyond the confines of the childcare
center. Chapter 8, “Extensions and Implications,” documents the presence
of several mechanisms uncovered in this book—such as repeated and durable
interaction, the assignment of cooperative tasks, the institutionalization of
trust, and the application of validation, storage, referral, and collaboration—
among churches, diners, bathhouses, beauty salons, colleges, and other
routine organizations. In so doing, it makes a case for the importance of
organizational embeddedness in the study of personal networks. The chap-
ter concludes by suggesting that the organizational embeddedness perspec-
tive points to different kinds of questions in the current study of social
inequality.
I conclude this introduction by clarifying what this book does and does
not purport to accomplish. This book probes personal networks as under-
stood by the people who form and sustain them, and in the day-to-day
organizational contexts in which these processes take place. The book does
not constitute a standard social network analysis of the mothers’ personal
networks, a project in which relationships are conceived mainly as nodes
and the ties between them. The focus is not structure but interaction, thus
revealing those elements of social capital that formal structural analysis
largely sets aside. In addition, this book is not a study of children or child-
care, issues that have been studied many times before.82 It is also not, finally,
a study in the organization of centers—an examination of how centers
balance their books, hire and fire staff, select managers, develop curricula,
or strive to meet government regulations. Organizational sociologists will
learn little from this book about the issues that have traditionally concerned
them—in fact, I employ their concepts, and those of formal network analysis,
to inform and enrich our discussion.
Instead, this book examines one aspect of the relationship between the
center and the mother: how the former shapes the social and organizational
ties of the latter. It makes a case for a revaluation of social capital theory,
and offers a related but alternative perspective on personal ties, one
concerned less with choices than with contexts, less with “unceasing efforts”
than with structured interaction, less with purposive action than with unex-
pected gains. It uncovers how the manifold advantages that people gain
from their networks may derive from little more than those organizations
in which they happen to participate routinely.
2
Childcare Centers and Mothers’ Well-Being

Whether Mothers Did Better When Their Children
Were in Daycare
Effective brokers connect people to useful others. They not only enlarge
social networks but also produce those particular connections that yield
concrete benefits, such as well-paying jobs, better health, discounts on
services, or lesser material difficulty. Consequently, being embedded in
the right organizations can either lift an actor to a higher rung on the
socioeconomic ladder or, as explored in this book, improve her current
well-being. That is, it represents a significant advantage. I have suggested
that childcare centers, though they differ dramatically from one to the
other, nonetheless tend to be effective brokers for the mothers of enrolled
children, because they tend to generate useful connections. Do they?
And if so, then do mothers enrolled in centers thereby experience greater
well-being?1
The present chapter addresses those two questions. Based on the
national survey of urban mothers of young children (Fragile Families),
it tests whether mothers who enrolled their children in centers im-
proved their circumstances when compared to those who did not.2 The
chapter begins by grounding the analysis of the national data. Relying
mostly on the in-depth interviews with New York City parents, it
examines how and why parents decide to employ childcare centers in
the first place. Then, based on the Fragile Families survey data, it
assesses whether mothers who employed centers thereby enlarged their
social networks, particularly the number of friends they reported. The
third section asks whether these networks improved mothers’ well-being,
specifically, their ability to avoid material and mental hardship. The
results are consistent with expectations: (a) most mothers who patron-
ized childcare centers made new friends therein, and (b) making friends
in centers was strongly and robustly associated with lower material and
mental hardship.
28
Childcare Centers and Mothers’ Well-Being 29
CHOOSING CHILDCARE
While experts, journalists, and politicians continue to debate the merits of

placing children in centers, parents have decided to do so at rapidly rising
rates. Figure 2.1 exhibits the percentage of 3- and 4-year-olds in the United
States enrolled in nursery, preschool, or kindergarten from 1965 to 2005.
The figure reveals that centers have become an increasingly common choice
over the past generation. In 1965, only 11% of children 3 to 4 years of age
were enrolled in nursery, preschool, or kindergarten; by 2000, more than
50% were.3
Figure 2.1 suggests that practical considerations may be driving this
trend. Today, many more mothers of children under six participate in the
labor force.4 In 1965, less than a quarter of all such mothers were either
employed or searching for work; by 2005, the figure was 60%.5
To examine how the parents we interviewed arrived at the decision to
employ center-based care, I begin by identifying the options available in
New York City. I then examine how they weighed priorities and ultimately
made a choice. The trends in figure 2.1 hint at what I found: for most
parents, higher order worries, such as the quality of care provided, were
100%
90%
Percent married mothers of Percent single mothers of
children under 6 in LF children under 6 in LF
80%
70%
60%
50%
40%
30%
20% Percent 3–4 year-olds in nursery, pre-

school, or kindergarten
10%
0%
1965 1970 1975 1980 1985 1990 1995 2000 2005
Figure 2.1. Labor force (LF) participation of mothers of young children and enrollment
of young children in daycares, 1965–2005.
Sources: Historical Tables, Table A-2. Census Bureau, census.gov/population/www/
socdemo/school.html (accessed 2/7/07) and Statistical Abstracts of the United States,
various years, census.gov/compendia/statab/ (accessed 4/20/07).
surpassed by practical considerations, such as the need to find care quickly

in order to return to work. For many mothers, childcare centers were the
most affordable or practical option.
Diverse Options
Parents searching for childcare today face many options. These options may
be categorized into three general alternatives: (1) childcare centers, which
are formally licensed organizations—such as Head Start centers, prekinder-
gartens, and nurseries—in which more than five children are cared for,
usually in age-sorted classrooms, by teachers and staff; (2) family daycare
services, which are licensed or unlicensed homes in which one adult cares for
up to five children if licensed or more than that (though rarely much more)
otherwise; and (3) paid or unpaid informal care arrangements, by which
relatives, friends, babysitters, or nannies care for children in their or the
child’s home.
Formally licensed childcare centers vary widely in form and orientation;
consequently, they defy categorization. For the present purposes, I catego-
rize them by their primary funding source: government-funded nonprofit,
privately funded nonprofit, and for-profit centers.6 Those in the govern-
ment-funded category include Head Start centers, part of the federal
program designed to improve the social and cognitive development of low-
income children and integrate the family into the education of the child; and
locally funded centers, such as New York City’s Administration for Chil-
dren’s Services centers, aimed at providing subsidized care to low-income
and working-class families, under principles consistent with those of Head
Start.7 Other government-funded centers specialize in care for particular
populations, such as children with physical disabilities. Centers funded
primarily by the government are usually either free or deeply discounted
based on household income.
Privately funded nonprofit centers exist in many forms and styles of
operation. Some are run by large nonprofit organizations aimed at child
well-being. For example, the Children’s Aid Society, a 150-year-old organiza-
tion with operating expenses of more than $70 million, runs a half
dozen childcare centers, many in low-income areas, with active education
programs and no charge or nominal fees for income-qualifying families.8
Many nonprofit centers are run by religious organizations, such as
churches and synagogues, as part of their community service operations,
or with the explicit purpose of providing religious instruction. Still others
are operated by ethnic organizations (e.g., Puerto Rican, Dominican,
or Jewish nonprofit organizations) and provide group-specific services,
such as bilingual instruction and the celebration of religious holidays.
Among nonprofit centers, fees vary widely, and nonprofit status by no means
implies low cost. Some nonprofit centers are free; the rest of them range, for
a child 2 or 3 years of age, from about $75 to about $1,800 a month.9
For-profit centers also vary widely. Some are corporate centers with
multiple branches, others are private centers reserved for employees of
large companies, and others are small, family-run businesses. Most for-profit
centers in New York depend on private fees for funding; the least expensive
cost about $340 a month for a child 2 or 3 years of age. However, for-profit
centers also serve low-income clients. The federal and local governments
provide childcare vouchers to a number of populations, including women
enrolled in Temporary Assistance for Needy Families and foster parents,
and some for-profit centers rely heavily on these vouchers to run their
operations (which illustrates the difficulty in categorizing centers by funding
source). One for-profit center in our field study accepted government vouch-
ers from more than half of its clients.
A Decision Governed by Practical Concerns

How do parents decide among these options? Empirical studies suggest that
price and location may be the primary motivating factors. For example, in a
study based on the 1985 wave of the National Longitudinal Survey of Youth,
Sandra Hofferth and Douglas Wissoker found that center-based care was
often cheaper than other paid options and that price was a critical variable in
predicting which form of childcare families employed.10 In a survey of
parents in Maryland, about 80% preferred a childcare center near their
homes, and proximity was the most important factor associated with use of
center-based care, surpassing even cost, quality, or hours of operation.11
When we asked New York mothers and fathers what factors weighed on their
decisions, they also expressed concern about price and location, but they
reported numerous other worries as well. Many mothers declared the diffi-
culty of facing what Sharon Hays called an inherent contradiction in moth-
erhood, by which mothers are both expected to devote themselves to raising
their children and yet implicitly berated for doing house and care work
instead of “real” paid work.12 Others worried about the center’s hours of
operation, the availability of slots in their neighborhood, the safety of the
environment, the level of caring they felt during their visit to the site, the
amount of leave from work the mother could appropriate, the number of
different alternatives they would have to employ, and the available time
they had to search for options.
In spite of their diversity, their answers revealed two tendencies. For
many, finding care proved a difficult and exhausting process, drawn out over
months and complicated by the pull of multiple criteria; for most, the
decision to employ centers was governed not by philosophical but by practi-

cal concerns. When deciding among childcare options, parents might,
in theory, make two separate decisions: first, which type of childcare to
employ (whether center care, family care, or informal care); and second,
which particular center or family care provider or babysitter to employ
among the alternatives. In practice, however, the decisions were often
made jointly, as parents shifted and balanced preferences nonstop until
cost or convenience won out. Finding childcare in a city such as New York
required flexibility, hard work, some luck, either affluence or sacrifice, and,
often, the willingness to satisfice rather than maximize—that is, to opt for
the first available among the good-enough alternatives over the best of the
best, including settling on a center when one preferred a sitter. In this
process, having access to financial resources helped dramatically, but did
not guarantee an easy task.13
Consider Iris, a thin, petite, blond lawyer whose husband worked full time
for pay while she worked pro bono, from home, until she was ready to return
to paid employment.14 Not returning to work was economically infeasible.
Her only daughter, an infant, was enrolled in Little Friends Day Care, a
center in an upper middle-class neighborhood in the city, even though
Iris and her husband first preferred an at-home nanny. She explained why:
“[I]nitially I started looking for a nanny because my husband and I had
discussed it and we just thought that the flexibility that a nanny gives is
probably better suited to our needs. Unfortunately after interviewing about
ten candidates [we came] to the conclusion that we weren’t comfortable with
any of [them].” As she explained later in the interview, the nannies had not
met her standards.
There’s just, you know some of them barely spoke enough English to have a
conversation with us. Which doesn’t mean they’re not good and loving but our
concern is that if anything, God forbid, happened, I wasn’t sure that . . . they
mastered enough English to be able to call an ambulance, or call the police . . .
which was a huge issue for us. Also [for] a lot of them . . . we don’t know anything
about them . . . [since so] . . . many of them are illegal immigrants and you have
no recourse [if something bad were to happen]. . . . Also, I think having taken
some time off before I started studying for the bar I spent a great deal of time in
the park and in the neighborhood. And there are things about some of the
nannies that just bothered me. . . . I’ve seen so many children strapped to their
strollers for long extended periods of time while their nannies are either chatting
with their friends on their cell phone or they’re doing their own shopping. And
that’s just not the . . . sort of childcare that I consider adequate.
In spite of her impressions, one candidate met their expectations. But

the reality of the cost quickly set in.
IRIS: There might have been one candidate with whom we would have been
comfortable. . . . Because she had a background in child development and
she spoke very good English and all of that. [But] she actually got a
competing offer which was much out of our league [economically]. So, we
had to sort of scramble and in talking to another mother in the park she
suggested that I look into the Little Friends Day Care. . . .
INTERVIEWER: So you just met this parent in the park?
IRIS: That’s right. Someone, actually she was someone from the neighborhood
and since we had just moved into the neighborhood she knew of the place.
But she had said, “Well you can call them but usually it’s quite difficult
because there’s a long waiting list.” And I just happened to be very lucky
that [my daughter, at seven months] was at the age [that] she fit into the
slot that was open. ’Cause I actually have friends who have a little boy her
age who applied six months later and were on the waiting list for a year and
a half.
While Iris and her husband worried about quality, English-speaking abilities,
accountability, and the full attention that a nanny could provide, cost and
the opportune opening of a slot near their home ultimately settled their
decision.
In fact, it was not uncommon for parents to jump from alternative to
alternative during the process. Naomi, a white mid-level manager, placed
her name on a waiting list at a center near her home while she was pregnant,
since, given that her job had provided only three months of leave, she
wanted a backup plan. However, having grown up in a family daycare
setting, she wanted the same for her young daughter, believing it provided
the type of warmth and attention Naomi treasured. For this reason, she
searched for and located an organization that provided a list of licensed
family daycare providers, by Zip code:
I looked up my home Zip code, my work Zip code and everything in between,
figuring I could drop [my daughter] off on my way . . . . Many of
them . . . I didn’t even call because I knew the block they lived on and it was
not an acceptable neighborhood for me to take my child to. One I called and
she was full. . . . And my last resort [was] this woman who sounded nice on the
phone [and] invited me to come to her home, interview with her.
Naomi’s last resort had given her hope. However, she soon realized what she
would not settle for, when she visited the home with her husband:
It was so horrible. It was . . . [in] the projects down on [another neighborhood]
and we got to outside the building and I was like, “Oh this is it? I’m not going in
here,” and [my husband] was like, “You’ve got your cell phone, call her and
cancel the appointment.” I [thought it would be rude to cancel, so] . . . we went
in and there were . . . broken needles on the steps outside and little crack vials
and all kinds of sort of signage in this very dim and filthy lobby that made me
nervous. Oh, things about “Rat poison” and “Danger” and “Don’t touch
these wires; they’re live” and . . . this woman’s apartment was so . . . dark and
dirty and had no air and no light. And she’d done the best she could with
it . . . [and] . . . she was the loveliest woman . . . and already had one little baby
she was taking care of and we conducted the interview as though we were
serious about it. . . . [But in the end, I told my husband] “I’m not taking my
baby to the projects everyday.” And even for that, that was going to cost me
$300 a week. I mean, childcare’s expensive.
Naomi began to worry, because her backup plan, enrolling her child in a
center, was not materializing as she expected: “I had huge anxiety about
solving the daycare problem. . . . [W]hile I was still pregnant . . . I got on the
wait list, and the plan was [my daughter] was supposed to have gone to them
at three months old when I came back to work.” However, a space was not
available when she was ready to return to work.
I found out I had maybe two or three weeks before I was coming back to work
and I panicked. Nannies can cost, I’d say probably going rate for a nanny in
Manhattan is $500 a week. I also got quotes as high as $800 a week which
I thought was insane. Through the friend of a friend I found an amazing
woman who came to me for $400 a week. She was somebody else’s family
nanny . . . but they were going on vacation for two months and she had some
down time. And it was exactly the eight weeks that I needed until [the slot at
the center was available]. So, that was like somebody smiled on me.
In the end, she considered herself fortunate. “Yeah, I totally lucked out.”
The experiences of Iris and Naomi were made more difficult by the fact
that both their children were infants (for whom there are fewer slots) but
also easier by the fact that they could afford $400 a week in childcare. But
Irina, a second-generation Latina mother of two in her late twenties who
worked as a staff assistant in a government office, could not. While she had
strong preferences for center-based care, her only practical options were the
four childcare centers near her apartment in a predominantly Latino neigh-
borhood with a high poverty rate. One of them, a church-based center, was
not only open too few hours but also in a general state of disrepair:
[The church] has a daycare center now but it’s only [open] from like 9:00 to
12:00. That wasn’t working for me. They opened that up not too long ago, but
I wasn’t interested. . . . It was not . . . perhaps they’ve done renovations now, but
a few years ago it was not up to par. I was not pleased. A lot of the information
was outdated; a lot of their supplies were ruined. It was . . . it just looked really
bad.
Another center, by a local college, had recently opened: “I did have my

eye on [that] one. . . . [B]ut they wanted $275 a week and I couldn’t hack it.
And then . . . they did take [subsidized housing vouchers for low-income
families in the city], but I guess some parents were like starting to get
involved. So, by the time I did come around, they weren’t accepting it no
more.”
A third center, costing $375 a week, was even more out of reach. The final
center, Alegre, was an organization funded by the Administration for Chil-
dren’s Services in which parents paid on a sliding scale according to income.
When her first son, whom she had as a teenager, was almost two, she tried
enrolling him in the center, only to be met by a daunting wait list: “[My first
son] never had the privilege of going to daycare because [the waiting list
was] backlogged. Everybody was on waiting lists. And at the time, I was a
working parent. My income wasn’t enough to go in privately.” She had to use
ad hoc arrangements with family and friends to get herself through her
first son’s early years. She learned her lesson: “So, when Samuel [, my second
son, was born,] the second time around, you can say I did it smart and I put
him on the waiting list while I was pregnant. Not a lot of daycare centers
around here, I mean, ours is one of the most affordable.”
In sum, our interviews confirm and expand on the findings of the existing
literature: when parents enroll their children in centers, they do so after a
great deal of trial and error in which practical concerns have won out.
MOTHERS’ NETWORKS AND WELL-BEING
Many mothers felt fortunate to have found center-based care. Yet perhaps
they were more fortunate than they realized. The rest of the chapter
carefully engages the quantitative data to answer two questions—did
mothers make new ties as a result of enrolling their children in centers?
And was enrolling their children in centers associated with greater well-
being?
To answer these questions, I estimated statistical models based on the last
two waves of the Fragile Families survey, the national survey of urban
mothers identified in chapter 1 (for details, see appendix B). The survey
was first conducted immediately after mothers had given birth (wave 1), and
mothers were then reinterviewed when their child was one, three, and five
(waves 2–4). I refer to this child as the “focal child.” For simplicity, if the
focal child is enrolled in a center I refer to the mother as “enrolled.”15 Of the
two waves employed in this study (waves 3 and 4), only the latest (conducted
around the year 2004) contained information about mothers’ experiences in
childcare centers needed to answer all the questions posed here. Wave 3,
however, allowed me to take mothers’ prior circumstances into account.16
The survey has an important limitation: it does not provide informative
data on the alternatives to childcare that parents employed. For example,
enrolling in family care may benefit or constrain mothers just as much as

enrolling in a childcare center.17 I cannot assess this possibility because of
how the survey was administered; I can only compare mothers with children
in centers to all other mothers.18 The details behind the survey design and
all of the statistical models estimated in this chapter are described in
appendix B. Here, I mainly summarize the results and discuss the most
important aspects of the basic rationale behind my analysis.
Table 2.1 exhibits the basic characteristics of enrolled and non enrolled
mothers in the survey. The statistics in table 2.1 are weighted to account for
the design of the survey; numbers represent the characteristics of urban
mothers of children who were newborns 5 years earlier in cities with
populations of at least 200,000. In table 2.1, a “center” is any formal organi-
zation providing care, whether it is referred to as a Head Start, preschool,
prekindergarten, kindergarten, nursery, or any other type of center-based
care.19 (Later chapters examine which differences among centers affect the
networks of parents.)
Table 2.1 makes clear that enrolled and non enrolled mothers differ in
important ways. Enrolled mothers have slightly fewer children and higher
probabilities of being married, college educated, and employed or in school
full time. Notably, the number of adults in the household, which is a
standard indicator of social support, does not differ among the sets of
mothers. But perhaps the greatest difference is income, with the average
childcare center mother making more than twice as much (>$62,000) as one
Table 2.1. Characteristics of urban mothers of 5-year-olds with children in centers

and not in centers
Characteristic Focal child not in center Focal child in center
Married 41% 53%

Cohabiting 11% 9%
Number of children 2.86 (2.86) 2.61 (1.51)
[mean (SD)]
Worked or school 40 hr 58% 68%
High school degree only 61% 79%
College degree or more 13% 29%
Household income [mean (SD)] 28,661.06 (58,653.42) 62,163.45 (87,614.88)
Age [mean (SD)] 30.61 (11.96) 32.60 (7.31)
Number of adults in 2.00 (1.61) 2.00 (0.83)
household [mean (SD)]
Black 29% 36%
Latina 39% 26%
White 25% 30%
Source: Fragile Families Study. Results are weighted.

not in a childcare center. These numbers do exhibit large standard devia-

tions; there is considerable overlap between the income distributions of
the two sets of mothers.20 Nevertheless, these differences must be taken
into account statistically when comparing the networks or well-being of
enrolled and non enrolled mothers.
DID MOTHERS MAKE NEW FRIENDS IN CENTERS?
Before assessing the impact of center participation on well-being, I examined

mothers’ acquisition of new friendships. The survey specifically asked en-
rolled mothers, “Do you currently have any friends whom you’ve met
through your child’s center?” If they answered “yes,” they were asked how
many. The results are exhibited in figure 2.2.
Figure 2.2 makes clear that most enrolled mothers made friends in
centers. About 60% made at least one friend, and more than 40% made
three or more friends in centers. Given the prevalence of friendship forma-
tion, we should not be surprised if enrolled mothers had larger social net-
works than non enrolled mothers.
We shall see that they do, suggesting that centers increased enrolled
mothers’ total friendship network size. To be clear, the Fragile Families
survey cannot be used to state definitively that these mothers would not
have made new friends had they not enrolled in centers. Since wave 3 of the
survey did not ask how many friends mothers had before enrolling in centers,
70%
60%
Percentage of mothers
50%
40%
30%
20%
10%
0%
0 1 or more 2 or more 3 or more 4 or more 5 or more 6 or more
Figure 2.2. Number of friends met in center: Mothers with child in childcare center.
Source: Fragile Families Study.
0
Child is not in center Child is in center
Figure 2.3. Predicted number of close friends for a

statistically average mother of a 5-year-old.
Source: Fragile Families survey. Negative binomial regression
model; includes controls for income, race, age, education,
number of children, number of adults in household,
employment, and marital and cohabiting status. Robust
standard errors account for city clustering.
we do not know whether friendly mothers were more likely to enroll in

centers in the first place.21 Nevertheless, since we know for certain that
mothers made new friends through their centers, it is informative to compare
the networks of enrolled and non enrolled mothers before and after taking
into account center-made friends.
Figure 2.3 and table 2.2 present these results. All mothers in the survey
were asked how many close friends they had. They averaged 5.4 friends, and
8.4% of them described themselves as having no close friends whatsoever,
the worst type of social isolation.22 I ran separate statistical regressions
estimating the relationship between enrolling in a center and both (a) the
number of close friends and (b) the probability of having no close friends at
all. To ensure the mothers were as comparable as possible, I adjusted for
variables affecting whether mothers enroll their children in centers: income,
race, age, education, number of children, number of adults in the household,
employment status, and marital and cohabiting status. The estimates also
adjust statistically for several important elements of the survey design. (The
first model is a negative binomial regression; the second, a logistic regression
model. The details and technical justifications behind all models in this
chapter are discussed in appendix B. The coefficients for figure 2.3 and
table 2.2 are presented in appendix B, tables B.1 and B.2, respectively.)23
Figure 2.3 presents the results for the relationship between center enroll-
ment and total number of friends. For ease of interpretation, instead of
presenting regression coefficients, figure 2.3 exhibits the predicted number
of close friends separately for statistically average enrolled and non enrolled
mothers. Statistically average enrolled mothers have about 4.7 close friends;
non enrolled mothers have about 3.5 close friends. This difference of 1.2
friends is statistically significant. To assess whether those friends made in
centers could account for the difference, I reestimated the model reported in
figure 2.3 after adjusting for whether the mother had specifically made
friends in the center. After these friends were taken into account, the
difference was reduced to 0.2 friends and no longer statistically significant.
This is consistent with the idea that enrolled mothers have more friends
specifically because of the friends they made in centers.
Table 2.2 exhibits the results for the relationship between center enroll-
ment and the probability of having no close friends. Fewer than 10% of
mothers in the sample are this radically isolated, so the question is whether
enrolling in a center reduces that probability even further. As shown in
table 2.2, this appears to be the case. The predicted probability of having no
close friends is about 8% for a non enrolled mother, and 6% for an enrolled
mother. The difference of about 2 percentage points (or about a fourth of
the baseline probability) is statistically significant. As I did in the earlier
estimates, I separately assessed whether the friends mothers made in centers
could account for the difference in probability of being isolated. The differ-
ence was reduced to 1.3 percentage points and was no longer statistically
significant.
In sum, most mothers in childcare centers made friends there; as a result,
enrolling in a center is statistically associated with both having more friends
and being less likely to be isolated. We now turn to our broader question,
whether enrolling their children in centers, and making friends there, con-
Table 2.2. Predicted probability of being socially

isolated: Urban mothers of 5-year-olds (after controls)
Enrollment Status %
In center 5.8%
Not in center 7.7%
Difference statistically significant at 0.05 level.
Source: Fragile Families Survey. Logistic regression model;

includes controls for income, race, age, education, no. children,
no. adults in household, employment, and marital and
cohabiting status. Robust standard errors account for city
clustering.

Unanticipated Gains 1-50

Uploaded by

Copyright:

Available Formats

Unanticipated Gains 1-50

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Unanticipated Gains 1-50

Uploaded by

Copyright:

Available Formats

Unanticipated Gains

This page intentionally left blank

Mario Luis Small

Oxford New York

Copyright # 2009 by Oxford University Press, Inc.

Published by Oxford University Press, Inc.

Library of Congress Cataloging-in-Publication Data

York University were instrumental. I thank Irwin Garﬁnkel at Columbia and

PART I: PERSONAL TIES IN ORGANIZATIONAL SETTINGS

PART II: SOCIAL TIES

PART III: ORGANIZATIONAL TIES

PART IV: BEYOND CHILDCARE CENTERS

APPENDICES: A MULTIMETHOD CASE STUDY

Social Capital and Organizational

advantage, I suggest, might be either much greater or greatly diminished,

to the resources characteristic of their roles. If her college strongly enforced

them—respond to institutional constraints, imperatives, and opportunities.

SOCIAL CAPITAL THEORY

institutionalized relationships of mutual acquaintance and recognition”—

The Missing Question: How Do People Make Social Ties?

different reading, Bourdieu is arguing that, regardless of how they are

Why the Missing Question Matters

Consider that, without knowing how network inequality arises, developing

ASSUMPTIONS UNDERLYING THIS BOOK

it argues that participating in organizations that effectively broker social

Actors May Form Ties Either Purposely or Nonpurposely

of the pursuit of another aim. Consider a local example: in a grocery store a

Forming Social Ties Either Purposely or Nonpurposely Depends

enrolled in larger classes. In fact, opportunities to interact may be critical

The Context of Interaction Is Shaped Signiﬁcantly by Organizations

An institution in the normative sense may be deﬁned, following Victor

college was externally motivated to establish a place likely to build social

OTHER ASPECTS OF PERSONAL TIES

motivated to affect social capital by either internal or external pressures. The

What Organizational Contexts Affect

Social and Organizational Ties

Resources, Access, and Mobilization

Actor-Driven and Institution-Driven Brokerage

Purposive and Nonpurposive Brokerage

to their clothing. While such strategies vary in their effectiveness, they

the case of the Dominican pharmacist, and the institutional imperatives

We have seen that when people participate in organizations, they encounter

beneﬁt participants. One organization that, on average, is a remarkably

CASE STUDY: MOTHERS AND CHILDCARE CENTERS

Childcare centers are often thought of as the impersonal alternative to

Table 1.1. Principal data sources for the study

Individual Fragile Families and child well- In-depth interviews (67

Childcare Centers and Mothers’ Well-Being

While experts, journalists, and politicians continue to debate the merits of

20% Percent 3–4 year-olds in nursery, pre-

surpassed by practical considerations, such as the need to ﬁnd care quickly

A Decision Governed by Practical Concerns

decision to employ centers was governed not by philosophical but by practi-

In spite of her impressions, one candidate met their expectations. But

Another center, by a local college, had recently opened: “I did have my

MOTHERS’ NETWORKS AND WELL-BEING

enrolling in family care may beneﬁt or constrain mothers just as much as

Table 2.1. Characteristics of urban mothers of 5-year-olds with children in centers

Characteristic Focal child not in center Focal child in center

Married 41% 53%