Theories of Marketing – Summary Week 5 Author: Huisingh
MA Summary Article 5.1 Kanagaretnam, Mestelman, Nainar & Shehata (2010)
Trust and reciprocity with transparency and repeated interactions
Abstract
This paper uses data from a controlled laboratory environment to study the impact of transparency
(i.e., complete information versus incomplete information) and repeated interactions on the level of
trust and trustworthiness (reciprocity) in an investment game setting. The key findings of the study
are that:
- transparency (complete information) significantly increases trusting behavior in one-shot
interactions. This result persists (houdt aan) in repeated interactions.
- transparency appears important for trustworthiness in one-shot interactions. In addition,
repeated interaction increases trust and reciprocity with or without transparency.
These results suggest that transparency is important in building trust in business environments such
as alliances and joint ventures which are loosely connected organizational forms that bring together
otherwise independent firms. It also provides support for the Sarbanes–Oxley Act of 2002 (SOX) and
similar legislation elsewhere which attempt to regain investors' trust in corporate management and
financial markets by stipulating enhanced disclosures.
Keywords: Transparency, Trust, Reciprocity, Repeated interaction, Business alliances, SOX
Reciprocity trustworthiness (betrouwbaarheid)
One-shot interaction effective after only one attempt.
Sarbanes–Oxley Act of 2002 (SOX) Amerikaanse wet op het terrein van het bestuur van
ondernemingen.
Introduction
Study we create a controlled laboratory environment in which participants interact in an
investment game (game theory, lab situation).
Our paper extends the Berg et al. (1995) investment game to study the impact of transparency (i.e.,
complete information versus incomplete information) and reputation building through repeated
interactions on the level of trust and reciprocity.
Our experiment resembles a business context of a simple two-partner alliance with a dominant
partner (who is the receiver in our experiment and decides on the profit-sharing rule). The other
partner's contribution to the alliance is similar to the investment game's sender (who provides
resources to the dominant partner). The synergy from an alliance relationship is captured by the
tripling of the investment in our experimental design.
There is anecdotal evidence in the popular press regarding the importance of the role of trust in
business transactions and how the loss of trust has led to the current dystopia in the financial
markets.
1
,The key finding of the study is that transparency significantly increases trusting behavior. Both
transparency and repeated interaction are important in eliciting trust from investors. The effect of
repeated interaction on trust persists over repeated rounds of investment decisions.
While transparency appears to be important for reciprocity in the one-shot investment game, it
does not have a significant impact in the repeated game environment. Participation in the repeated
game environment, however, leads to a significant increase in reciprocity in the first round of the
repeated investment game independent of transparency. This indicates the importance of repeated
interaction effects on reciprocal behavior.
Literature review, theoretical development and hypotheses
Trust is defined by Robbins and Langton (2003) as the “positive expectation that another will not
— through words, actions, or decisions — act opportunistically”.
Reciprocity is an echo of trust by the party that was trusted.
Information
Under complete information in the investment game, both the sender (investor) and the receiver
(responder) know each other's initial endowments (gift/donatie?) and the investment multiplier
(technology). We introduce incomplete information in order to price the veil of uncertainty in the
investment game by making the participants' endowments random and unknown to each other.
With incomplete information, the receiver is unable to determine the sender's actual endowment
and does not know how trusting the investor actually is. Croson and Buchan (1999) suggest that trust
is important in generating reciprocity. If an investor is unable to unambiguously signal her trust
through her investment, she may anticipate that the receiver will not reciprocate in a way consistent
with the actual level of trust revealed. This may lead to lower levels of trust and a lower group
surplus than under complete information
Hypothesis 1a: The null hypothesis is that in one-shot games, the level of trust is invariant to the
information condition, while the alternative is that incomplete information will
lead to lower levels of trust.
Hypothesis 1b: The null hypothesis is that in one-shot games, the level of reciprocity is invariant to
the information condition, while the alternative is that the level of reciprocity may
differ with incomplete information relative to complete information.
2.2. Repeated interactions
Most investment activities and business transactions are conducted on an ongoing basis rather than
as one time encounters. In a repeated interaction environment, one's reputation may be an
effective a priori control on ex-ante opportunism. We conjecture that in a repeated multi-period
investment game, subjects may attempt to create incentives that induce the other party to
cooperate (which is how we have defined building reputation).
This reputation building mechanism(the repeated interaction), is expected to encourage the sender
to trust more in order to influence the receiver to honor the increased trust with greater reciprocity.
Hypothesis 2a: The null hypothesis is that the level of trust is invariant to opportunities to build
reputation, while the alternative is that trust will increase within the context of a
repeated game relative to a one-shot game.
2
, Hypothesis 2b: The null hypothesis is that the level of reciprocity is invariant to opportunities to
build reputation while the alternative is that reciprocity will increase within the
context of a repeated game relative to a one-shot game.
The relation between trust (and reciprocity) and repeated interactions is also impacted by situational
variables, Church et al. (2005). Contexts, such as the state of information in the environment, may
affect trust and reciprocity.
Hypothesis 3a: The null hypothesis is that in repeated investment games, the level of trust is
invariant to the information condition, while the alternative is that the level of
trust may differ with incomplete information from that with complete information.
Hypothesis 3b: The null hypothesis is that in repeated investment games, the level of reciprocity is
invariant to the information condition while the alternative is that the level of
reciprocity may differ with incomplete information from that with complete
information.
Experimental design
The investment game
The session in which each individual participated consisted of a series of periods of a computer
mediated investment game. Senders were asked to make a decision about how much of their
endowment they wished to invest with their paired anonymous person. After the receivers received
this information, they decided how much of the resources they controlled they would return to the
sender. These values were reported to the senders and the total payoffs to the senders and receivers
appeared on their screens. Individuals were then rematched and a second period of sending and
responding took place. After the second period, individuals were reassigned to new partners and to
new roles. For the next four periods the investment game was repeated as described for the first
period. After the fourth period of the repeated game, partners were reassigned and roles were
reversed. At the end of the session, subjects completed a short questionnaire.
Trust and reciprocity
Senders' decisions in the investment game provide a measure of trust. Trust is measured as the
proportion of the endowment that is invested.
Reciprocity is measured as the proportion of the grossed up (in this case, tripled) investment
returned by the responder to the sender.
Results
One-shot game and the first round of the repeated game
- Trust in the one-shot game and the first round of the repeated game
The mean value of the differences between trust in the first round of the repeated game and trust in
the one-shot game across the individuals in this experiment is 0.173 and this is significantly different
from zero (p=0.000). This result rejects hypothesis H2a, that the opportunity to build reputation will
have no effect on trust in favor of the alternative that trust will be enhanced.
The differences between trust displayed under the conditions of complete information
(transparency) and incomplete information are statistically significant in the one-shot game and in
the first round of the repeated game (p=0.000 and p=0.013). These results reject hypotheses H1a
and H3a, that transparency has no effect on trust in favor of the alternative that trust will be
enhanced by transparency in the one-shot game and that it will be affected by transparency in the
repeated game. (see table 2 in article)
3
MA Summary Article 5.1 Kanagaretnam, Mestelman, Nainar & Shehata (2010)
Trust and reciprocity with transparency and repeated interactions
Abstract
This paper uses data from a controlled laboratory environment to study the impact of transparency
(i.e., complete information versus incomplete information) and repeated interactions on the level of
trust and trustworthiness (reciprocity) in an investment game setting. The key findings of the study
are that:
- transparency (complete information) significantly increases trusting behavior in one-shot
interactions. This result persists (houdt aan) in repeated interactions.
- transparency appears important for trustworthiness in one-shot interactions. In addition,
repeated interaction increases trust and reciprocity with or without transparency.
These results suggest that transparency is important in building trust in business environments such
as alliances and joint ventures which are loosely connected organizational forms that bring together
otherwise independent firms. It also provides support for the Sarbanes–Oxley Act of 2002 (SOX) and
similar legislation elsewhere which attempt to regain investors' trust in corporate management and
financial markets by stipulating enhanced disclosures.
Keywords: Transparency, Trust, Reciprocity, Repeated interaction, Business alliances, SOX
Reciprocity trustworthiness (betrouwbaarheid)
One-shot interaction effective after only one attempt.
Sarbanes–Oxley Act of 2002 (SOX) Amerikaanse wet op het terrein van het bestuur van
ondernemingen.
Introduction
Study we create a controlled laboratory environment in which participants interact in an
investment game (game theory, lab situation).
Our paper extends the Berg et al. (1995) investment game to study the impact of transparency (i.e.,
complete information versus incomplete information) and reputation building through repeated
interactions on the level of trust and reciprocity.
Our experiment resembles a business context of a simple two-partner alliance with a dominant
partner (who is the receiver in our experiment and decides on the profit-sharing rule). The other
partner's contribution to the alliance is similar to the investment game's sender (who provides
resources to the dominant partner). The synergy from an alliance relationship is captured by the
tripling of the investment in our experimental design.
There is anecdotal evidence in the popular press regarding the importance of the role of trust in
business transactions and how the loss of trust has led to the current dystopia in the financial
markets.
1
,The key finding of the study is that transparency significantly increases trusting behavior. Both
transparency and repeated interaction are important in eliciting trust from investors. The effect of
repeated interaction on trust persists over repeated rounds of investment decisions.
While transparency appears to be important for reciprocity in the one-shot investment game, it
does not have a significant impact in the repeated game environment. Participation in the repeated
game environment, however, leads to a significant increase in reciprocity in the first round of the
repeated investment game independent of transparency. This indicates the importance of repeated
interaction effects on reciprocal behavior.
Literature review, theoretical development and hypotheses
Trust is defined by Robbins and Langton (2003) as the “positive expectation that another will not
— through words, actions, or decisions — act opportunistically”.
Reciprocity is an echo of trust by the party that was trusted.
Information
Under complete information in the investment game, both the sender (investor) and the receiver
(responder) know each other's initial endowments (gift/donatie?) and the investment multiplier
(technology). We introduce incomplete information in order to price the veil of uncertainty in the
investment game by making the participants' endowments random and unknown to each other.
With incomplete information, the receiver is unable to determine the sender's actual endowment
and does not know how trusting the investor actually is. Croson and Buchan (1999) suggest that trust
is important in generating reciprocity. If an investor is unable to unambiguously signal her trust
through her investment, she may anticipate that the receiver will not reciprocate in a way consistent
with the actual level of trust revealed. This may lead to lower levels of trust and a lower group
surplus than under complete information
Hypothesis 1a: The null hypothesis is that in one-shot games, the level of trust is invariant to the
information condition, while the alternative is that incomplete information will
lead to lower levels of trust.
Hypothesis 1b: The null hypothesis is that in one-shot games, the level of reciprocity is invariant to
the information condition, while the alternative is that the level of reciprocity may
differ with incomplete information relative to complete information.
2.2. Repeated interactions
Most investment activities and business transactions are conducted on an ongoing basis rather than
as one time encounters. In a repeated interaction environment, one's reputation may be an
effective a priori control on ex-ante opportunism. We conjecture that in a repeated multi-period
investment game, subjects may attempt to create incentives that induce the other party to
cooperate (which is how we have defined building reputation).
This reputation building mechanism(the repeated interaction), is expected to encourage the sender
to trust more in order to influence the receiver to honor the increased trust with greater reciprocity.
Hypothesis 2a: The null hypothesis is that the level of trust is invariant to opportunities to build
reputation, while the alternative is that trust will increase within the context of a
repeated game relative to a one-shot game.
2
, Hypothesis 2b: The null hypothesis is that the level of reciprocity is invariant to opportunities to
build reputation while the alternative is that reciprocity will increase within the
context of a repeated game relative to a one-shot game.
The relation between trust (and reciprocity) and repeated interactions is also impacted by situational
variables, Church et al. (2005). Contexts, such as the state of information in the environment, may
affect trust and reciprocity.
Hypothesis 3a: The null hypothesis is that in repeated investment games, the level of trust is
invariant to the information condition, while the alternative is that the level of
trust may differ with incomplete information from that with complete information.
Hypothesis 3b: The null hypothesis is that in repeated investment games, the level of reciprocity is
invariant to the information condition while the alternative is that the level of
reciprocity may differ with incomplete information from that with complete
information.
Experimental design
The investment game
The session in which each individual participated consisted of a series of periods of a computer
mediated investment game. Senders were asked to make a decision about how much of their
endowment they wished to invest with their paired anonymous person. After the receivers received
this information, they decided how much of the resources they controlled they would return to the
sender. These values were reported to the senders and the total payoffs to the senders and receivers
appeared on their screens. Individuals were then rematched and a second period of sending and
responding took place. After the second period, individuals were reassigned to new partners and to
new roles. For the next four periods the investment game was repeated as described for the first
period. After the fourth period of the repeated game, partners were reassigned and roles were
reversed. At the end of the session, subjects completed a short questionnaire.
Trust and reciprocity
Senders' decisions in the investment game provide a measure of trust. Trust is measured as the
proportion of the endowment that is invested.
Reciprocity is measured as the proportion of the grossed up (in this case, tripled) investment
returned by the responder to the sender.
Results
One-shot game and the first round of the repeated game
- Trust in the one-shot game and the first round of the repeated game
The mean value of the differences between trust in the first round of the repeated game and trust in
the one-shot game across the individuals in this experiment is 0.173 and this is significantly different
from zero (p=0.000). This result rejects hypothesis H2a, that the opportunity to build reputation will
have no effect on trust in favor of the alternative that trust will be enhanced.
The differences between trust displayed under the conditions of complete information
(transparency) and incomplete information are statistically significant in the one-shot game and in
the first round of the repeated game (p=0.000 and p=0.013). These results reject hypotheses H1a
and H3a, that transparency has no effect on trust in favor of the alternative that trust will be
enhanced by transparency in the one-shot game and that it will be affected by transparency in the
repeated game. (see table 2 in article)
3
