Network Brokerage

and Competitive Advantage:

First Two Rules of Social Capital
Network Brokerage and Competitive Advantage (page 1)

For text on this session, see Chapters 1

and 2 in Brokerage and Closure (including
adjunct bits from Neighbor Networks).

Appendices:

I. Example Network Questionnaire for a Web Survey (from 2010 Neighbor Networks, 2017 Management
and Organization Review)
Strategic Leadership

II. Measuring Access to Structural Holes (from 1992, Structural Holes, 2010 Neighbor Networks)

III. Competitive Advantage in Social Networks and Stigler's "Economics of Information" (1961 Journal of
Political Economy)

This handout was prepared as a basis for discussion in executive education (Copyright © 2020 Ronald S. Burt, all rights reserved).
To download work referenced here, or research/teaching materials on related topics, go to www.ronaldsburt.com.
 This is Generic Organization Chart
Showing the Formal Network
in a Hypothetical Company
CEO

COO CFO HR

Americas Asia Europe

North South
Network Brokerage and Competitive Advantage (page 2)

Crop Seed Farm

Seeds
Protection Treatment Management

Corn
Strategic Leadership

Soybeans

Other
 Sociogram of Formal Network in a Large
EU Healthcare Company
CEO C-Suite Heir Apparent
Other, Respondent
Other, NonRespondent

Yanjie

Bob
Network Brokerage and Competitive Advantage (page 3)
Strategic Leadership
 Social Network EU and Emerging Asia
at the Top Markets
of the Company
Lines indicate frequent and
substantive work discussion;
heavy lines especially
close relationships.
Yanjie

US
Back
Office
Network Brokerage and Competitive Advantage (page 4)

B
Bob

Jim
B
B
B

B
B
Strategic Leadership

CEO R&D
C-Suite
Front Jie
Heir Apparent
Office
Other Senior Person

Figure 2 in Burt, "Network disadvantaged entrepreneurs" (Entrepreneurship Theory & Practice, 2019)
 _________________________________
Handin Assignment #1: Network Data Name, Section
(Complete this page, scan, and submit by midnight Saturday of the week when we finish discussing this handout.)
SOCIOGRAM
The "network" around a person is a pattern of relationships with and between colleagues. graphic image of
a network in which

A
This worksheet is completed in four steps: dots represent
nodes (a person,
(1) In the oval, write your first name. group, etc.) and
lines represent
(2) In the squares, write the first name or connections
nickname of five people with whom you
have had the most frequent and substantive
contact while you have been at Booth. This
could be other Booth students, professors,
staff, people with whom you B
work, or just good friends with
whom you have had frequent
E
Network Brokerage and Competitive Advantage (page 5)

and substantive contact while at

Booth. Please don't use full
names; just first names or
(3) Draw a line between each pair of contacts niknames. Ron will report
who are connected in the sense that they network averages in class
have frequent and substantive contact with session after these come in.
one another. If you're worried about Ron
seeing names, scratch out
(4) Compute network density (# / #possible). names before sending. Only
Count the number of lines between contacts the lines are network data.
(exclude relations between you and the
Strategic Leadership

D C
contacts). Divide by the number possible
(n[n-1]/2, where n is number of contacts,
which is 5 in this example). Multiply by 100
and round to nearest percent.

DENSITY = _____________

Appendix I contains an illustrative survey webpage used to gather network data.

 YOU YOU YOU

Network A Network B Network C

Network Brokerage and Competitive Advantage (page 6)

Redundancy
YOU
Network & Information
by Cohesion

Contact
Redundancy
Contacts as
Strategic Leadership

Redundancy
by Structural YOU Source vs. Portal
Equivalence

from Figures 1.1 and 1.3 in Burt (1992, Structural Holes) and Figure 1.2 in Brokerage and Closure
 Network Structure Maps Distribution of Information

Paul Bob Merton Solomon Asch Leon Festinger Hal Leavitt Elihu Katz
Lazarsfeld 1910-2003 1907-1996 1919-1989 1922-2007 1926 -
1901-1976
Network models of advantage are grounded in two facts about the social distribution of
information from the 1950s “golden age” of social psychology (e.g., Festinger, Schachter & Back, 1950;
Network Brokerage and Competitive Advantage (page 7)

Asch, 1951; Leavitt, 1951; Katz & Lazarsfeld, 1955): (1) people cluster into groups as a result of interaction
opportunities defined by the places where people meet, and (2) communication is more frequent and
influential within than between groups so that people in the same group develop similar views.
People tire of repeating arguments and stories explaining why they believe and behave the way they do.
Within a group, people create systems of phrasing, opinions, symbols and behaviors defining what it means to be a
member. Beneath the familiar arguments and experiences are new, emerging arguments and experiences awaiting a
label, the emerging items more understood than said within the group. What was once explicit knowledge interpretable
by anyone becomes tacit knowledge meaningful primarily to insiders. With continued time together, information in the
group becomes “sticky” – nuanced, interconnected meanings difficult to understand in other groups (Von Hippel,
Strategic Leadership

1994). Much of what we know is not easily understood beyond the colleagues around us. Holes tear open in the flow of
information between groups. These holes in the social structure of communication, or more simply structural holes
(Burt, 1992), are missing relations indicating where information is likely to differ on each side of the hole and not flow
easily across the hole. In short, the bridge and cluster structure in social networks indicates where information is
relatively homogeneous (within cluster) and where information is likely to be heterogeneous (between clusters).
From Burt, "Network disadvantaged entrepreneurs" (Entrepreneurial Theory and Practice, 2019, page 22)
 Bridge & Cluster: Small World of Organizations & Markets

A
1

B
7 3
Network Brokerage and Competitive Advantage (page 8)

2
James

Robert

5
4
6

Density Table
C&D 85 Group A
5 25 Group B Network
indicates
Strategic Leadership

0 1 100 Group C
Network Constraint distribution
(C = Σj cij = Σj [pij + Σq piqpqj]2, i,j ≠ q) 0 0 29 0 Group D of sticky
person 3: .402 = [.25+0]2 + [.25+.084]2 + [.25+.091]2 + [.25+.084]2
information,
which defines
Robert: .148 = [.077+0]2 + [.154+0]2 + [.154+0]2 + [.154+0]2 + [.154+0]2 + [.154+0]2 + [.154+0]2 advantage.

From Figure 1.1 in Brokerage and Closure.

 Social Network JIM and JIE are
EU and Emerging
WARLORDs in their Asia
at the Top Markets
businesses, Illustrating
of the Company Rule 3 of Network
Lines indicate frequent and Advantage:
substantive work discussion;
heavy lines especially Close the
close relationships. network around your
contacts
Yanjie to promote trust
and efficiency.
US
Back
Office
(page4)9)

B
Advantage(page

B
Bob
Competitive Advantage

Jim
B
B
B
and Competitive

B
B
Front
Analysis
Brokerage and
Leadership

Office
NetworkBrokerage

CEO R&D
Social Network

C-Suite
Front Jie
Heir Apparent
Office
Strategic
Network

Other Senior Person

Figure
(“The Bull,” 1917 2 in political
Berlin Burt, "Network
cartoondisadvantaged entrepreneurs" (Entrepreneurship Theory & Practice, 2019)
of Bavarian bourgeois)
 Strategic Leadership
Network Brokerage and Competitive Advantage (page 10)

(Huateng "Pony" Ma, founder-CEO Tencent)

 Here is the core network for a job BEFORE and AFTER the employee
expanded the social capital of the job by reallocating network time and
energy to more diverse contacts.

Create Value It is the weak contact connections (structural holes) in

the AFTER network that provides the expanded social
by Bridging capital.

Structural 2
1
The employee AFTER is more positioned
at the crossroads of communication
Holes between social clusters within the firm
BEFORE and its market, and so is better
3
positioned to craft projects and

53
.6 con
STICKY INFORMATION 4
policy that add value across
Information expensive to move clusters.

s
5
because: (a) tacit, (b) complex,

t rai
(c) requires other knowledge to

nt
absorb, or (d) interaction with
sender, recipient, or channel.
information
Network Brokerage and Competitive Advantage (page 11)

Research shows that

STRUCTURAL HOLE employees in networks
breadth,
disconnection between two like the AFTER network, timing, and
groups or clusters of people arbitrage
spanning structural holes,
are the key to integrating
BRIDGE
operations across functional 20
relation across structural hole .0 1
and business boundaries. In co
ns 2
research comparing senior people tra
NETWORK ENTREPRENEUR int
*
with networks like these BEFORE and
or "broker," or "connector:"
AFTER networks, it is the AFTER networks AFTER
a person who coordinates 3
that are associated with more creativity, faster
across a structural hole
Strategic Leadership

learning, more positive individual and team 4

BROKERAGE evaluations, faster promotions, 5

act of coordinating across and higher earnings.
a structural hole
*Network scores refer to direct contacts.

From Figure 1.4 in Burt (1992 Structural Holes), and Figure 1.2 in Brokerage and Closure.
See Appendix I on survey network data, Appendix IV on measuring network constraint.
 (Q201) Sociograms of the bridge-and-cluster structure to an organization or its
market provide a map of how information is distributed. All of the below are
true except:

A. Clusters indicate where information Bill

is likely to differ.
Bob

B
B

B. Clusters indicate where information

B

is likely to be sticky.
CEO
C-Suite B
Network Brokerage and Competitive Advantage (page 12)

Heir Apparent
Other Senior Person

C. The lack of bridge relations indicates where there is a problem

for integrated company operations.

D. Bridge relations indicate which people are positioned to be

network brokers.
Strategic Leadership

E. Clusters indicate where there are structural holes in the

organization or market.
 (Q261) Information and ways of understanding are more
homogeneous within than between social clusters. Being able to
see the boundaries between social clusters is therefore critical to
identifying the structural holes that define rewarding opportunities
for brokerage. Fortunately, (circle best completion to the
sentence):

A. the boundaries around social clusters are sharply defined by the

Network Brokerage and Competitive Advantage (page 13)

absence of connections between in clusters.

B. the boundaries around social clusters are often ambiguous but can be
identified by the absence of connections between clusters relative to
presence within clusters.

C. the boundaries around social clusters are often ambiguous so we define

them by talking with people in each cluster to learn whether opinion and
Strategic Leadership

practice in one cluster differs from opinion or practice in the other cluster.

D. successful network brokers can ignore boundaries.

 (positive evaluation, high compensa
Z-Score Business
MEASUREMENT: contrast is between people rich in
access to structural holes versus people without
(cosmopolitans vs locals in Merton 1949; opinion leaders vs followers in Katz & Lazarsfeld 1955; extensive
vs intensive search in Rees 1966; leaders vs managers in Kotter 1990; exploration vs exploitation in March
1991; cultural omnivores vs univores in Peterson 1992; open vs closed networks, on the edge of worlds
vs at the center; and of course, Schumpeter's 1911 touchstone image of entrepreneurial "leaders" bringing
together elements from separate production spheres within which people live by routines)

Disconnected 100% in one

contacts group provides
provide rich r = -.73 no access
access to to structural
structural holes
Network Brokerage and Competitive Advantage (page 14)

holes
Network Constraint
— few many ——— Structural Holes ——— few
Here network constraint – the extent to which a person’s network is limited to a
single group, which means they have no access to structural holes (other popular
measures are size, density, and ego-network betweenness). Constraint increases
as a network becomes small (few alternative contacts), dense (strong relations
Strategic Leadership

between contacts), or hierarchical (central contact holds others together)

Data are easily available from surveys, 360˚, email, and other electronic trace
(badges, chat rooms, social media, virtual worlds, etc.).
See Appendix I on network survey data, and Appendix II on measuring access to structural holes.
 In-Class Worksheet for Network Metrics
Ego is not presented. These are just ego's key contacts. All connections are symmetric and binary (zei = zej = 1).
EgoNetwork Betweenness = ∑i ∑j>i ([zei zej - zij] / [∑k zki zkj]), k ≠ i, j, and index k includes ego e.

A
A A A
B

C B C B C B D
C

size (degree) _____ _____ _____ _____

density (# / #possible) _____ _____ _____ _____
betweenness (holes) _____ _____ _____ _____
Network Brokerage and Competitive Advantage (page 15)

A B C D
A A A
A __
E B E B E B
B __

D C D C D C C __

D __
Strategic Leadership

size (degree) _____ _____ _____

density (# / #possible) _____ _____ _____
betweenness (holes) _____ _____ _____

Discussed in Appendix II.

 Handin Assignment #2: Network Metrics
(Complete this page, scan, and submit by midnight Saturday
of the week when we finish discussing this handout.) _________________________________
Name, Section
Compute the network metrics for ego in these six networks. Ego is not presented; just
ego's key contacts. All connections are symmetric and binary (zei = zej = 1).

A A A

E B B E B
D

D C C D C

size (degree) _____ _____ _____

density (# / #possible) _____ _____ _____
Network Brokerage and Competitive Advantage (page 16)

betweenness (holes) _____ _____ _____

A A A

B E B E B
D
C D C D C
Strategic Leadership

size (degree) _____ _____ _____

density (# / #possible) _____ _____ _____
betweenness (holes) _____ _____ _____
 Network Metrics
for More Usual Networks
Graph plots ego-network betweenness scores
few ——— Structural Holes ——— many

against network constraint scores for a probability

sample of 700 Chinese entrepreneurs
Network Betweenness

See Appendix II for more on similarity between

alternative measures of access to structural holes.
Network Brokerage and Competitive Advantage (page 17)
Strategic Leadership

r = -.92

Network Constraint
many ——— Structural Holes ——— few
 (Q149) Compute network density for John's network (to simplify
the sociogram, John's connections are not shown).

A. 30%

B. 47%
Network Brokerage and Competitive Advantage (page 18)

C. 60%

D. 70%

E. 90%
Strategic Leadership
 (Q245) Compute network betweenness for John's network
(number of structural holes to which John has monopoly access; to
simplify the sociogram, John's connections are not shown).
Susan

A. 0.5

B. 1.7 Pierre Elena

John
Network Brokerage and Competitive Advantage (page 19)

C. 2.0

D. 2.3 S — Robert
Yanjie
E 0 —
E. 5.0 R .33 0 —
Strategic Leadership

Y 0 .5 .5 —
P 0 .33 0 0 —
 Social Network EU and Emerging Asia
at the Top Markets
of the Company
Lines indicate frequent and
substantive work discussion;
heavy lines especially
close relationships.
Yanjie

US
Back
Office B
Competitive Advantage (page 20)

B
Bob

Jim
B
B
B

RULE 1, Brokers Do Better: For top-line growth, large open

networks
B facilitate creativity,
B
innovation, and achievement via
information breadth, timing, and arbitrage advantages from
and 3)
Leadership

CEO bridging structural holes (Milgram 1969; Granovetter R&D1973;

(page
Leadership

C-Suite
Freeman 1977; Burt Front 1980, 1992, 2005; 2021; Lin et al. 1981;
Brokerage
Foundations

Heir Apparent
Gould &Back FernandezOffice1989; Ahuja 2000; Lin 2001; Aral & Van
Strategic

Other Senior Person

Alstyne Office
2011; Fleming & Waguespack 2007; Zaheer & Soda
Strategic
Network

2009;
Figure Goldberg
1 in Burt, et al. 2016;
"Network disadvantaged Soda, Tortoriello
entrepreneurs" & Iorio
(Entrepreneurship 2018).
Theory & Practice, 2019)
 Now to establish the
Define Z-Score empirical fact that

(compensation, evaluation, promotion rate)

the people known as

Raw Performance Indicator

Bob’s performance "network brokers"
Z-Score Relative Performance

enjoy achievement
is higher than
expected

Jim’s performance
is lower than and rewards higher
expected
than their peers.

Manager Background Brokers are to the

Network Brokerage and Competitive Advantage (page 21)

(e.g., job rank, age, geography, kind of work,

organization division, education, etc.)
left on the horizontal
axis contrasting
open with closed
networks.
Strategic Leadership

Network Constraint (x 100)

large, open
many ——— Structural Holes ——— few small, closed
Robert James
 Define Z-Score

(compensation, evaluation, promotion rate)

Achievement and

Raw Performance Indicator

Bob’s performance rewards are
Z-Score Relative Performance

is higher than
expected
distinguished on the
Jim’s performance
is lower than
expected
vertical axis,

measuring the
Manager Background extent to which a
Network Brokerage and Competitive Advantage (page 22)

(e.g., job rank, age, geography, kind of work,

organization division, education, etc.)
person is doing
better than his or
her peers.
Strategic Leadership

Network Constraint (x 100)

large, open
many ——— Structural Holes ——— few small, closed
Robert James
 Business Success Decreases
as the Network Around a Person Closes
Define Z-Score
(compensation, evaluation, promotion rate)
Raw Performance Indicator

Z-Score Relative Performance

Bob’s performance median network
is higher than constraint (49 points)
expected

Jim’s performance
is lower than
expected
Network Brokerage and Competitive Advantage (page 23)

Manager Background
(e.g., job rank, age, geography, kind of work,
organization division, education, etc.)

Managers in the U.S.

(n = 2085, 7 study pops, r = -.75)

Managers in Europe
(n = 1094, 3 study pops, r = -.73)

Managers in Asia, mostly China

Strategic Leadership

(n = 1342, 3 study pops, r = -.78)

Network Constraint (x 100)
many ——— Structural Holes ——— few

NOTE — Plotted data are average scores within five-point intervals of network constraint within each study population (2018 survey added to Burt, Social
Networks 2019: Figure 1; see footnote 2 there for data sources; cf. Figure 1.8 in Brokerage and Closure). Correlations are computed from the plotted data
using log network constraint. Inset graph to the upper left contains hypothetical data illustrating computation of z-score relative performance.
 (Q173) For network brokerage to provide advantage, what is the
most essential quality required in the surrounding organization or
market?

A. Exclusive access to people

Network Brokerage and Competitive Advantage (page 24)

B. Multiple social clusters

C. Information accuracy

D. Information variation
Strategic Leadership

E. Large numbers of people

 Variation Indicates Sticky Information. Below graphs
show variation in fish prices before and after cell phones
are available to fishermen and wholesalers.
Weekly surveys
were conducted
with sample
wholesalers in
three regions
for a common
category of fish
Network Brokerage and Competitive Advantage (page 25)

sold (sardines).
Regions are
administrative
districts in the
Indian state of
Kerala.

Network brokers
Strategic Leadership

are a mechanism
that clears sticky
information in a
Figure 4 in Jensen, "The digital divide: information (technology), market performance, and
market. welfare in the south Indian fisheries sector" (2007 Quarterly Journal of Economics). See
Appendix III for network broker analogy in Stigler's economics of information.
 (Q150) Using network density as a rough indicator of John's
access to structural holes, do you expect him to be doing well or
not so well in his career?

A. Well, John’s contacts are well

connected.

B. Well, Susan serves as a

Network Brokerage and Competitive Advantage (page 26)

partner to John.

C. Not so well, John is disconnected.

D. Not so well, too many connections missing between

contacts
Strategic Leadership

E. Not so well, too many connections between contacts.

 (Q186) The displayed graph illustrates the competitive advantage enjoyed by
network brokers by showing that brokers enjoy compensation, evaluations, and
promotion rates better than peers. True or false?

2.5

(compensation, evaluation, promotion)

A. True
Z-Score Relative Performance
1.5

B. False 0.5
Network Brokerage and Competitive Advantage (page 27)

0.0

-0.5

-1.5

Z = 2.78 - .82 ln(C)

Strategic Leadership

-2.5 r = -.53
5 15 25 35 45 55 65 75 85 95

Network Constraint (C)

many ——— Structural Holes ——— few
 Second Life prediction from Second Life prediction from

Returns to
nonredundant social contacts constrained social network
Z-score Achievement

Predicted Avatar Z-score Achievement

Brokerage Are
Also Evident in
Online Networks.
Avatar

These are
28)

the returnsEverQuest
to II prediction from
Predicted(page

brokerageversus
innonredundant
two(lower)
Network Brokerage and Competitive Advantage

social (upper)
economic contacts
EverQuest II prediction
virtual worlds. from constrained social (upper)
versus economic (lower) network

25+

Effective Size Network Constraint (x 100)

(Number of NonRedundant Contacts)
Strategic Leadership

Dots are average Y scores within integer (left) or five-point (right) intervals on
horizontal axis. EverQuest II achievement variable is the predicted character
level in Model 8, Tables 3.4 and 3.5. Second Life achievement is the canonical
correlation dependent variable in Model 15, Tables 3.5 and 3.6.

from Burt (2020 Structural Holes in Virtual Worlds).

 3
7 7

1 2 2

3 3
7 7

Monopoly Broker Disconnected Brokers

5 (WHEEL) Network (DB) Network 6

ID Position
Network Brokerage and Competitive Advantage (page 29)

5 5
1 6-hole broker
2 3-hole broker
4 4
3 1-hole broker
4 3-hole broker 6 6
5 5
5 5-person clique
6 3-person clique Connected Brokers
CLIQUE (CB) Network
Network 7 pendant
Strategic Leadership

(or, overlapping cliques)

Even with Random Assignment to Networks,

Network Brokers Are Perceived To Be Leaders (setup)
 t-test Network
beta (N=385) Constraint
-.29 -5.24 Defined

(Leader Cites to Subject / Total Leader Cites in Team)

-.45 -6.74 Behavioral
-.63 -13.91 Relative
Percent Leader Cites

Person half
as constrained
as average
teammate
gets 70% of
leadership
vote.
Network Brokerage and Competitive Advantage (page 30)

Person as constrained as
average teammate gets
20% of leadership vote.
r = -.94 with ln(relative network constraint)

Relative Network Constraint

(Actual Constraint on Subject / Average Actual Constraint on Teammates)
Strategic Leadership

Even with Random Assignment to Networks,

Network Brokers Are Perceived To Be Leaders (results)
Data are averaged within .05 intervals of relative network constraint. Inset table contains standardized regression coefficients
and test statistics for constraint predictions with controls. Burt, Reagans, and Volvovsky (Social Networks, 2021:Fig 10).
 Returns to Brokerage Are Evident
in Low Returns to Over-Specialized Students
Recent scholarship on the returns to labor market specialization often claims
that being specialized is advantageous for job candidates. We argue, in contrast,
that a specialist discount may occur in contexts that share three features: strong
institutionalized mechanisms, candidate profiles with direct investments that
signal their value, and a high supply of focused candidates relative to demand.
We then test whether there is a specialist discount for graduating elite MBAs,
as it is a labor market that exemplifies these conditions under which we expect
specialists to be penalized. Using rich data on two graduating cohorts from a top-
tier U.S. business school (full-time students, 2008-2009), we show that elite MBA
graduates who established a focused (specialized) market profile of experiences
relating to investment banking before and during the program were less likely to
receive multiple job offers and were offered less in starting-bonus compensation
than similar MBA candidates with no exposure or less-focused exposure to
investment banking. Our theory and findings suggest that the oft-documented
specialist advantage may be overstated.
Network Brokerage and Competitive Advantage (page 31)

Figure 1 displays predicted (marginal) probabilities of receiving multiple offers for

candidates who have mean values for each of the control variables but different
profiles.

Figure 2 compares the starting bonuses of hypothetical job candidates with different
profiles. Each hypothetical candidate is a single white male who graduated from a
top-20 undergraduate institution, has above a 3.8 GPA, received more than one
job offer, has the mean age and work experience characteristics (months, number
of firms), accepts a job in I-banking, and earns the mean base salary for I-banking
jobs in his 2008 cohort year. The only difference is the candidate’s profile in terms
of exposure to I-banking.

FOCUSED (career history in finance before mba, concentration in finance, joined

an i-banking club during mba, and i-banking internship; 61% of students who
Strategic Leadership

graduate to a job in i-banking were focused on i-banking)

NON-SEQUENTIAL exposure (neither of the above categories, but some mba
program contact with i-banking)
PARTIAL sequential exposure (prior experience in finance + concentration in
finance or participation in i-banking club)
Figures and text are from Merluzzi and Phillips (2016 Administrative Science Quarterly),
PRE-MBA exposure (only exposure before mba program)
“The Specialist Discount." For more applied discussion, see Merluzzi, (June 2016 HBR),
"Generalists get better job offers than specialists." Looking later in the career, Kleinbaum (2012 ASQ) "Organizational misfits,"
shows with email data that managers with unusual patterns of communication are most likely to emerge the valued network brokers.
 Returns to Brokerage Aggregate
to Companies, Industries, and Communities
People with phone networks
that span structural holes
live in communities higher
in socio-economic rank
Networks are defined by land-line & mobile
phone calls (map to left). Socio-economic
rank is UK government index of multiple
deprivation (IMD) based on local income,
Network Brokerage and Competitive Advantage (page 32)

employment, education, health, crime,

housing, and environmental quality (graph
below). Units are phone area codes.
Strategic Leadership

figures from Eagle, Macy, and Claxton (2010 Science), “Network diversity and economic development”
 (Q259) We discussed the below graph of prices before and after a
technology change. The transition from wide variation to narrow
variation shows the effect of:

A. Broader access to market

Network Brokerage and Competitive Advantage (page 33)

information

B. Less noise in previously

available information

C. More competitors in the market

Strategic Leadership

D. Fewer competitors in the market

E. More sticky information in the market

 (Q15) The great thing about building a
brokerage network is that once you have A. True
it in place, you can sit back and enjoy your
competitive advantage over your peers. B. False
True or false?
Network Brokerage and Competitive Advantage (page 34)

(Q83) Returns to brokerage depend on

A. True
senior management recognizing the
importance of social networks for
B. False
employee performance. True or false?
Strategic Leadership
 Network Broker Achievement
(Q289) Bridging structural holes can improve the odds of
achievement, but achievement improves the odds of
bridging structural holes. We cannot say that network
structure is always causal for achievement, but given
supportive evidence from experiments with random
Network Brokerage and Competitive Advantage (page 35)

assignment to networks, we believe it can be causal. True

or False?

A. True.
Strategic Leadership

B. False.
 HOW IT WORKS: Creativity and Innovation
Are at the Heart of It

Achievement & Rewards

(What benefits?)
Brokerage
across
Structural Holes What in your work
Adaptive Implementation
(How to frame it & who should be involved?)
improves the odds
that you will discover
Network Brokerage and Competitive Advantage (page 36)

the value of something

you don't know you don't know?
Creativity & Innovation
(What should be done?) Alternative Perspective (how would this problem look from the perspective of
a different group, or groups — thinking “out of the box” is often less valuable than seeing
the problem as it would look if you were inside a specific “other box”)

Best Practice (something they think or do could be valuable in my operations)

Strategic Leadership

Analogy (something about the way they think or behave has implications for how I can enhance the value of my operations; i.e., look for the value of
juxtapositioning two clusters, not reasons why the two are different so as to be irrelevant to one another — you often find what you look for)

Synergy (resources in our separate operations can be combined to create a valuable new idea/practice/product)
from Burt, "The social capital of structural holes" (2002 The New Economic Sociology). The consequences of the
information diversity associated with network brokerage is productively elaborated at length in economist Scott
Page's 2007 book, The Difference: How the Power of Diversity Creates Better Groups, Firms, Schools and Societies.
 Illustration: Where did the M-16 come from?
Network Brokerage and Competitive Advantage (page 37)

Discussion Question*
Consequential ideas are typically attributed to special people, geniuses, in part to make us feel less
uncomfortable about our own ideas. True to form, an American armament expert describes Eugene
Stoner, the engineer who developed the M-16 assault rifle, as "an engineering genius of the first order."
Another describes him as "the most gifted small-arm designer since Browning." (Browning patented the
Strategic Leadership

widely-adopted BAR and 45 automatic.)

1. Based on the brief history video, how would you describe Stoner's genius?
2. What circumstances might allow you or your colleagues to be as creative?
*Photos are from the video shown during the session. For discussion and references, see page 73 in Brokerage
and Closure. For sampling on the dependent variable, see Rosenzweig, “Misunderstanding the nature of
company performance: the halo effect and other business delusions,” 2007 California Management Review.
 Brokerage, Good Ideas, and Innovation,
Digging a Little Deeper
J
3.5 0.9
E E

J
J 0.8
E J
E
Y = a + b ln(C) ^
3 E P(no idea)
Management Evaluation of Idea's Value

^
a
^b t 11.2 logit test statistic 0.7
E Judge 1 6.42 -1.04 -5.8
J
Judge 2 4.08 -.63 -3.9 J
0.6
G E Combined 5.51 -.91 -7.4
E J
2.5 G
E J

Probability
J
E 0.5
Network Brokerage and Competitive Advantage (page 38)

G E C
E
E J J C
J 0.4
G
2 E J C C C
E J C ^ C 0.3
P(dismiss)
G J 5.5 logit
G
J C C C C
J test statistic
G
G G J C 0.2
1.5
G G
E
G ". . . for those ideas that were
C either too local in nature, 0.1
Strategic Leadership

C
G incomprehensible, vague,
G C
G G or too whiny, I didn't rate them"
1 C 0
10 20 30 40 50 60 70 80 90 100 10 20 30 40 50 60 70 80 90 100

Network Constraint (C) on Manager Offering Idea

from Figure 2.1 in Brokerage and Closure (or Figure 5 in Burt, "Structural holes and good ideas," 2004 American Journal of Sociology,
point is elaborated in Burt and Soda, "The social origins of great strategies," 2017 Strategy Science).
 Network Brokers Use More Familiar Language

Network Brokers: Clique Managers:

Relatively Average Relatively Closed Probability
Open Networks Networks Networks No Difference

Familiar Text 56.32 46.67 34.34 P < .001

Network Brokerage and Competitive Advantage (page 39)

Idea Dismissed 14.38% 36.94% 43.42% P < .001

NOTE — Columns distinguish the bottom, middle, and top third of supply-chain managers on network constraint
(horizontal axis in the two graphs on the previous page). “Familiar Text” is the number of words in a manager’s text
that are familiar in the sense that they are found in the LIWC language software dictionary. Probability test is based on
a -9.49 z-score from a Poisson regression of word count over the three network categories (-1, 0, and 1), controlling
for number of words in the manager’s idea text. “Idea Dismissed” is the percent of managers whose idea is dismissed
Strategic Leadership

by the executives as not worth rating. Probability test is based on a 5.14 chi-square with 2 degrees of freedom,
controlling for number of words in the manager’s idea text.
Table 1 in Burt, "Social network and creativity" (Handbook of Research on Creativity and Innovation,
2020). For more general results on broker advantage depending on brokers using familiar language, see
Goldberg, Srivastava, et al., "Fitting in or standing out?" (2016, American Sociological Review)
 Three Team Histories for Person A
Further Back Prior Final
Relatively closed
A A A Colleagues = 3
team history (Ca
B C B C B C Constraint = 92.6
score of 92.6 is a
D D D
z-score of 0.8)

A A
B C B C
A Colleagues = 5
About average D
B C Constraint = 59.9
team history D
B C B C
D D
Network Brokerage and Competitive Advantage (page 40)

A
B C
A
B C
A A Colleagues = 8
D
Relatively open B C B C Constraint = 33.1
team history (Ca D
score of 33.1 is a B C
z-score of -1.2) A D
Strategic Leadership

B C NOTE — Persons B, C, and D are members

in A’s final team. Each dot is a different
B C person in prior teams. “Colleagues” is the
D number of people with whom A has worked.
B C Constraint is 100 x A’s network constraint
D score (horizontal axis in Figure 2B).

Figure 3 in Burt, "Social network and creativity" (Handbook of Research on Creativity and Innovation, 2021).
 Episode Role
Creativity Creativity
-0.018 -0.018
Constraint
(-6.73) (-6.44)
0.003 -0.005
# Episodes
(0.51) (0.36)
Maximum Episode Creativity

-0.013 -0.014
# in Fallow Period
(-0.99) (-0.73)

Intercept 5.32 5.30

R2 .31 .29

N observations 200 200

Maximum episode creativity

Mean of the other two measures

Maximum role-creativity in episode
Network Brokerage and Competitive Advantage (page 41)

Network Constraint
(lack of structural holes within and between
producer-director-writer teams in which person worked)

Maximum Career Creativity by Career Access to Structural Holes

Graph is from Soda, Mannucci, and Burt (2018). The observations are all 200 people who worked as producers, directors, or writers in any of
the 273 episodes of the BBC series, Dr. Who. The horizonal axis is a person’s network constraint score for the network of people with whom
the person worked. High scores indicate the person worked with people who primarily worked with one another. Low scores indicate the
Strategic Leadership

person worked with many different people, who themselves came together from working with many different people. Constraint and creativity
are averaged within 5-point intervals on the horizontal axis (two intervals containing a single person are combined with the closest adjacent
interval). Creativity is measured on the vertical axis in two ways: (1) maximum creativity score a person ever received for an episode on
which s/he worked (mean 1-5 creativity score from two expert critics, hollow circles), and (2) maximum creativity score a person ever received
for his or her role as producer, director, or writer (mean 1-5 creativity score from two expert critics, hollow squares). The table to the right
contains OLS regression models showing the strong creativity-network association after holding constant the number of episodes on which a
person worked, and the person’s number of episodes during a fallow period in the Dr. Who series (coefficients presented with test statistics in
parentheses). Picture is an evil alien in the series.
 Number highly creative episodes
Mean of the two measures Episode Role
Number of high role-creative episodes Creativity Creativity
Number of High-Creativity Episodes

-1.11 -0.92
Log Constraint
(-8.11) (-6.32)
0.04 0.05
# Episodes
(5.81) (6.89)
-0.04 -0.05
# in Fallow Period
(-5.19) (-5.19)

Intercept 4.18 3.27

Pseudo R2 .46 .42

N observations 200 200

Network Brokerage and Competitive Advantage (page 42)

Aggregate
Network Constraint Career Creativity
(lack of structural holes within and between by Career Access
producer-director-writer teams in which person worked)
to Structural Holes
Graph is from Soda, Mannucci, and Burt (2018). The observations are all 200 people who worked as producers, directors, or writers in any
of all 273 episodes of the BBC series, Dr. Who. The horizonal axis is a person’s network constraint score for the network of all people with
whom the person worked. High scores indicate the person worked with people who primarily worked with one another. Low scores indicate the
person worked with many different people, who themselves came together from working with many different people. Constraint and creativity are
Strategic Leadership

averaged within 5-point intervals on the horizontal axis (two intervals containing a single person are combined with the closest adjacent interval).
Creativity is measured on the vertical axis in two ways: (1) number of a person’s episodes that were judged highly creative by one or both of two
expert critics (hollow circles), and (2) number of episodes in which a person was judged by either or both of the two expert critics to have played
their role as producer/director/writer in a highly creative way (hollow squares). To be highly creative in multiple episodes, one has to work on
multiple episodes, so the table to the right contains Poisson regression models showing the strong creativity-network association after holding
constant the number of episodes on which a person worked, and the person’s number of episodes during a fallow period in the Dr. Who series
(coefficients presented with test statistics in parentheses).
 (Q130) We discussed creativity/innovation as an import-export
game that does not require genius. Which of the below is most
responsible for the truth of the statement?

A. A person can be lucky whether or not she is a genius.

B. Genius is a word the ignorant use to describe competent.

Network Brokerage and Competitive Advantage (page 43)

C. To be seen as creative, find people more ignorant than

yourself.

D. Idea value resides in the audience.

Strategic Leadership

E. Idea value resides in the inventor.

 (Q203) Closed networks do not identify A. True
unintelligent managers so much as they
identify specialists. True or false? B. False

(Q204) Innovation does not depend on

individual genius so much as it depends on A. True
employees finding opportunities to broker
knowledge from where it is routine to where
B. False
Network Brokerage and Competitive Advantage (page 44)

it would create value. True or false?

(Q262) Network brokers are more likely to

propose creative, innovative ideas that appeal
A. True
to top management. One reason for the
Strategic Leadership

appeal of broker proposals is the creative,

innovative language brokers use. True or B. False
false?
 (Q260) The below team of four people, A B C D, are working on
current project. They worked together on their previous project.
They worked together on the project before that. Which of the
below sentences best describes the current team?
Network Brokerage and Competitive Advantage (page 45)

A. The team is likely to produce innovative content.

B. The team is unlikely to produce innovative content.

Strategic Leadership

C. The team depends on executive sponsorship.

D. The team has a loose partner structure.

 Social Network EU and Emerging Asia
at the Top Markets
of the Company
Lines indicate frequent and
substantive work discussion;
heavy lines especially
close relationships.
Yanjie

US
RULE 2, Brokerage is Contingent:
Back Successful brokerage requires
(page 3)and Competitive Advantage (page 46)

Office B acceptance in the target audience

B
(acceptance can be measured by
Bob
job status, network status,
Jim
B reputation;
B
Granovetter, 1985,
B
1992; Tullock, 1985; Greif, 1989;
Ellickson, 1991; Bernstein, 1992,
B
B
2019; Uzzi, 1996, 1999; Burt,
1997, 1998; Kleinberg, 1999; Macy
Leadership
StrategicLeadership
Network Brokerage

CEO R&D
& Centola, 2007; Rider, 2009;
C-Suite
Front Tortoriello & Krackhardt, 2010;
Foundations

Heir Apparent
Office Burt & Merluzzi, 2016).
Strategic

Other Senior Person

Figure 1 in Burt, "Network disadvantaged entrepreneurs" (Entrepreneurship Theory & Practice, 2019)
 Access to bridge connections doesn’t
guarantee brokerage benefits. Returns are
contingent on factors not yet discussed.
Graph A below is from Brokerage & Closure and the previous handout showing
achievement increasing with more access to structural holes. Circles are z-score
residual achievement for 1,986 observations averaged within five-point intervals
of network constraint in each of six management populations (analysts, bankers,
and managers in Asia, Europe, and North America, see discussion of Figure 2.3 in
Chapter 2; heteroscedasticity is negligible, X2 = 2.97, 1 d.f., P ~ .08). Bold line is
the vertical axis predicted by network constraint.
Graph B to the right shows the raw data that were averaged to create B. But Vary Widely between
Graph A. Vertical axis is wider to accommodate more variable achievement. the Advantaged Individuals
Heteroscedasticity is high due to achievement differences between advantaged (overall r = -.24,
individuals (X2 = 269.5, 1 d.f., P < .001), but the association between achievement t = -9.98, n = 1,989)
and network advantage remains statistically significant when adjusted for
heteroscedasticity (Huber-White, t = -8.49).

A. Achievement Scores for

People in Open Networks Are
Z-Score Residual Achievement
(evaluation, compensation, promotion)
Network Brokerage and Competitive Advantage (page 47)

Higher than Peers on Average

(r = -.58, t = -6.78, n = 85)
Strategic Leadership

Network Constraint
many ——— Structural Holes ——— few
Figure adapted from Figure 1 in Burt (2012, "Network
Related Personality," American Journal of Sociology).
 When a broker proposes something new,
there is no guarantee that the proposal
will work in our market, with our company
processes, staffed by our people. There
is risk to accepting the proposal. Chains

Access to Structural Holes

(mean network constraint, lower
Network Advantage of
of command broken in service of company

scores indicate more access)

interests can just as easily be broken in
service of personal interests, or in service of
well-intentioned but strategy-eroding interests.
How will you be viewed in the target audience
as the proposal source?

Manager Age
For Example,
Network Brokerage and Competitive Advantage (page 48)

Are You the Right Age

To Be Accepted as the
Source of a Proposal?

t-test for network effect in Figure 1)

Returns to Advantage
The graphs plot averages across 2,206 senior managers in
(within-age estimates of
six organizations in electronics, finance, software, and supply
chain. The top graph shows the age at which people have
the most access to structural holes (more open networks at
the top).
Strategic Leadership

The bottom graph shows the age at which people have the
greatest returns to brokerage. Vertical axis is test statistic
for the strength of association between a manager's relative
achievement and his or her network constraint (calculate for
each age group the returns to brokerage graph).
Figure 4.2 in Burt, “Life course and network advantage”
(2019 Social Networks and the Life Course).
 Age is Not a General Caution; More a Function
of Company Culture: "Peak" Periods in Manager Life-Cycle
Access to Structural Holes
(mean network constraint, lower
scores indicate more access)
Network Advantage of
Network Brokerage and Competitive Advantage (page 49)

Shaded areas
HR in a Commercial Software Engineering enclose “peak”
Bank (n=283) in Electronics
(correlation between log network

years - ages in which

constraint and z-score residual

Supply Chain in Electronics (n=113)

Returns to Advantage

returns to network
achievement in Figure 1)

(n=455)
advantage are
similar to maximum.

Computers
(n=170)

Financial Services Investment

(n=654) Bank (n=531)
Strategic Leadership

Manager Age Manager Age Manager Age

(returns increase with age) (returns decrease with age) (returns increase & decrease)
Figure 3 in Burt (2018, "Life course and network advantage")
 r = -.58
t = -4.8
RULE 2: Broker Job Status
Reassures, or Lack of It Concerns,
the Target Audience
Which means the network around a senior
Structural Holes and Good Ideas
person is especially important for his or her
TABLE 1
achievement.
Predicting Performance
r = -.28
t = -2.5
1 2 3 4
Salary Graphs for executives,
Salary managers, Promotion
Evaluation
and junior managers to the right
Manager 1 . . . . . . . . . . . . 31,099** (2,882) 35,707** (3,498)compensation
show z-score .973 (.678) .689 (.670)
Manager 2 . . . . . . . . . . . . 16,652** (2,745) 19,892** (3,479) .863 (.631) 1.165 (.648)
relative to peers (controlling for
Manager 3
Network Brokerage and Competitive Advantage (page 50)

(reference) . . . . . . . . . . . . . background
. . . differences)
. . . across . . .
Sr. manager . . . . . . . . . . . 19,638** (3,782) levels of network
15,484** (4,143) constraint.
.116 Not .635 (.885)
(.843)
Executive . . . . . . . . . . . . . 65,394** (4,522) only do more
61,930** (4,835)senior.423
people(1.01)
have .221 (1.08)
Purchasing . . . . . . . . . . . . 754 (1,351) more
1,811 open networks.410
(1,884) (on average),
(.313) .478 (.345)
Age . . . . . . . . . . . . . . . . . . . 338** (52) they
300**earn higher
(71) returns
.085**to(.013)
having .084** (.013)
Bachelor . . . . . . . . . . . . . . 1,610 (1,003) 200 networks
open (1,401) (also
.211 (.237)if
pay more .118 (.240)
Graduate . . . . . . . . . . . . . . 734 (864) 451 (1,155) .208 (.203)
they don't have an open network). .182 (.204)
Hightech . . . . . . . . . . . . . . 3,516** (880) 3,150* (1,189) .087 (.209) .162 (.210)
Lowtech . . . . . . . . . . . . . . 6,927** (1,481) 6,607* (2,375) .351 (.342) .409 (.378) r = -.02
Urban 1 . . . . . . . . . . . . . . . 3,613** (1,046) Table to the
3,947** left is from
(1,456) .423 page 371 .152
(.247) (.252) t = -0.2
Urban 2 . . . . . . . . . . . . . . . 5,049** (1,010) of Burt, "Structural
5,585* (1,427) .564 holes and good.052
(.238) (.243)
Network constraint . . . . 7 (25) ideas"
1 (2004, (38) American
.014** Journal
(.004) of .022** (.006)
Sociology).
Strategic Leadership

Mgr2 # constraint . . . . 19 (35) 47 (58) .004 (.008) .008 (.009)
Mgr3 # constraint . . . . 47 (38) 159* (59) .007 (.009) .003 (.009)
SrMgr # See pp. 156-162 and Figure 3.8 in
constraint . . . . . . . . . . . (75)
214*
Brokerage(84)
216* .005
and Closure for(.017)
general .010 (.019)
Executive #
constraint . . . . . . . . . . . 681** (124) discussion(132)
697** showing the form
.011 of
(.028) .024 (.030)
N ..................... 673 contingency
398 functions. 673 638
Note.—Coefficients in models 1 and 2 are change in salary dollars with a unit increase in row variable
(respectively .80 and .83 squared multiple correlations; network effect plotted in fig. 4). Coefficients in
model 3 predict three levels of evaluation for an ordinal logit model (114.8 x2 with 17 df; network effects
are plotted in fig. 4 holding age constant). Coefficients in model 4 are for a logit model predicting whether
 RULE 2: Broker Network Status
Reassures or Concerns the

(eigenvector score / mean score)

Target Audience

Network Status
Network status is on the vertical axis of the top graph. Status
is defined in the same way that price is defined in the general
equilibrium model: Si = Sj zji Sj, where Si is status of person i, and
zji is connection from j to i. Like price, status is only meaningful
in reference to the status of some numeraire benchmark person.
Here, status is normalized at the mean, so a score of 1.0 indicates a
person of average status in the network.
R2 = .74
Si = Sj zji Sj
Regions indexed
by shading, functions
Network Constraint
by shape
Network Brokerage and Competitive Advantage (page 51)

Sales
Regional Ops
Product Support
Administration r = -.96 for
high status

Z-Score Compensation
r = .03 for
low status
Strategic Leadership

Network Constraint
Sociogram is Figure 3.2 in Neighbor Networks and the graphs are from Figures 1 and 2 in Burt & Merluzzi discussion of the link between
brokerage and network status as a reputation measures (2013, "Embedded brokerage," Research in the Sociology of Organizations)
 Returns to
Brokerage High Status (top 50%)
n = 21

Predicted Avatar Z-Score Achievement

Contingent on
Z-Score Achievement

r = -.92
t = -10.35
High Status (top 50%)
Network Statusb = .57 forn =x,25-.01 for x 2

Can Also Be t = 30.12 for x, -16.86 for x 2

Found in Online
Networks
Avatar
52)

Other Status (bottom 50%)

n = 14 Other Status
(page

b = .28 for x, -.01 for x2 (bottom 50%)

Predicted

Results here are from networks and

t = 3.90 for x, -2.44 for x2 n = 19
Network Brokerage and Competitive Advantage

achievement in the virtual world of r = -0.21

Second Life. Achievement is the t = -0.88
canonical correlation dependent variable
in Model 15, Table S5. “High” status is
above median. Scores on the y-axis are 24+
average achievement scores for avatars
NonRedundant
within integer Contacts
intervals of nonredundant Network Constraint (x 100)
few ——— Structural Holes ——— many many ——— Structural Holes ——— few
contacts (left) or five-point intervals of
network constraint (right). Statistics are
Strategic Leadership

based on averages plotted in the graph.

See Table S8 for regression results
Brokerage-Achievement Association,
with controls showing higher returns to
brokerage for individual avatars with high
status.Contingent on Network Status in Second Life
Achievement is the canonical correlation dependent variable in Model 15, Table S5. “High” status fromis above median.
Burt (2020, Structural Scores on the
Holes in Virtual Worlds).
y-axis are average achievement scores for avatars within integer intervals of nonredundant contacts (left) or five-point intervals of
network constraint (right). Statistics are based on averages plotted in the graph. See Table S8 for regression results with controls
showing higher returns to brokerage for individual avatars with high status.
 Reputation Can Substitute for Status,
making reputation valuable as the key to being accepted as a broker.
Graph plots investment banker reputation by levels of network status. Reputation is measured by average colleague evaluation. Boxes
span 25% to 75% with bold horizontal at the mean. Whiskers extend down to minimum reputation, up to maximum.

Reputation Is
Correlated with

(mean colleague evaluation)

Status, but Is

Banker Reputation
Distinct
Network Brokerage and Competitive Advantage (page 53)

High Status Is a Good Signal

of Positive Reputation.

Low Status Is an
Ambiguous Signal

GENERIC DEFINITION: "Differences in detail aside,

most social scientists agree upon two aspects of
reputation: first, knowing a business partner's past
Strategic Leadership

behavior mitigates uncertainty about his future

performance; second, reputation demonstrates the
person's credibility as an honest business partner Network Status
and reduces the uncertainty associated with trusting (eigenvector score / mean score)
him." (Hillmann and Aven, 2011, AJS, page 485)

From Burt (2020, Structural Holes in Virtual Worlds). The boutique investment bank, Moelis — "Best Global Independent Investment Bank"
in 2010 and "Most Innovative Boutique of the Year" in 2011 — nicely illustrates the competitive advantage of reputation as an entree to
brokerage opportunities (download free Moelis case from www.sbs.oxford.edu/reputation/cases).
 RULE 2: Broker Advantage Is Contingent on Reputation
Graph plots relative banker compensation
across levels of constraint in the banker's
discussion network. Compensation is R2 = .73
averaged within intervals of network (t = -3.31, n = 226)
constraint, but the test statistic is for all 469

Z-Score Compensation
observations, holding constant job rank,
peer evaluation, years with the organization, Top 50% of reputations

(total annual)
minority, and working in US headquarters Bottom 50%
(Burt, Neighbor Networks 2010:91-93).
There are two predictions: one for
bankers with above-average reputations
(solid squares), the other for bankers with
below-average reputations (hollow squares).
R2 = .21
Network Brokerage and Competitive Advantage (page 54)

Network status is added to each prediction

(t = -0.42, n = 243)
as a control for a banker's social standing
across all senior people in the bank.
As Rider (ASQ 2009:578-579) explains
for placement agents: “a broker’s reputation for
consistently representing actors of high quality is a
valuable, intangible asset that enables a broker to
realize future rents on the brokerage position. . . If a
positive reputation reduces the costs of assuaging Network Constraint (C)
potential exchange partners’ concerns, then the many ——— Structural Holes ——— few
returns to brokerage should be positively related to
Strategic Leadership

a broker’s reputation.” Similarly, Nee and Opper (Capitalism from Below 2012: 211) describe Chinese entrepreneurs building
reputation in the course of brokering connections: “Through personal introductions and fine-grained information passed
through social networks, the ‘broker’ typically signals trustworthiness and reputation of the prospective business partners.
Moreover, it is in the broker’s interest to make good recommendations, as most business partners will tend to reward their
networking contacts in one way or another. Such introductions can span the social gaps, or ‘structural holes’ between
groups.”
For discussion, see Burt & Merluzzi (2014, “embedded brokerage”). The boutique investment bank, Moelis — "Best Global Independent Investment Bank"
in 2010 and "Most Innovative Boutique of the Year" in 2011 — nicely illustrates the competitive advantage of reputation as entree to brokerage opportunities
(download free Moelis case from www.sbs.oxford.edu/reputation/cases).
 Most important,
44%
reputation enables a
wider population of
people to be brokers.
Number of Bankers

Relative to
17% job rank and
network status,
reputation opens
organizations
5% and markets
Network Brokerage and Competitive Advantage (page 55)

to the largest
number of
Senior Network Reputation people with
Job Rank Status 200 bankers eligible good ideas.
by their reputation
25 bankers 80 bankers
eligible by their eligible by their -0.6

senior job rank network status

Compensation and Log Network Constraint

About 200 cases have reputation positive

Average Subsample Correlation between

enough to benefit from brokerage
-0.4

Value of Status or Reputation

-0.4
Strategic Leadership

-0.2

-0.2
Horizontal axis ranks banker observations from highest status (hollow dots) or About 80 cases
have enough
most-positive reputation (solid dots) to the opposite extreme. Vertical axis is the -0.0
status to benefit
from brokerage

correlation between compensation and log network constraint for a sample of

0.0
observations adjacent to each banker (24 of higher social standing plus 24 of lower). 0 50 100 150 200 250 300 350 400 450

Displayed data are smoothed by averaging across 24 adjacent observations. Rank Order of Bankers from First to Last in Social Standing
(hollow dots for network status, solid dots for reputation)

From Burt (2020, Structural Holes in Virtual Worlds).

 (Q127) The higher compensation for network brokers displayed in
the graph below is typically a spurious network effect since people
in higher job ranks receive higher compensation and are more
likely to be network brokers. True or false?

2.5

A. True

(compensation, evaluation, promotion)

Z-Score Relative Performance
1.5

B. False
Network Brokerage and Competitive Advantage (page 56)

0.5

0.0

-0.5

-1.5
Strategic Leadership

Z = 2.78 - .82 ln(C)

-2.5 r = -.53
5 15 25 35 45 55 65 75 85 95

Network Constraint (C)

many ——— Structural Holes ——— few
 Z-Score Residual Achievement
(evaluation, compensation, promotion)
Network Brokerage and Competitive Advantage (page 57)

Network Constraint
many ——— Structural Holes ——— few

(Q158) The graph to the left above shows average returns to brokerage. The graph
to the right shows individual returns before they are averaged. What do the two
graphs tell you about the returns one can expect to earn as a network broker?

C. Network advantage does not

A. You can earn more as an
Strategic Leadership

guarantee high performance.

individual.
D. Average returns are heteroskedastic.
B. Average returns are high.
 (Q161) The graph below displays returns to brokerage for two
categories of bankers in a large financial organization. From the
graph, you can infer that:

A. Broker access to
structural holes is The bold line is for people with
r = -.86 clearly positive reputation.
valuable.

Z-Score Compensation
Dashed line is for other people.

B. Broker status in the

Network Brokerage and Competitive Advantage (page 58)

(total annual)
informal organization is
valuable.

C. Broker reputation is r = -.28

valuable.
Strategic Leadership

D. Broker compensation
increases with job rank. Network Constraint (C)
many ——— Structural Holes ——— few
 (Q206) After age 50, your opportunities to
broker across structural holes decrease but
returns to brokerage increase. True or false?
Network Brokerage and Competitive Advantage (page 59)

A. True

B. False
Strategic Leadership
 (Q195) We discussed network brokerage facilitated by a broker’s
job rank, status, and reputation. As contingency factors for
successful brokerage the three variables are most similar with
respect to:

A. Authority of eligible brokers

Network Brokerage and Competitive Advantage (page 60)

B. Visibility of eligible brokers

C. Broker eligibility to broker

D. Cost of the broker’s proposal

Strategic Leadership

E. Number of eligible brokers

 _________________________________
Name, Section

HandIn Assignment #3:

Network Broker
(One page due by midnight Saturday of the week when we finish discussing this handout.)

Drawing on our discussion of this handout, describe an event in your experience

during which you believe you acted as a network broker.

What was the structural hole you brokered?

Network Brokerage and Competitive Advantage (page 61)

How did you become aware of the hole?

Why was the hole there?

How did you go about brokering across the hole? Did you have to
overcome resistence from either side of the hole?
Strategic Leadership

What was the outcome?

 Three Summary Points
Network Structure Is a Proxy for the Distribution of Information
For reasons of opportunity, shared interests, experience — simple inertia — organizations and
markets drift toward the bridge-and-cluster structure known as a “small world.”

RULE 1 of Social Capital: Brokers Do Better

Bridge relations across the structural holes between clusters provide information breadth, timing, and
arbitrage advantages, such that network brokers managing the bridges are at higher risk of “productive
accident” in detecting and developing good ideas. By clearing the sticky-information market across
organizations, brokers tend to be recognized leaders, better compensated than peers, more widely
celebrated than peers, and promoted more quickly than peers. Some specifics:
- Closed networks do not identify unintelligent managers so much as expert specialists.
- Creativity and innovation are an import/export process. Value is not at the innovation source. It
Network Brokerage and Competitive Advantage (page 62)

emerges each time productive new knowledge is adopted in a target audience.

- Creativity and innovation depend on the network as well as the inventive person. It does not
depend on individual genius so much as it depends on employees finding opportunities to broker
knowledge from where it is routine to where it would create value.

RULE 2 of Social Capital: Contingent Returns to Brokerage

To the extent that a broker is proposing something new, there is no guarantee that the proposal will
work in our market, with our company processes, staffed by our people. There is risk to accepting the
proposal. Chains of command broken in service of company interests can just as easily be broken
Strategic Leadership

in service of personal interests, or in service of well-intentioned but strategy-eroding interests. Social

standing in the form of job rank, network status, or reputation is the way would-be brokers overcome
the suspicions with which brokers can be viewed. Reputation is particularly valuable. It legitimates the
largest number of people to be successful network brokers.
These are core rules for Leonardi and Contractor's
2018 HBR piece, "Better people analytics."
Strategic Leadership
Network Brokerage and Competitive Advantage (page 63)

Materials
Appendix
 Appendix I:

Example
Network
Questionnaire
for a
Web Survey
Network Brokerage and Competitive Advantage (page 64)

for discussion
of these slides and
how to collect
network data,
see Appendix A,
"Measuring the
Network," in
Neighbor Networks.
Strategic Leadership

For a similar offline

exercise, see Hermi
Ibarra's 2008 network
exercise on the HBSP
website (item 9-497-993).

Figure A1 in Neighbor Networks

 Strategic Leadership
Network Brokerage and Competitive Advantage (page 65)

continued
Appendix I,

Figure A2 in Neighbor Networks

 Confidential
Figure A2. Business Time Line Worksheet
Business Event
Time Line for an Example Firm
Name Generator
1993, secured technology partner 2008, secured current
primary export customer
The next five questions generate a summary 1999, first bank loan
picture of the business network. To draw the
picture, you will be asked about people, but we business | | | today
do not want to know any one's name. I will go founded 1997
_____ 2002
_____ 2007
_____ 2012
through this network worksheet with you, asking 1992
_____
about people who were useful to your business 2000, critical supplier
no longer available 2004, first export contract
in one way or another. Without mentioning
anyone's name to me, please write on your
worksheet the names of people who come to
mind in response to the questions. We will
create a list of names then refer to people by
Network Brokerage and Competitive Advantage (page 66)

their order on the list. No names. You will keep Time Line for Your Firm
the worksheet to yourself.

Q1. Let me begin with an example so you can

see how the interview protects your
confidentiality at the same time that a picture of
the business network emerges. Your business business | | | today
time line shows that your firm was founded in founded _____ _____ _____ 2012
_____
_(say founding year)_. Please think back to
your activities in founding the firm. Who
was the one person who was most valuable
to you in founding the firm?
Strategic Leadership

Q2. Now please do the same thing for each of the significant events you listed on your business time line. The first significant
event you listed was __(say first event)__ in _(say year)_. Who was the person most valuable to you during that event?
Please write on the first line below the person's name. The person most valuable in this event could be the same person who was
most valuable to you in founding the firm. You would just enter the name again.

from Burt and Burzynska, "Chinese entrepreneurs, social networks, and

guanxi" (2018 Management and Organization Review)
 Figure A3. Name Interpreters Flesh Out Relationships
and Define Connections among Cited Contacts
§  Contact Gender (male, female)
§  Emotional Closeness to Contact (especially close, close, less close, distant)
§  Duration of Connection with Contact (years known)
§  Frequency of Contact (daily, weekly, monthly, less often)
§  Trust (1 to 5, low to high trust) “Consider the extent to which you trust each of the listed people.
For example, suppose one of the people asked for your help. The help is not extreme, but it is
substantial. It is a level of help you cannot offer to many people. To what extent would you trust each
Network Brokerage and Competitive Advantage (page 67)

person to give you all the information you need to decide on the help? For example, if the person was
asking for a loan, would they fully inform you about the risks of them being able to repay the loan? If the
person was asking you give a job to one of their relatives, would they fully inform you about their relative's
poor work attitude or weak abilities, or other qualities that would make you prefer not to hire the relative?”

§  Role (all that apply: family, extended family,

neighbor, party, childhood, classmate, military,
colleague, business association)

§  Matrix of Connections between Contacts

Strategic Leadership

(especially close, distant,

or something in between)

from Burt and Burzynska, "Chinese entrepreneurs, social networks, and

guanxi" (2017 Management and Organization Review)
 Appendix II: Measuring Access to Structural Holes*
from Burt, "Formalizing the argument," (1992, Structural Holes); "Gender of social capital" (1998, Rationality and Society); Appendix B "Measuring Access to Structural
Holes," (2010, Neighbor Networks). See the Pfeffer Stanford case #OB-66 for a productive overview ("A note on networks and network structure").

Network brokerage is typically measured in terms of opportunities to connect people. When everyone you know is connected
with one another, you have no opportunities to connect people. When you know a lot of people disconnected from one another,
then you have a lot of opportunities to connect people. “Opportunities” should be emphasized in these sentences. None of
the usual brokerage measures actually measures brokerage behavior. They index opportunities for brokerage. Reliability and
cost underlie the practice of measuring brokerage in terms of opportunities. It is difficult to know whether or not you acted on a
brokerage opportunity. One can know with more reliability whether or not you had an opportunity for brokerage. Acts of brokerage
could be studied with ethnographic data, but the needed depth of data would be expensive, if not impossible, to obtain by the
practical survey methods used to measure networks.
Good reasons notwithstanding, the practice of measuring brokerage by its opportunities rather than its occurrence means
that performance has uneven variance across levels of brokerage opportunities. Performance is typically low in the absence
of opportunities. Performance varies widely where there are many opportunities: (1) because some people with opportunities
do not act upon them and so show no performance benefit, (2) because it is not always valuable to move information between
disconnected people (e.g., explain to your grandmother the latest technology in your line of work), or (3) because the performance
Network Brokerage and Competitive Advantage (page 68)

benefit of brokerage can occur with just one key bridge relationship. A sociologist might do more creative work because of
working through an idea with a colleague from economics, but that does not mean that she would be three times more creative
if she also worked through the idea with a colleague from psychology, another from anthropology, and another from history. The
above three points can be true of brokerage measured in terms of action, but under the assumption that people invest less in
brokerage that adds no value, the three points are more obviously true of brokerage measured in terms of opportunities. It could
be argued that people more often involved in bridge relations are more likely to have one bridge that is valuable for brokerage,
and to understand how to use bridges to add value, but the point remains that the network measures discussed below index
opportunities for brokerage, not acts of brokerage.

Bridge Counts
Bridge counts are an intuitively appealing measure. The relation between two people is a bridge if there are no indirect connections
Strategic Leadership

between the two people through mutual contacts. Associations with performance have been reported measuring brokerage
with a count of bridges (e.g., Burt, Hogarth, and Michaud, 2000:Appendix; Burt, 2002).

Constraint
I measure brokerage opportunities with a summary index, network constraint. As illustrated on the next page, network constraint
begins with the extent to which manager i’s network is directly or indirectly invested in the manager’s relationship with contact j
(Burt 1992: Chap. 2): cij = (pij + Σqpiqpqj)2, for q ≠ i,j, where pij is the proportion of i’s network time and energy invested in contact
 Illustrative A

Network and
Computation B

Constraint C

measures the
extent to which a E
network doesn't D

span structural
Network Brokerage and Competitive Advantage (page 69)

Network constraint measures the extent to which your network time and energy
is concentrated in a single group. There are two components: (direct) a contact
holes consumes a large proportion of your network time and energy, and (indirect) a
contact controls other people who consume a large proportion of your network
time and energy. The proportion of i’s network time and energy allocated to j, pij,
is the ratio of zij to the sum of i’s relations, where zij is the strength of connection
between i and j, here simplified to zero versus one.

cij = (pij + Σq piqpqj)2 q ≠ i,j

contact-specific network data
constraint (x100):
A . 1 0 0 1 1 1
Strategic Leadership

A 15.1 B 1 . 0 1 0 0 1
B 8.5 C 0 0 . 0 0 0 1
C 2.8 D 0 1 0 . 0 0 1
D 4.9 100(1/36) E 1 0 0 0 . 0 1
Figure 2.2 in Structural Holes. E 4.3 F 1 0 0 0 0 . 1
F 4.3 gray dot 1 1 1 1 1 1 .

total 39.9 = aggregate constraint (C = Σj cij)

 j, pij = zij / Σqziq, and variable zij measures the strength of connection between contacts i and j. Connection zij measures the lack
of a structural hole so it is made symmetric before computing pij in that a hole between i and j is unlikely to the extent that either
i or j feels that they spend a lot of time in the relationship (strength of connection “between” i and j versus strength of connection
“from” i to j; see Burt, 1992:51). The total in parentheses is the proportion of i’s relations that are directly or indirectly invested
in connection with contact j. The sum of squared proportions, Σjcij, is the network constraint index C. I multiply scores by 100
to discuss integer levels of constraint.
The network constraint index varies with three network dimensions: size, density, and hierarchy. Constraint on a person
is high if the person has few contacts (small network) and those contacts are strongly connected to one another, either directly
(as in a dense network), or through a central, mutual contact (as in a hierarchical network). The index, C, can be written as
the sum of three variables: Σj(pij)2 +2Σjpij(Σqpiqpqj) + Σj(Σqpiqpqj)2. The first term in the expression, C-size in Burt (1998), is a
Herfindahl index measuring the extent to which manager i’s relations are concentrated in a single contact. The second term,
C-density in Burt (1998), is an interaction between strong ties and density in the sense that it increases with the extent to which
manager i’s strongest relations are with contacts strongly tied to the other contacts. The third term, C-hierarchy in Burt (1998),
measures the extent to which manager i’s contacts concentrate their relations in one central contact. See Burt (1992:50ff.;
1998:Appendix) and Borgatti, Jones, and Everett (1998) for discussion of components in network constraint.

Size
Network Brokerage and Competitive Advantage (page 70)

Network size, N, is the number of contacts in a person's network. In graph-theory discussions, the size of the network around
a person is discussed as “degree.” For non-zero network size, other things equal, more contacts mean that a manager is more
likely to receive diverse bits of information from contacts and is more able to play their individual demands against one another.
Network constraint is lower in larger networks because the proportion of a manager’s network time and energy allocated to any
one contact (pij in the constraint equation) decreases on average as the number of contacts increases.

Density
Density is the average strength of connection between contacts: Σ zij / N*(N-1), where summation is across all contacts i and
j. Dense networks are more constraining since contacts are more connected (Σqpiqpqj in the constraint equation). Contact
connections increase the probability that the contacts know the same information and eliminate opportunities to broker information
between contacts. Thus, dense networks offer less of the information and control advantage associated with spanning structural
holes. Density is only one form of network closure, but it is a form often discussed as closure.
Strategic Leadership

Hypothetical networks in the figure below illustrate how constraint varies with size, density, and hierarchy. Relations are
simplified to binary and symmetric in the networks. The graphs display relations between contacts. Relations with the person
at the center of the network are not presented (that person at the center is referenced by various labels such as "you," "ego,"
or "respondent"). The first column in the figure contains examples of sparse networks (zero density). No contact is connected
with other contacts. The third column of the figure contains maximum-density networks (density = 100). Every contact has a
strong connection with each other contact. At each network size, constraint is lower in the sparse-network column.
 Hierarchy
Density is a form of closure in which contacts are equally connected. Hierarchy is another form of closure in which a minority of
contacts, typically one or two, stand above the others for being more the source of closure. The extreme is to have a network
organized around one contact. For people in job transition, such as M.B.A. students, that one contact is often the spouse. In
organizations, hierarchical networks are sometimes built around the boss.
Hierarchy and density both increase constraint, but in different ways. They enlarge the indirect connection component
in network constraint (Σqpiqpqj). Where network constraint measures the extent to which contacts are redundant, network
hierarchy measures the extent to which the redundancy can be traced to a single contact in the network. The central contact
in a hierarchical network gets the same information available to the manager and cannot be avoided in manager negotiations
with each other contact. More, the central contact can be played against the manager by third parties because information
available from the manager is equally available from the central contact since manager and central contact reach the same
people. Network constraint increases with both density and hierarchy, but density and hierarchy are empirically distinct measures
and fundamentally distinct with respect to social capital because it is hierarchy that measures social capital borrowed from a
sponsor.
To measure the extent to which the constraint on a person is concentrated in certain contacts, I use the Coleman-Theil
inequality index for its attractive qualities as a robust measure of hierarchy (Burt, 1992:70ff.). Applied to contact-specific constraint
scores, the index is the ratio of Σj rj ln(rj) divided by N ln(N), where N is number of contacts, rj is the ratio of contact-j constraint
Network Brokerage and Competitive Advantage (page 71)

over average constraint, cij/(C/N). The ratio equals zero if all contact-specific constraints equal the average, and approaches
1.0 to the extent that all constraint is from one contact. Again, I multiply scores by 100 and report integer values.
In the first and third columns on the next page, no one contact is more connected than others, so all of the hierarchy scores
are zero. Non-zero hierarchy scores occur in the middle column, where one central contact is connected to all others who are
otherwise disconnected from one another. Contact A poses more severe constraint than the others because network ties are
concentrated in A. The Coleman-Theil index increases with the number of people connected to the central contact. Hierarchy
is 7 for the three-contact hierarchical network, 25 for the five-contact network, and 50 for the ten-contact network. This feature
of hierarchy increasing with the number of people in the hierarchy turns out to be important for measuring the social capital of
outsiders because it measures the volume of social capital borrowed from a sponsor, which strengthens the association with
performance (this point is the focus of the later session on outsiders having to borrow network access from a strategic partner).
Note that constraint increases with hierarchy and density such that evidence of density correlated with performance can
Strategic Leadership

be evidence of a hierarchy effect. Constraint is high in the dense and hierarchical three-contact networks (93 and 84 points
respectively). Constraint is 65 in the dense five-contact network, and 59 in the hierarchical network; even though density is
only 40 in the hierarchical network. In the ten-contact networks, constraint is lower in the dense network than the hierarchical
network (36 versus 41), and density is only 20 in the hierarchical network. Density and hierarchy are correlated, but distinct,
components in network constraint.
 Broker Partner Clique Partners
Networks Networks Networks

Network Hierarchy
A A A

Small
Networks C B C D
B C B

contacts 3 3 3
density x 100 0 67 100
hierarchy x 100 0 7 0
constraint x 100 33 84 93
from: Cliques
A 11 44 31 ok

Br
B 11 20 31 e rs
C 11 20 31
nonredundant contacts 3.0 1.7 1.0
betweenness (holes) 3.0 0.5 0.0 Network Density

A A A

E B E B E B
Network Constraint
Network Brokerage and Competitive Advantage (page 72)

Larger
Networks D C D C D C decreases with number of contacts
contacts 5 5 5 (size), increases with strength of
density x 100 0 40 100
hierarchy x 100 0 25 0 connections between contacts
59 65
constraint x 100
from:
20
(density), and increases with sharing
A
B
4
4
36
6
13
13
the network (hierarchy).
C 4 6 13
D 4 6 13
E 4 6 13 This is Figure 1 in Burt, "Reinforced Structural Holes,"
nonredundant contacts 5.0 3.4 1.0
betweenness (holes) 10.0 3.0 0.0 (2015, Social Networks, an elaboration of Figure B.2
in Neighbor Networks). Graph above plots density
Strategic Leadership

Still Larger and hierarchy for 1,989 networks observed in six

Networks management populations (aggregated in Figure
contacts 10 10 10 2.4 in Neighbor Networks to illustrate returns to
density x 100 0 20 100 brokerage). Dot-circles are executives (MD or more
hierarchy x 100 0 50 0
constraint x 100 10 41 36 in finance, VP or more otherwise). Hollow circles are
nonredundant contacts 10.0 8.2 1.0 lower ranks. Executives have significantly larger,
betweenness (holes) 45.0 18.0 0.0
less dense, and less hierarchical networks.
To keep the diagrams simple, relations with ego are not presented.
 R2 = .99

(Number Monopoly-Access Holes)

Ego-Network Betweenness
(Number Monopoly-Access Holes)
Ego-Network Betweenness
R2 = .92
NonRedundant Contacts

-.90 correlation -.71 correlation

with log constraint with log constraint
Network Brokerage and Competitive Advantage (page 73)

Network Constraint (x 100) Network Constraint (x 100) NonRedundant Contacts

many ——— Structural Holes ——— few many ——— Structural Holes ——— few few ——— Structural Holes ——— many

The Network Measures of Access to Structural Holes

Are Strongly Correlated
Strategic Leadership

These are network metrics for 801 senior people in two organizations analyzed in Burt, "Reinforced structural
holes" (2015, Social Networks). One organization is a center-periphery network of investment bankers (circles).
The other is a balkanized network of supply-chain managers in a large electronics company (squares). The
point is that networks rich in structural holes by one measure tend to be rich in the other measures.
 Structural Folds Indicate Access to Structural Holes
3 7 11 14

4 2 6 8 12 10 15

5 9 13 16

Ego-Network Reinforced Holes

Network Effective Size Ego-Network
Network Betweenness (RSH)
Kind of Network Size (NonRedundant Modularity
Constraint (Structural
Network Brokerage and Competitive Advantage (page 74)

(Contacts) Contacts) (Newman Q)

Holes)
Raw Normalized
Closed (3, 4, 5, 7,
8, 9, 11, 12, 13, 3 1.0 92.6 .00 .00 0% .00
14, 15, 16)

Broker (1) 2 2.0 50.0 1.00 .75 75% .00

Broker (2, 6) 4 2.5 58.3 3.00 1.75 29% .00

Fold Broker (10) 6 4.0 46.3 9.00 6.00 40% .50

Strategic Leadership

This is Figure 3 in Burt, "Reinforced structural holes" (2015, Social Networks), Log Constraint 1.00
based on the above networks in Figure 1 of Vedres and Stark, "Structural
Effective Size -.90 1.00
folds: generative disruption in overlapping groups" (2010, American Journal of
Sociology). Correlations to the right are across the 801 bankers and managers EN Betweenness -.71 .88 1.00
analyzed in the 2015 article.
RSH -.71 .93 .91
 Appendix III: Competitive Advantage in
Social Networks and Stigler’s “Economics
of Information,” JPE 1961*
"The expected saving from given search will be greater, the
greater the dispersion of prices.” When price varies greatly
between sellers, it is worth a buyer’s time to search for the
lowest price. It makes no sense to search for the lowest price George Stigler,
of a commodity good; all prices are similar. 1960

The potential value of search is an incentive for entrepreneurs to aggregate price

Network Brokerage and Competitive Advantage (page 75)

information by enforcing localized transactions, as in medieval markets, or by

becoming "specialized traders whose chief service, indeed, is implicitly to provide a
meeting place for potential buyers and sellers.”

In short, the value of search is proportional to information variation, and search is

more productive for people more exposed to the variation.

As referenced in Stigler’s 1982 Nobel acceptance speech: "The proposal to study

the economics of information was promptly and widely accepted, and without even a
Strategic Leadership

respectable minimum of controversy." "All I had done was to open a door to a room
that contained many fascinating and important problems."
*Discussed in Burt and Soda, "The social origins of great strategies" (Strategy Science, 2017). Photo is from University
of Chicago Photographic Archive [apf1-07960], Special Collections Research Center, University of Chicago Library.