Social Exchange

1
Social Exchange

• Background
• Theoretical Roots
• Classical Anthropology Malinowski (Kula
  exchange), Levi-Strauss (Classificatory Kinship
  studies), Durkheim
• Homans, Blau, Emerson, Ekeh
• Direct Exchange
• Emerson, Cook, Yamagishi
• -Power Dependence, Vulnerability
• Willer, Markovski, Skvoretz, Lovaglia
• - Network Exchange Theory (NET)
• Game Theory
• - Bienenstock Bonacich
• Central questions about power
• Generalized Exchange
• Bearman
• Nobuyuki Takahashi
• Central question is about solidarity Social
  cohesion

2
Social Exchange Direct
Theoretical Background
Peter Blaus Exchange and Power in Social Life
and Homans Elementary Forms are central starting
points for much of this work

• Blau
• Most of social life rests on interaction
• Interaction is rarely a purely disinterested
  affair
• People seek something from interaction and give
  something in turn
• Blau focused on
• what distinguished economic from social exchange
• the forces propelling reciprocity
• the importance of ambiguity in social exchange
• Social debt
• How power comes from controlling resources

3
Social Exchange Direct
Theoretical Background
The view of actors is individualistic rational.
With Homans this rested mainly on a
Skinner-esque behaviorism and basic rational
actor models. Blau dropped most of the
behaviorism issues (though different
psychological issues were taken up by Emerson
others) in favor of a simplified economic
view. Blau argues that, An apparent
altruism pervades social life people are
anxious to benefit one another and to reciprocate
for the benefits they receive. But beneath this
seeming selflessness an underlying egoism can
be discovered the tendency to help others is
frequently motivated by the expectation that
doing so will bring social rewards.
4
Social Exchange Direct
Theoretical Background
From this, Blau develops a theory of exchange and
power. We exchange with others for the things we
cant get ourselves (favors from colleagues,
romantic interest from people we are attracted
to, skills). Those that control these resources
have power, since others are willing to provide
something for them. Social exchange differs
from economic exchange in the extent of ambiguity
underlying the exchange (how much is a half hour
of a colleague's time worth? Or an hour of the
attention of someone you are attracted to?),
which has multiple implications for the dynamics
of social interaction
5
Social Exchange Direct
Theoretical Background
People have extended (or, in many cases, reacted
against) Blaus propositions in multiple ways.
Most versions of network exchange theory starts
with Blaus assertion that power follows from the
control of resources. That unilateral control
leads to power in a dyad is one thing, but how
does control over exchange differ when we move
beyond the dyad to larger exchange structures?
This is the branch of ideas that lead to Network
Exchange Theory and Power Dependence Theory The
focus is less on the ambiguity or uniquely social
aspect of the exchange event, but rather on how
any exchange relation is affected by the social
structure that restricts exchange partners.
6
Social Exchange Direct
Cook, Emerson, Gilmore and Yamagishi
A relatively early paper in a long sequence of
work (see handout). This piece sets up one
branch the exchange theory. For a great review of
the other dominant branch, see Network Exchange
Theory (Willer, 1999)
A very active research community, interested in
identifying how the structure of a network can
give particular members of the network greater
control over resources. In most cases, the work
is experimental and formal, moving very carefully
along theoretically defined lines, and testing
each step with experiments.
7
Social Exchange Direct
Basic Concepts
Many of the social networks of interest to
social scientists can be analyzed fruitfully as
exchange networks, provided that the specific
content of the social relations in the network
involves the transfer of valued items Consists
of 1) a set of actors 2) a distribution of
valued resources among those actors 3) for each
actor a set of exchange opportunities 4) a set
of historically used exchange relations (subset
of 4) 5) a set of network connections linking
exchange relations into a single network
structure
8
Social Exchange Direct
Basic Concepts

• Definition 1 Two exchange relations between
  actors A-B and A-C are connected to form the
  minimal network B-A-C to the degree that exchange
  in one relation is contingent on exchange (or
  nonexchange) in the other relation.
• The connection is positive if exchange in one
  relation is contingent on exchange in another
  relation
• The connection is negative if exchange in one
  relation is contingent on nonexchange in the
  other.

9
Social Exchange Direct
Basic Concepts

• Definition 2 A position in a graph or network is
  a set of one or more points whose residual graphs
  are isomorphic (I.e. automorphic equivalence)
• Used to simplify the analysis of otherwise more
  complex networks
• Position in the network determines exchange
  behavior

10
Social Exchange Direct
(Used in previous experiments)
(Used in the experiment)
11
Social Exchange Direct
Basic Concepts

• In these networks,
• Each actor has a resource which the other actors
  want
• each line represents an opportunity for exchange
• Solid lines represent a more profitable exchange
  opportunity than dashed lines.
• They expect that the high profit opportunities
  will be converted into relations
• The emergent networks are negatively connected
  any use of one opportunity means that another is
  forgone
• Actors have no knowledge of the structure beyond
  their own set of relations

12
Social Exchange Direct
Research Question Do predictions based on power
dependence notions and those based solely on
structural centrality yield the same results in
negatively connected networks?

• Compare Betweenness centrality and Closeness
  Centrality
• Hypothesis
• In figure 1c, D gt E gt F in power

D
(note that this results from the weak connection
between F, otherwise the graph would be a simple
circle)
13
Social Exchange Direct
Power Dependence Theory
Def. 3. In any dyadic exchange relation Ax By
(where A and B are actors and x and y are
resources introduced in exchange), the power of A
over B (PAB) is the potential of A to obtain
favorable outcomes at Bs expense. Def. 4. The
dependence (DAB) of A on B in a dyadic exchange
relation is a joint function (1) varying directly
with the value of y to A, and (2) varying
inversely with the availability of y to A from
alternate sources. Power Dependence
hypothesis PAB DBA
14
Social Exchange Direct
Power Dependence Theory
H2 As the exchange process proceeds, E will
display more power use than the occupants of
position D and F, as seen by (a) an increase over
time in the amount of benefits gained and (b) a
greater absolute level of exchange benefit by E
in the final exchange phase.
H3 Power of E over F will be seen before that of
E over D H4 Position E will exert equal levels
of power over the occupants of F and D by the
final or stable phase of power use.
15
Social Exchange Direct
Experimental Exchange Process

• Communication only through computer, to restrict
  opportunities to the form listed in the network
• Subjects negotiate with each other for profit
  points by sending offers and counter offers (the
  value differed by exchange partner, so that F to
  F would lead to lower profits for each point
  traded (for F) than an F to E would, in keeping
  with the broken / solid line structure of the
  network.)
• Subjects could not compare their benefits to
  others benefits

16
Social Exchange Direct
Experimental Results
Results confirm Power dependence theory
17
Social Exchange Direct
Simulation Results
18
25
20
E
E
D
D
15
Points
10
5
0
1
2
3
4
5
6
Time
19
Social Exchange Generalized

• The questions behind generalized exchange differ
  from those in direct exchange
• Goods can often be transferred long distances,
  which cant occur in the negatively connected
  exchange networks.
• People do not directly benefit by being a giver
  -- no immediate reciprocity
• Interest in how exchange unites a large society
• Interest in problems of free riding and compliance

20
Social Exchange Generalized
History
The Kula Ring. One of the most cited examples of
a generalized exchange process is the Kula Ring.
...
...
...
...
Necklaces
Armbands
21
Social Exchange Generalized
Basic types
Network Generalized Exchange
Examples include Giving blood, reviewing
journal articles, carpools
22
Social Exchange Generalized
Basic types
Chain Generalized Exchange
Examples Kula Ring Some forms of Kinship
23
Social Exchange Generalized
In all generalized exchange systems, those who
give are not necessarily those who receive, and
thus there is great opportunity for free
riding. Bearman is interested in (a) identifying
a generalized exchange system and (b) explaining
how it came to be and how it is maintained.
24
Social Exchange Generalized
Why Exchange?
We exchange because some of the things we have we
cant use. Economic exchange rests on the
conversion of use-value to exchange-value.
Social exchange rests on exchanging use-values
directly. Something that person A has is useful
to B, but not to A, which makes it available for
exchange. In the Groote Eylandt case, the
incest taboo makes sisters unavailable for
marriage, and thus items of exchange. Almost all
classificatory kinship systems have a known
structure, based on who is allowed to marry who.
The puzzling point on Groote Eylandt was that the
normative rules guiding marriage were
self-contradictory, making it impossible to
develop a coherent marriage system.
25
Social Exchange Generalized
26
Social Exchange Generalized
The normative alternatives
Normatively, A male should marry his FZD
(Fathers Sisters Daughter)
27
Social Exchange Generalized
The normative alternatives
1) Ego in group 1 seeks a wife. Where does he go?
FZD
E
2)Es Father is in Group 3.
3) Es Fathers Sister was married into group 4.
FZ
F
4) Es FZs daughter goes to Group 1
5) E should marry his FZD, who is in 1.
28
Social Exchange Generalized
The normative alternatives
The bilateral cousin marriage system in western
genealogy terms
29
Social Exchange Generalized
The normative alternatives
B
A
30
Social Exchange Generalized
The normative alternatives
E
MMBDD
Who Should E marry? Es F is in 5 Es M is in
7 Es MM is in 4 Es MMB is in 4 Es MMBD is in
6 Es MMBDD is in 1
F
MMBD
M
MMF
MMB
MM
MF
31
Social Exchange Generalized
The normative alternatives
MM
MMB
F
MMBD
M
F
MMBDD
E
(Second Cousins)
32
Social Exchange Generalized
The normative alternatives
On Groote Eylandt, names for kin could fit in
either system. But logically, the two cannot
occur at the same time. Ethnographers of Groote
Eylandt concluded that the kinship system there
was a jumbled mess. The only thing that all
ethnographers agreed on was that people could not
marry within their own moiety. But is it?
People seemed to know who to marry, what patter,
if any, did their marriages fall under? To test
this, use data on kinship status among the
aborigines and block model the movement of wives
across the system.
33
Social Exchange Generalized
The raw data for the analysis are 5 relationship
matrices

• MB, M, MBS, MBD
• MMBDS, MMBDD, W, HZ,WB,H
• ZS, ZD, FZD, FZS
• FMB, FM, DD, DS
• MF, MFZ
• M Mother, B Brother, D Daughter, S Son, Z
  Sister
• Note these are classificatory equivalents on
  Groote E.

34
Social Exchange Generalized
Marriage patterns across named section, grouped
by Moiety
Log-Linear model shows that, with respect to
named section, mixing is random within moiety
35
Social Exchange Generalized
After block modeling the kinship relations,
Berman calculated the flow of wives across
blocks. The result was a near-perfect cycle.
36
Social Exchange Generalized

• Where does this structure come from?
• It can be maintained, once in place, in many ways
  (Balance, self interest), but that is not
  sufficient to explain where it came from.
• Bearman argues that chain generalized exchange
  follows because of the demographic pressure
  induced by the the great age difference in
  marriage (18 years) and polygamy.