Week 04 - Comparing alliances and game theory

In the readings this week about alliances, there were a lot of similarities to game theory which is often studied in political science and economics.  Game theory is defined as, “a branch of applied mathematics that is used to model interdependent decision making.” (Ishiyama, 2011).  In game theory, the goal for both parties involved is to minimize the impacts or loss to them and evaluate the situation and decisions using a payoff matrix (Banerjee, 2005).  The payoff matrix can help decide which decision, whether to cooperate or defect, provides the most benefit and the “sole intention of every player involved in a game is to maximize his own payoff” (Banerjee, 2005, p. 1).  This goal is very similar with alliances in that before an organization enters into an alliance, the organization evaluates various aspects of the relationship to determine what kind of advantages are available and also if this relationship will further their own strategic goals (Gommes-Casseres, 1993). With game theory and alliances, there are two parties evaluating one another while guarding their own interests and trying not to be taken advantage of.  

Trust and trustworthiness are two factors in alliances that are also factors in game theory. The definition of trust in an alliance is “mutual confidence that no party to an exchange will exploit another’s vulnerabilities” (Barney & Hansen, 1994, p.3).  Barney & Hansen (1994) also describe three different types of trust that can shape an alliance and determine its success. The same definition of trust can be applied to game theory, specifically to the prisoner’s dilemma which was first explored by mathematician Albert Tucker in the 1950’s (Darity, 2008).

The prisoner’s dilemma is a hypothetical situation in where two people are arrested for a crime. After being arrested, the two people are then separated into different rooms and interrogated. The police suspect the two suspects collaborated on the crime. The suspects have some choices, they can remain silent (cooperate) or confess (defect).  If both stay silent, the police cannot convict either suspect.  If one defects and tells the police what happened, that suspect will have a lesser sentence while the other will have a more severe one.  If both suspects confess, their sentences will not be as severe as if just one of them confessed, but is still more severe than if both had remained silent.  A hypothetical payoff matrix from Banerjee (2005) is pasted below to summarize the various options.

 

Here's a link to a Dilbert video that explains the prisoner's dilemma:



The payoff matrix in game theory can be compared to semi-strong trust and strong form trust with alliances; the same external factors as well as morals apply to the prisoner’s dilemma.  The governance or external laws and regulations that impose penalties on opportunistic behavior that exist with alliances (semi-strong trust) also exist in the prisoner’s dilemma (Barney & Hansen, 1994). In the prisoner’s dilemma, it can be assumed that both suspects are rational and want to minimize their prison sentence.  This is a form of governance that punishes someone for taking advantage of another’s vulnerabilities. If the two suspects also know one another or are related, you can assume that they have some morals and would feel guilty or immoral if they told on the other one and left them with a longer prison sentence, which is an example of strong form trust in alliances (Barney & Hansen, 1994). 

There are additional factors that can impact how a person makes a decision in the prisoner’s dilemma, such as repeated play or interaction. If the one suspect will interact with the other suspect in the future or if the two have a broader strategy, then one is probably not going to confess and leave the other behind. This same factor can impact alliances. If one organization enters into an alliance as part of a broader strategic goal, then the organization will not be as likely to expose the other’s vulnerabilities and will put more effort into making the relationship successful (Kanter, 2010).

Game theory and alliances are both about protecting your own vulnerabilities and trying to maximize on benefits.  Trust and payoff evaluation are two key factors that are seriously evaluated by the individuals or organizations considering entering into an alliance with another organization, and are also two key factors evaluated in game theory.   

References

Banerjee, A. (2005). A dynamic game-theoretic approach to the modified prisoner’s dilemma. Oxford University Press, 493-502. doi:10.1093/jigpal/jzi037

Barney, J. & Hansen, M. (1994). Trustworthiness as a source of competitive advantage. Strategic Management Journal, 15, 175-190.

Darity, W. A. (2008). In International encyclopedia of the social sciences (2^nd ed). Detroit, MI: Macmillian Reference USA

Gomes – Casseres, B. (1993). Managing international alliances: A conceptual framework. Harvard Business School, 1-20.

Ishiyama, J.T. (2011). In 21^st century political science: A reference handbook. (2011 Ed).

Kanter, R. (2010). Collaborative advantage: The art of alliances. Harvard Business Review. 1-16.