Friday, September 7, 2018

Is an Unpredictable Leader Good for National Security? Think the goal is to keep your enemies guessing? Game theory suggests otherwise.

19 June 2018
Author: Drew Calvert
Featured Faculty: Sandeep Baliga, Professor of Managerial Economics & Decision Sciences at Northwestern University's Kellogg School of Managment

Summary:
The urgency of having a strong nuclear deterrence strategy during the Cold War sparked a debate between economist and Nobel laureate, Thomas Schelling, and President Richard Nixon about the effectiveness of the “madman” theory of international relations.  Nixon believed that having a reputation for unpredictability was the most effective deterrent, while Schelling, by applying game theory, proved that building a reputation of consistent behavior was much more effective. 
In this article, Sandeep Baliga, a professor of management at Northwestern University’s Kellogg School of Management, echoes Schelling’s game theory insights to explore the effects of unpredictable leadership on present day national security.  Using game theory to study international relations, Professor Baliga has found that meticulous strategy, not unpredictability, is the best approach in conducting foreign policy.  As Baliga puts it: “in national security, predictability can definitely pay.” 

Author Drew Calvert brings clarity to the connection between game theory and international relations by understanding incentives and how they work.  Incentives in game theory are best described in the principal-agent problem.  In the principal-agent problem, a principal needs to incentivize an agent whose interests may not align with those of the principal.  The author gives the example of an employer who might incentivize an employee to be productive by linking salary or bonuses to outputs such as sales.  “If the employee believes the employer will renege on paying him, or that her payment is random and independent of performance, there is no incentive for the employee to work.” 
On the other hand, an agent is more likely to respond positively to incentivization if he believes the principal will stay true to his word in an agreement. 

This notion is where consistency becomes imperative for the US to build a reputation for following through on policies designed to incentivize the actions of both allies and adversaries.  The example given in the article is the United States’ efforts to influence North Korea by incentivizing the de-acceleration of North Korea’s nuclear arms program.  Calvert states “if North Korea is left without a clear sense of how the US might respond to aggression or appeasement, its leaders might choose a more reckless and destabilizing course of action.” This could hold especially true when considering other internal, and external factors the country may weigh as important. The underlying logic for designing incentives in the principal-agent problem is, in many ways, parallel to the US-foreign adversary relations.  As Schelling stated, “if your opponents believe you will keep your word, then your word can shape their actions.”

Critique:
I think many that find issue with consistency in foreign policy do so because, in an age of deliberate deception and asymmetric warfare, predictability carries with it a stigma of “outdatedness.”  International relations is complex and situational, with each situation requiring a different response.  I find game theory to be a beneficial approach to the intelligence context because of its effort to model interaction between decision makers, accounting for the unpredictability of human nature. In his book “Game Theory: Analysis of Conflict,” Roger Myerson defined game theory to be the study of mathematical models of strategic interaction between RATIONAL decision-makers.  The article attempts to apply game theory models to understand the interaction between volatile decision-makers.  The article makes the case that game theory proves volatility and unpredictability to be detrimental to incentivizing both allies and adversaries, and I don’t disagree.  My critique comes in the form of a suggestion for more research to fill the white space in the article: even if game theory proves it doesn’t pay to be “the madman,” how do we apply game theory to incentivize “the madman”?




8 comments:

  1. Bryant,

    It seems that the idea behind this article would work the best for national security issues such as the use of nuclear weapons due to its reliance on most actors being rational. Do you think this currently or will eventually apply to the realm of cyber warfare? Can the U.S. realistically incentivize "the madman"?

    ReplyDelete
    Replies
    1. Billy - Essentially the anonymity and unpredictability of the cyber landscape can turn any actor into a "madman" by default. I think the current cyber warfare landscape exemplifies the dilemma with unpredictable behavior by the principal. Unclear cyber deterrence policies and inconsistent responses to acts of cyber warfare and cyber crime by the US government has cultivated a landscape that leaves are government, its critical infrastructure, and our nations businesses prone to breaches and attacks.

      Delete
  2. Bryant,

    I'd agree that deception and asymmetric warfare are complicating factors to this analysis of game theory as applied to international relations. Theoretically, it should cause actors to respond asymmetrical to an asymmetric attack but so far current examples have shown this not to be the case. In response to asymmetric attacks from Russia, the US has acted quite predictably despite the unpredictability effect of the "madman." Even before the madman, asymmetry was met with conventional, predictable outcomes. I'd argue there asymmetric responses to asymmetric attacks poses a potentially higher risk of escalation although this is due to ingrained nature of nuclear escalation ladders into decision-making against adversaries. Does the author every address whether we should totally reform of the way we think about game theory and shift it away from nuclear escalation?

    ReplyDelete
    Replies
    1. Harry - No he doesn't suggest reformation and the nuclear escalation was just a concrete example that was timely and relevant. I think Calvert, through the insights of Baliga, is using game theory to address the ineffectiveness of unpredictable behavior by nation-state leaders. Because game theory has obvious parameters their suggestions are limited but follow a line of logic that makes sense to me, specifically in reference to relations with North Korea, Iran, and even possibly ISIS.

      Delete
  3. Bryant,
    Your discussion on the interaction of game theory and international relations is both timely and insightful. Considering it is the decisions of both "players" at the international level that produces a mutual outcome, I find it essential that decision-makers consider their ally/adversary's historical behavior and personal agenda when making their own decision. I agree with Dave Calvert that a predictable (and trustworthy, reliable, etc.) strategy is a more effective strategy than that of the "madman". My discussion on applying game theory to determine how personal preferences influence decision-making may help as a next step to answering your final question. If we can identify the preferences of our ally/adversary, we may be able to take advantage of those preferences to incentivize the "madman" to cooperate with our own agenda. Great post!

    ReplyDelete
    Replies
    1. Thanks Tom! You bring up a great point that is echoed in the article as Calvert talks about Baliga's body of work: "Using this model, he has found that meticulous strategy, not volatility, is ultimately the best approach."

      Delete
  4. So far much of what I have read on game theory says that is has been proven that game theory is successful if the players act rational. I agree with you when you say you find game theory to be beneficial to intelligence for the sake of creating an advantageous and honest interaction between the players. Unfortunately, rational actors do not exist as much as they should when it comes to international relations or even in the intelligence community.

    ReplyDelete
  5. The article I read talked about game theory and the benefits of bluffing or not. I think your article is hinting the same thing by determining if the decision maker is acting rational or not. If the decision makers is always bluffing or acting irrational it is safer to assume that pattern will continue. Yes the decision maker is still not rational, however it does make them predictable.

    ReplyDelete