Rethinking reputation: Group membership and trustworthiness

1
Rethinking reputation Group membership and
trustworthiness

• Margaret Foddy
• Carleton University
• Ottawa, Canada
• Third CEFOM/21 International Symposium,
• Sapporo, Japan

2
Outline of talk (for the tired ones)

• Review theoretical background for ingroup trust
• Summarize work on selective interaction that
  results from ingroup trust
• Convince you that category based trust is robust
  but not necessarily strong
• Consider reputation in the context of ingroup
  identity, and its effect on group-based trust
• Report a study on reputation (when it means, what
  does x know about what y did to a person in the
  ingroup/outgroup)
• Speculate on possible implications across
  societies, and theoretical models (especially the
  role of category based trust)
• Consider links to other processes such as
  exclusion from groups, willingness to stay in, or
  leave groups,

3
BACKGROUND

• The allocator studies and trust in strangers
• Focus on unilateral trust, not cooperation
• Explored the utility of social identity and
  self-categorization theory for understanding
  group-based trust in strangers
• Collaboration with Michael Platow, (now
  Australian National University) Toshio
  Yamagishi and Toko Kiyonari, Hokkaido University

4
Trust expectation that an other person will
perform an action which is beneficial, or not
detrimental to us

• When we trust, we commit ourselves to an action
• agree to put ones outcomes in anothers hands
• take a cooperative action
• Make oneself vulnerable
• Before we know what the other will do
• When there is no assurance mechanism

5
Trust in strangers is important in establishing
new relationships

• Whom shall we trust?
• People whose interests are aligned assurance
  (encapsulated trust Hardin)
• People who share category and/or group membership
  with us group-based trust
• People who have proven trustworthy to us, or to
  others (reputation)

6
(In)group heuristic?naïve theory of groups

• People believe others in their own groups will
  treat them well
• People do in fact show higher ingroup fairness,
  are subject to principles of ingroup fairness
  and trustworthiness (Yamagishi)
• Ingroup trust may be a second order belief, or
  one based on taking the perspective of other(s),
  producing a consensus that groups can be relied
  on if you are included within its boundaries
  (Tyler)
• The effects are relatively weakpeople trust
  ingroup more, but still place trust in strangers
  outside own group
• (because)
• Group boundaries are relatively fluid (Brewer)
• Salience principles
• People extend and contract the boundaries of the
  group this may limit the utility of group-based
  trust for both theorist and actors
• Interesting question is this a psychological,
  sociological or economic question? (for
  discussion later)

7
Group identity and trust what is the
basis?Shared group membership affects

• Evaluations of ingroup members vs outgroup
  (ingroup more positive? And if so, why?) Turner
• Differential distribution of positive outcomes to
  ingroup members Tajfel
• Expectations of fairness and cooperation from
  others in the group (Yamagishi)
• ?Provide a categorical basis for trust in
  strangers

8
Experimental paradigm

• Allocator game a 2-dictator game
• Allocator has X to split between self and
  recipient focus is on recipient
• 2 Allocators one each from
• the persons ingroup (e.g.La Trobe Uni)
• an outgroup (e.g. Melbourne Uni)
• RECIPIENT chooses one allocator to receive money
  from, and gives reasons for choice, predicts how
  much each allocator would have given him or her
• We assess relative, not absolute, trust, but
  results are similar when third choice option of
  sure thing is provided similar when there is
  direct reciprocity as in the Berg game)

9
Results from choose the allocator studies

• Strong preference for ingroup allocator, but not
  extreme differences in expectations of positive
  treatment
• Preference holds for large and small ingroup
  categories (in the West and East?) for intact,
  minimal and categorical groups
• Preference disappears if allocator does not know
  of shared identity blocks group based trust
• Preference for ingroup allocator even if outgroup
  stereotype is more positive (even economics
  students prefer ingroup (economics) allocator)
• Group stereotype has weak effect on expectations
  and preferences
• Ingroup heuristic includes expectations of fair
  treatment from fellow group members?reduces
  uncertainty, forms basis for trust

10
Does shared identity matter only if there is no
other information?

• What about a persons reputation?
• Is it more important than shared group
  membership ?

11
How has reputation been conceptualized?

• the estimation in which one is held character
  in public opinion the character to attribute to
  a person, thing or action repute
• (plus a large literature in economics, political
  science, sociology etc. indicating that there are
  multiple dimensions to reputation)
• Personal reputation first or second hand
  knowledge of the pattern of a persons past
  behavior (e.g. fair cooperative competent)
• Organizational reputation as above
• Summarized in trait terms, generalized or a
  record of behavior and outcomes (e.g. a
  university)
• reputation is owned by a person, and can be
  manipulated, communicated, tested in law, etc.
• Reputation reduces uncertainty can be a positive
  or negative resource
• Little research attention to the group contexts
  in which a person establishes a reputation
• Who is the relevant public?
• What mechanisms are in place to govern true and
  false reputation (Toshio)

12
Groups, identity and reputation

• Category vs target based reputation Reputation
  developed by/attached to the group may be
  transferred to individuals with no information
  about the individuals behavior (e.g. nurses
  used car salesmen) naïve theory says people can
  be expected to act in accordance with the group
  stereotype, but individuating information may
  reverse the stereotype-based inference
• Ta da! (Main point)
• Effects of reputation on trust depend on the
  relationship between the truster and the target
  of trust
• Group membership may condition interpretation of
  information about anothers behavior
• Expect different behavior from ingroup and
  outgroup members towards self
• A third persons behavior towards an outgroup
  member may not be informative of their likely
  behavior towards an ingroup
• Category based ingroup trust may
• Replace reputation
• Combine with reputation
• Be replaced by reputation

13
How does reputation workin the group context?

• Reputation information may individuate the
  allocator and produce attributions to traits, or
  inferences about the persons values (e.g. fair
  shady exploitative)
• Individuating information may be regarded as more
  reliable/informative than shared group membership
  as basis for trust
• Information about allocators behavior may block
  group categorization effects block effects of
  group heuristic and in-group favoring norm
• Forecasting cross-cultural questions
• Is reputation information more important in the
  East or the West?
• Are contextual effects of the group relationship
  between truster and trustee more important in
  East or West/

14
Or, is reputations effect on trust moderated by
the group identities of the people involved?

• Information about outgroup members treatment of
  their own groups may not be informative fairness
  within the outgroup does not predict fairness to
  those not in it
• Information about ingroup members treatment of
  outgroup may not be predictive of how they will
  treat ingroup members, or may confirm that
  positive ingroup bias will operate for ingroup
  (confirm the naïve theory about ingroup behavior)

15
Experimental paradigm reputation

• Choose between ingroup (A) and outgroup (B)
• Recipient told allocators scheduled to make two
  separate allocationsone to ingroup (A), one to
  outgroup (B) recipient
• Experimental conditions
• 1.There is reputation information about neither
  allocator (baseline)
• OR about what ONE allocator has given another
  recipient who was
• 2. From allocators ingroup
• 3. From allocators outgroup
• No reputation information about the other
  allocator, who has not done second allocation

16
Information about ONE allocators division of 16
in a previous trial
17
Predictions

• If individual reputation is basis of choice, the
  fair allocator (regardless of target) will be
  preferred over the allocator with no reputation
  information, regardless of group, and the
  no-reputation allocator preferred to the one
  who has been unfair in the past
• If group membership affects the meaning of past
  fair behavior, then
• Past fairness within the outgroup (B?B) will not
  increase choice of outgroup by A
• Fairness by outgroup to ingroup in past (B?A)
  will reduce ingroup preference (not a typical
  intergroup situation)
• Unfairness within the ingroup (A?A) will reduce
  ingroup preference (allocator violates ingroup
  norms)

18
Frequency of choice of allocator when allocator
has been FAIR (all choosers from group A)

• Ingroup(A) Outgroup(B)
• Control 18 3
• Reputation info
• A?A 18 2
• A?B 17 4
• B?B 17 4
• B?A 6 14

19
Expected of funds Allocator was FAIR

• Ingroup(A) Outgroup(B)
• Reputation info
• A?A .48 .34
• A?B .46 .38
• B?B .48 .34
• B?A .37 .42
• Note recipients made predictions for both
  allocators, regardless of choice made

20
Frequency of choice of allocator when allocator
has been UNFAIR (all choosers from group A)

• Ingroup(A) Outgroup(B)
• Reputation info
• A-A 10 9
• A-B 10 10
• B-B 17 3
• B-A 17 3
• Control 18 3

21
Expected of funds to new recipient Allocator
was UNFAIR

• Ingroup(A) Outgroup(B)
• !A?A .32 .28
• A?B .28 .32
• B?B .39 .18
• B?A .32 .29
• Overall lower expectations than in FAIR
  condition
• Smaller difference in expectations from A and B
• Reputation information does not undermine trust
  in ingroup allocator as in ingroup allocator
• (!here is an ingroup member who does not treat an
  ingroup member well
• (here is an outgroup person who does not treat
  an ingroup member well)

22
When ingroup allocator has been unfair, why do at
least half still choose ingroup?

• Recipients gave reasons for choice of allocator
• Reasons coded
• Those choosing the unfair ingroup allocator
  referred to shared reciprocal obligation, common
  interests and similarity as a basis for expecting
  fairer treatment from ingroup compared with
  outgroup
• Those choosing outgroup did not refer to shared
  group membership main reasons were the unfair
  past behavior of the ingroup allocator
  (reputation), and decision to take a gamble with
  the outgroup

23
Summary of results

• People do NOT prefer an outgroup allocator who
  has been generous to a fellow outgroup member
  (allocators ingroup), over an ingroup allocator
  about whom they have no information
• People DO prefer an outgroup allocator who has
  been generous to a member of the persons own
  ingroup (B was fair to A)attribution to
  individual traits? People are evenly divided
  about reputation information when
• Ingroup member is unfair to the ingroup
• Ingroup member is unfair to the outgroup

24
Conclusions 1

• Reputation is not set up in a vacuumbehavior of
  a target towards others is interpreted in part in
  terms of the relationships of self and the
  target, and the target with others (this
  information may not always be known)
• Expectation of ingroup fairness is the default
  option not so clear that outgroup unfairness is
  the default (Brewer, 1981 Mummenday, 2000)
• Reputation may help identify the prevailing group
  norms, and a given actors adherence to them
  clearer basis for trust or distrust
• When reputation information is ambiguous, person
  may apply a group, individual, or universalistic
  template to interpret the information may
  reflect level of self-categorization, situational
  cues
• We know little about the relationship between the
  source of reputation information (ingroup
  outgroup) and its influence on future trust of
  ingroup and outgroup members

25
Group-based reputation and trust the same
across cultures?

• Will simple shared category activate trust in
  cultures where group implies a network of
  obligation and sanction (assurance)?
• Is reputation information more important in
  collectivist or individualistic cultures?
  Will it destroy the ingroup preference
  completely?
• Are the norms of ingroup favouritism the same
  across cultures?
• Are the effective groups the same across
  cultures? Categoriesminimal groupsconnected
  groupsdyads? Reputation can operate in all of
  these, but may have different effects
• How is reputation information conveyed? Do we
  trust some informants more than others
  (second-order trust may arise from shared group
  membership with informants). Does it matter who
  told us? (e.g. an outgroup member told me that
  Toshio is not reliable will I believe this? If
  an ingroup member tells me he is unreliable
  (joke), I may be more willing to believe it

26
Is individual reputation more important in
collectivist societies?

• The yes side
• If trust is based more on assurance and specific
  relationships in collectivist societies (i.e. not
  generalized trust), reputation should be more
  importanttarget-based trust should dominate
• Category-based trust is weaker in collectivist
  societies, so reputation information will
  dominate (evidence is mixed)
• formal assurance systems in Western cultures may
  make reputation irrelevant
• Group loyalty makes universal principles of
  fairness irrelevant
• The no side
• Individual reputation in Western societies is
  very important, and is used to make attributions
  to individual as a basis for predicting future
  behavior and trustworthiness the same applies to
  organizations
• Particularistic familism is as common in the West
  as in the East
• Reputation is an important part of English common
  law (you can sue someone for damaging your
  reputation)
• E-business puts a lot of emphasis on reputation
• However, the issue of whether reputation is
  interpreted in group terms has not been
  systematically explored across cultures

27
Why does reputation matter?

• Conditions the interpretation of behavior to
  ingroups and outgroups
• Forms the basis for selective interaction
• May form the basis for inclusion or exclusion of
  people from groups
• Affects peoples estimates of whether a common
  resource is viable, and therefore willingness to
  stay in a group (e.g. if all members have a
  negative reputation, group will be less viable
  than if all positive)
• Why do ingroup effects on reputation matter?
• Double standards in judging people (more lenient
  to ingroup)
• Informative about naïve theory of groups
• Important to theories of group identification
• Window on cross-cultural differences in how trust
  is formed, and how trustworthy behavior is
  generated

28
Conclusions 2

• Cultural differences in trust, the importance of
  reputation, and the willingness to exclude others
  from the group are not yet well-established
• Minimal group effects are obtained in Japan and
  China
• Trust in category members occurs in both types of
  culture (sometimes)
• There is not a lot of information about how group
  processes may differ between the East and the
  West
• Globalization means that the social structural
  sources of differences may be disappearing
• We have almost no idea how the Middle East fits
  into the picture!! Is behavior in these cultures
  a reflection of shared beliefs and second order
  beliefs, or coercion?
• We need theoretical frameworks to understand any
  differences
• Naïve theory of groups
• Content of cultural beliefs
• Social structural influences? incentive
  structures