Trust, Conflict, and Cooperation:

Trust, Conflict, and Cooperation: A Meta‐Analysis
Daniel Balliet VU University Amsterdam
Paul A. M. Van Lange
VU University Amsterdam
Balliet, D., & Van Lange, P. A. M. (December, 2012). Trust, conflict, and cooperation: A meta‐analysis. Psychological Bulletin.
Trust and Cooperation
• Morton Deutsch (1959)
– Cooperation in a PD must involve trust
Conceptualizations of Trust
• Beliefs about other’s reliability, confidence, and predictability (Dasgupta, 1988; McAllister, 1995; Sitkin & Roth, 1993)
• Beliefs about other’s concern for one’s well‐
being, a relationship, or group (Holmes & Rempel, 1989; Mayer et al., 1995; Rousseau et al., 1998; Yamagishi 2011)
Conceptualizations of Trust
• “Expectations of benign behavior from someone in a socially uncertain situation due to beliefs about the person’s dispositions (including his feelings towards you)” (Yamagishi, 2011, p. 27)
• “Psychological state comprising the intention to accept vulnerability based upon positive expectations of the intentions or behavior of another” (Rousseau et al., 1998, p. 395)
Situations that Afford Trust
• Social Interdependence – Each person’s behavior affects each other’s outcomes
• Conflict of Interest (or Vulnerability) – When a person has a tempting option to behave in a way that results in the best outcome for themselves, but reduces outcomes for another (or group). Research Question
• Does the degree of conflict across situations moderate the relation between trust and cooperation?
An Interdependence Perspective
• Non‐correspondence of outcomes (i.e., conflict of interests) is one dimension that describes
differences across interdependent situations • High conflict sitautions: A context when benevolent motives affect behavior and a situation when people condition thier behavior based on beliefs about other’s benevolence
• Low conflict sitautions: Benevolent motives less important determinants of behavior, and people are more motivated by other self‐serving motives
Rapoport (1967)
K = 3(R) – 1(P)/5(T)‐0(S) = .40
K = .00
K value
K = 1.00
Coordination
Zero sum
Mixed‐Motives Prisoner’s Dilemma
Conflict Across Social Dilemmas
• The index of cooperation has been shown to have a main effect on trust and cooperation in both 2‐person and n‐person social dilemmas (e.g., Krueger et al., 2012; Komorita et al., 1980; Murnighan & Roth, 1983)
• Index of cooperation has a negative relation with the effect size (i.e. trust and cooperation)
Overview of Meta‐Analysis
• Trust and cooperation in social dilemmas
• We only include studies on social dilemmas (common knowledge, simultaneous choice, mutual dependence)
• We include both dispositional trust and state trust (or exectations of others behavior)
• 212 effect sizes (198 published, 14 unpublished)
Codings
• K value (99 ES)
– (M = .45, SD = .19) Range = .16 to .95
• Type of social dilemma
– PD (99), PGD (77), RD (11), other (25)
• Expectations and dispositional trust
– Expectations (147), disposition (65)
• One‐shot versus iterated (# iterations)
– One‐shot (132), iterated (76)
Codings
• Group size
– Dyads (85), groups (94)
• Individual versus inter‐group
– Individual (192), inter‐group (20)
• Year of publication
– (M = 2001, SD = 13) Range 1960 to 2011
• Country
– 28 different countries
Analysis
• Correlation (r) is the DV
• Positive r indicates higher trust relates to higher amounts of cooperation
• Random Effects
Overall Average Effect Sizes
• State Trust
– r = .58, 95% CI = .54,.62
– No publication bias
• Dispositional Trust
– r = .26, 95% CI = .22,.31
– Slight bias to over estimate effect size (r = .21)
Moderators: K value
• State Trust (n = 70)
– b = ‐ .60, p = .004
– K .33 versus K .50, Q(1) = 9.32, p = .002
– K = .33 (31 studies), r = .65
– K = .50 (18 studies), r = .45
• Dispositional Trust (n = 25)
– b = ‐.14, p = .49 (random effects)
– b = ‐.16, p = .09 (fixed effects)
Moderators: K value
• Multiple regression including both state trust and dispositional trust effect sizes, controlling for the type of effect size
– K value: β = ‐.26, p = .003
Trust is more prective of cooperation in situations involving a larger conflict of interests
Multiple Regression Model
Discussion
• Trust matters more for behavior in situations involving a larger conflict of interests
• Trust is more strongly related to behavior in dyadic interactions compared to group interactions
Trust in Conflict
• We argue that trust involves beliefs about others benevolence
• Others benevolence is more predictive of their behavior in larger conflict situations
• Thus, people may more stongly condition thier behavior based on beliefs about other’s benevolence in larger conflict situations
Alternative Accounts
• Self‐Projection
• Self‐Rationalization
Implications
• Hostage Taking and Incentives
• When do people build trust?
• “Strain Test” situations in close relationships
Conclusion
• Paradox of trust (Yamagishi, 2011): When trust is difficult to establish, this is when trust is most important for behavior Thank You!
Balliet, D., & Van Lange, P. A. M. (December, 2012). Trust, conflict, and cooperation: A meta‐analysis. Psychological Bulletin, published
online first.
Balliet, D., & Van Lange, P. A. M. (2013). Trust, Punishment and cooperation across 18 societies: A met‐analysis. Perspectives on Psychological
Science, 8, 363‐379.
Prior Evidence
• A few past experiments find that both dispositional trust and state trust more strongly relate to cooperation in dilemmas that involve larger amounts of conflict (Parks Hulibert, 1995; Yamagishi & Sato, 1986)
• Close relationship research finds a larger difference between high and low trust individuals in stonger strain test situations (Campbell et al., 2005; Murray et al., 2003; Shallcross & Simpson, 2012)
Encapsulated Self‐Interest Perspective: A Competing Prediction
• Trust is a belief in other’s self‐interest (not benevolence) (Hardin, 1992; 2000)
• Trust predicts cooperation when people have a self‐interest to cooperate
• Trust should more strongly relate to cooperation in situations with less conflict
Trust in Social Dilemmas
• Social dilemmas always involve a conflict between self and collective interests (Dawes, 1980; Kollock, 1998; Komorita & Parks, 1994)
• Public goods dilemmas, common resource dilemmas, prisoner’s dilemmas
• Past research has clearly shown that trust positively relates to cooperation (e.g., Yamagishi, 2011)
Conflict Across Social Dilemmas
• Index of Cooperation ‐ indicates the amount of conflict of interests in a PD (Rapoport, 1967) – Temptation outcome (T)
– Reward for cooperation outcome (R)
– Punishment for non‐cooperation outcome (P)
– Sucker outcome (S)
Competing Predictions
• Interdependence Perspective
– Negative relation between the index of cooperation and the trust‐cooperation effect size • Encapuslated Self‐Interest Perspective
– Positive relation between the index of cooperation and the effect size
Analysis
• r (correlation) is the dependent measure
• Positive correlation indicates higher trust relates to higher amounts of cooperation
• Collapse non‐independent effect sizes
• Test for publication bias
– Duval and Tweedie (2000), Egar’s Regression • Mixed‐Effects and Random Effects Analyses
• Comprehensive Meta‐Analysis software
Correlation Amongst Moderators
Moderators: Null Findings
• Type of dilemma
• Measurement of expectations: Before or After choice
• Group Size
• Year of publication
• Participant payment
Moderators: Country
• Only included countries with 4 or more ES
• State Trust: Q(8) = 66.83, p < .001
• Dispositional Trust: Q(3) = 21.63, p <.001
In certain countries, trust has a stronger relation with cooperation
A note on Vulnerability
• Conflict
• Power
• Uncertainty
Individual versus Inter‐group
• Groups are more competitive and this reduces the relation between trust and cooperation
• Insko et al. (2005): When a group trusts that another group will cooperate, then they still compete with that group and attempt to take advantage of that cooperation
Culture and Trust
• This study provides additional behavioral data that there are subtantial differences in trust across countries in trust
One‐Shot versus Repeated Interactions
• Prior theorizing suggests that trust is a psychological mechanism that sustains reciprocal relations
• This implies that people will condition ther behavior on trust in others during repeated interactions, more than one‐shot interactions • We find no support for this perspective
Alternative Accounts
• Reduced variation in trust and cooperation in very low and high conflict situations
• Self‐Projection
• Self‐Rationalization
Moderators: Iterations
• State Trust: Q(1) = 0.00, p = .96
– One‐shot (94): r = .58
– Iterations (46): r = .58
• Dispositional Trust: Q(1) = 1.40, p = .24
– One‐shot (34): r = .24 – Iterations (24): r = .30
Moderators: Inter‐Group
• State Trust: Q(1) = 10.18, p = .001
– Individual (122): r = .60
– Inter‐group (20): r = .43
Trust is more predictive of cooperation between individuals compared to cooperation between groups