You are on page 1of 19

Counterterrorism Cooperation and the ASilver Bullet:@ A Game Theory Illustration

by Jeff Breinholt 1
Introduction Shortly after 9/11, President Bush presented the other nations of the world a binary choice in how they could respond to international terrorism. "Over time it's going to be important for nations to know they will be held accountable for inactivity," he said at on November 6, 2001, at a joint press conference with French President Jacques Chirac, AYou're either with us or against us in the fight against terror." 2 Although this statement may have served an important function at the time, it is clear almost five years later that foreign countries= counterterrorism efforts cannot be judged by reference to an either/or choice. Bilateral relations, as most diplomats tell us, are more subtle, and have multiple layers. They do not lend themselves to binary classifications. The same is true of a concept that has been discussed extensively since 9/11: information-sharing. The various commissions that studied what government failings contributed to the World Trade Center and Pentagon attacks found fault with how intelligence was handled in the U.S. intelligence community. These studies led to structural changes designed to assure that information flows more freely to personnel capable of connecting the dots and authorizing proactive counterterrorism actions. They have generally not delved in to the structural features of the intelligence and law enforcement communities, the incentives that developed among the players within these structures to hoard or share information, and the extent to which these factors or structures drive this behavior. Instead, these studies make the mistake of placing a more complex environment into a binary system: those who, before 9/11, refused to share information were wrong. Those who pushed for greater dissemination were prescient. This assessment is as simplistic as judging foreign countries as either Awith us@ or against us in J.D., UCLA 1988; B.A. Yale 1985; Deputy Chief, Counterterrorism Section, United States Department of Justice. The views expressed in this article are the author’s own and do not necessarily reflect those of his employer. New coverage of this statement can be found at: www.archives.cnn.com/2001/US/11/06/gen.attack.on.terror/.
2 1

-1-

Counterterrorism Cooperation & Game Theory J. Breinholt May 2006

Game theory analysis. Law Review.international cooperation against terrorism and information-sharing .fighting terrorism. Part III describes a negotiated settlement. One such concept is what I refer to as the Asilver bullet@ concession: a commitment by the recipient country that the source country maintains control over the use of the intelligence it is sharing.together. the factors that influence how terrorism-related intelligence flows between two countries is complex. however. the simplicity breaks down. This analysis can lead to the discovery of institutional reforms. generates information that is pertinent to counterterrorism actions. Part I of this paper describes the assumptions for the bilateral counterterrorism cooperation game . each of which have an intelligence capability which. suggests some institutional reforms and concepts that will maximize the prospect of an optimal equilibrium. Fall 1991. and to illustrate the dynamic comes into play as a result of the combination. This article seeks to bring two concepts . Part IV describes an actual example of the dynamic discussed in the game.two allied countries. AGame Theory. given the assumptions described. It seeks to answer complicated questions through the lens of a simple competition. What does game theory suggest about the ideal set of circumstances for fighting terrorism? As shown in this article. Law. and the strategy that comes into play by rational actors. based on their rational assessment of costs and benefits of the various choices available to them in the game.@ Univ of Cinc. including the right to pull it when it appears that an unpalatable amount of disclosure is being threatened. Part II describes the two classic game theory devices. Where the equilibrium is not the most mutually beneficial. It does so through an application of game theory. Over time. where two countries negotiated a solution that led to a significant counterterrorism result. The conclusion discusses the implications of these Martin Shubick. Part V discusses what circumstances and practices from both the hypothetical game and the real-life example might be institutionalized to maximize international counterterrorism cooperation. and considers them in the context of cross-border sharing of counterterrorism intelligence. a tool that over the last few decades has been applied legal analysis. 3 -2- Counterterrorism Cooperation & Game Theory J. the game theorist seeks to explain what circumstances led to this result. and bilateral counterterrorism cooperation is not subject to binary choices or jingoistic solutions. Game theory analysis results in the establishment of an Aequilibrium:@ the strategy that would and should be adopted by the players. 3 Game theory is the study of the basic elements of conscious conflict and cooperation among multiple people. and the Concept of Competition. Breinholt May 2006 . through sensitive sources and methods.

For example. This is necessary. In intelligence parlance. Each With An Intelligence Service The first assumption is that the players are two allied countries that are jointly interested in fighting international terrorism. Each country has an intelligence capability. they are motivated to help the other. Thus. so sensitive is some of the each country=s intelligence that it is not shared at all within anyone outside the country.@ an abbreviation for ANo Foreign Dissemination. a former NSA Counterterrorism Cooperation & Game Theory J. The Assumptions: Two Allied Countries. there is a concept known as Asingular intelligence.findings to counterterrorism and information-sharing generally. for example. 4 This decision is based on a decision that the particular sources and methods of its acquisition will be jeopardized by such sharing. Breinholt May 2006 -3- .@ Information is considered Asingular@ if its mere dissemination will allow the recipients to know how it was collected. singular information is more sensitive and less disseminated than non-singular information. because it expends resources developing them. 5 Each country is the owner of all of the intelligence it collects.even with a trusted ally . Because they are allies and have a mutual goal (and common enemy). as each country=s intelligence capability represents an extension of its sovereignty. Thus. even between allied countries. the spy will be prosecuted. Although the intelligence each country collects is sometimes shared with other nations. As a result. because the sharing of certain purloined information . can be obtained by diplomatic (or military. of course. if a nation catches a spy attempting to steal intelligence on behalf of a friendly foreign country. each side takes precautions against its intelligence being stolen by other countries. as long as this help if this can be done in a way that is consistent with their own national security. Consent. which means that it invests resources in the collection and analysis of information it uses to protect its citizens and its assets from threats. 6 Within the United States. The fact that the two countries are allies and equally committed in the battle against international terrorism does not mean that they do not hide intelligence from the other. It has an interest in maintaining its source and methods as long as possible. or economic) pressure. certain classified documents are marked with the letters ANOFORN.@ This sub-classification denotes that the document contains information that should not be shared with foreign partners. In fact. including its closest allies.ruins certain sources and methods forever. 6 5 4 This point is illustrated by the American prosecution of Jonathan Pollard. I. there is no way for one country to legally compel intelligence collected by another. such sharing is generally done in such a way that the recipient countries do not thereby become aware of the sender=s specific intelligence collection capability. Such sharing can only come from consent.

7 II. cannot be killed with stag-hunting equipment. Breinholt May 2006 7 -4- . the fact that a country is a close ally of the United States . The Games: The Stag Hunt and Prisoner=s Dilemma With game theory. internal consistency.does not mean that every request for terrorism-related intelligence will be honored. then. With these assumptions. each country also has operational decision makers who appreciate what type of intelligence . One is called Stag Hunt. Pollard. etc. Before the hunt starts. They can immediately grasp the relevance of intelligence that is discussed. Meanwhile. they speak the same professional language. who must be outfitted accordingly. v. can be used to justify a particular action. while a single player can successfully get a hare alone with the proper hare-hunting equipment. 1992) Note that the above description illustrates the limitations of the binary choice described by President Bush. 959 F. Each player. the payoff of a single stag (meat worth $4) is greater than what they would receive if they each caught their own hare (meat worth $2). source reliability and track record. however. Awith us@ . It works like this: there are two people who will be going hunting at the same location.in the President=s words. without the benefit of knowing what the other chooses.Despite the fact that each country controls its own intelligence. which approximate reality. is faced with a binary choice. For each player.C. Successfully hunting a stag requires to cooperation of another person. That is. Each country has adopted an intelligence system that rewards relevant personnel for coming up with information and analysis that proves useful in taking a certain proactive government action. This means that when people who occupy comparable position in each country=s intelligence apparatus get together.2d 1011 (D. The matrix showing the contingencies and payoffs looks like this: analyst convicted of spying for Israel. A hare. Cir. they are not in communication with each other. See U. there are a number of useful paradigms to illustrate the dynamic of intelligence-sharing between countries. Each must decide unilaterally whether to be outfitted to hunt a stag (a larger game animal) or a hare.S. based on their training.judged by volume. the effectiveness of its intelligence community is judged by how often it provides intelligence that is useful to operational decision-makers. Counterterrorism Cooperation & Game Theory J. the intelligence collectors and analysts within each country are trained in such a way that they yearn to have their products used for the larger good.

They are each offered a binary choice: cooperate. unless at least one of them turns state=s evidence. The Prisoner=s Dilemma. are arrested by the police. the result may be a loss of sources and methods without the benefit the full pooling of information. A and B. it illustrates the dynamic that is at play in the question of intelligence-sharing among countries who are jointly interested in combating international terrorism. in which case your best option would be to hunt for hare.Stag Stag Hare Player 1: $4 Player 2: $4 Player 1: $2 Player 2: $0 Hare Player 1: $0 Player 2: $2 Player 1: $2 Player 2: $2 The best result would be for both players to cooperate. involves a negative payout . and choose the stag hunting equipment. The police have insufficient evidence for a felony conviction. $4 per person. or stay silent. unlike the Stag Hunt. Assume that two countries with an intelligence apparatus and a mutual goal of fighting terrorism must decide whether to share their sensitive intelligence. you might conclude that the other player is going to look out for himself. Killing a stag would yield $8 of meat.time in prison . If only one country cooperates. A completely self-interested strategy means no cooperation. What if you opt for the staghunting equipment. The Stag Hunt game is intentionally unreal. where your payout will not depend on any cooperation. although it will be lower. If one testifies for the prosecution against the other and the other remains silent. they cannot communicate in order to confirm the other=s cooperative attitude. originally framed by Merrill Flood and Melvin Dresher working at the RAND Corporation in 1950 but formalized and named by Albert W. If both countries agree to cooperate. the betrayer goes free and the Counterterrorism Cooperation & Game Theory J. Breinholt May 2006 -5- . The prisoners are separated. Tucker. However. only to find that the other player is loaded for hare? You would be out of luck. A better game in which to discuss international terrorism cooperation might be the Prisoner=s Dilemma. If you are sufficiently risk averse or cynical about human nature. and that this decision must be made in the context of the Stag Hunt game.and two players whose goal is to minimize it for themselves. Two suspects. and would spend a day without any payout. However. their pooling of intelligence may allow them to connect the various pieces of a developing terrorist plot so that the one county=s law enforcement can move in to prevent it. as there are very few situations in which communication necessary for cooperation is not possible.

If each betrays the other. the outcome when both confess is worse for each than the outcome from remaining silent. 8 The equilibrium is established through a calculation that assumes that the other suspect is equally likely to cooperate with the authorities as to stay silent. A group whose members pursue rational self-interest may all end up worse off than a group whose members act contrary to rational self-interest. The question posed by the dilemma poses: What will happen? How will the prisoners act? The decision matrix looks like this: Remain Silent Remain Silent Cooperate Prisoner 1: 6 months Prisoner 2: 6 months Prisoner 1: freedom Prisoner 2: 10 years Cooperate Prisoner 1: 10 years Prisoner 2: freedom Prisoner 1: 2 year Prisoner 2: 2 years The dilemma arises when one assumes that both prisoners only care about minimizing their own jail terms. However. However. a rational player should calculate the likely payout by multiplying the likelihood the consequences of a particular decision by the value of the payout. the rational player would calculate his strategy (either cooperate or remain silent) as follows: Other Player=s Choice Chance Remain Silent . the police can sentence both prisoners to a misdemeanor charge and six months in prison each.org/wiki/Prisoner's_dilemma#References.wikipedia. If the other player=s choice is a 50-50 proposition. Breinholt May 2006 -6- . Even if they were able to talk to each other. Each prisoner must make the choice of whether to betray the other or to remain silent. neither prisoner knows for sure what choice the other prisoner will make. If that happens.silent accomplice receives the full 10-year sentence. Whatever the other does. If both stay silent. Counterterrorism Cooperation & Game Theory J.en. The outcome of each choice depends on the choice of the accomplice and neither prisoner knows the choice of his accomplice. neither could be sure of the other=s motives.50 Payout Cooperate: freedom Silent: 6 months Cooperate: 2 years Silent: 10 years Likely Payout freedom 3 months 1 year 5 years Cooperate 8 . If the payoffs are not assumed to represent self-interest. they will receive a two-year sentence.50 See www. each is better off confessing than remaining silent. a group whose members rationally pursue any goals may all meet less success than if they had not rationally pursued their goals individually.

necessitated by the rules of the game that prohibit communication. the rational player will choose to cooperate. This is not something that occurs frequently in the real world. we need to add a couple. Now let=s get very specific. Application of the Game: Building a Structure For Optimal Equilibrium The Stag Hunt and the Prisoner=s Dilemma each illustrate a situation where there is no communication permitted between the players. this is referred to as the Aextensive form@ game model. since this would mean freedom for both. Thus. silence. on the rational player=s assessment that the other player choice is a 50/50 proposition. since the sum of the expected payout from this choice is a 12 month sentence ( Afreedom@ plus one year). its neighbor realizes that it has intelligence which. each player would choose to be outfitted with stag-hunting equipment. In the real world. Say that the country of South has announced the indictment of a group of individuals involved in a credit card fraud scheme. if communication were allowed in the Stag Hunt. Going back to the set of assumptions described for the international counterterrorism cooperation. This choice depends. Breinholt May 2006 -7- . Although South=s intelligence suggests that the credit card scheme is part of a terrorist financing operation. What is not publicized is the fact that South has additional intelligence on these defendants. and allows them to reach a joint decision that is the best for each. Say that the two countries are contiguous. and split $ 8 dollars of meat. This is an estimate. Similarly. there is not enough Counterterrorism Cooperation & Game Theory J. Moreover. each game ignores the iterative nature of communication. As a result.something that would occur if the players were not separated and could communicate freely with each other .. whereas the sum of the likely payout from the alternative choice. of course. if shared. the intelligence each country collects domestically is likely to be have significance to the other. that citizens from one country often visit the other. and talk about countries that qualify for the series of assumptions we have posited. and that the two countries often play host to the same visitors. What if the player in the Prisoner=s Dilemma could eliminate uncertainty about what the other player will do? If there is no uncertainty . where players make a series of moves that are informed by the other player=s choices. will create some counterterrorism options that would not otherwise exist. If communication is permitted. communication would allow the players to beat the system in the Prisoner=s Dilemma. assuming rationality on each side? III. how would it play out. In game theory. Assume further that. where players have an opportunity to assess and recalibrate their strategy over the course of repeated interactions. and that this news is widely publicized. this is called negotiation. is 63 months (five years plus three months).they would presumably reach the joint decision to each remain silent. The ability of the players to communicate eliminates the uncertainty. developed though its sensitive sources and methods.Under this scenario. due to the actions of one country.

the Southerners need a boost. -8- Counterterrorism Cooperation & Game Theory J.C.S. at least up to a point. North: No. who realize that she has information that suggests that some of the South defendants are receiving directions from a North-based Al Qaida operative. South does not obtain the benefit of the Northern intelligence.is among its most guarded national security secrets.intelligence yet to justify seeking charges on the Southern crime of Aproviding material support. Unlike the Stag Hunt and Prisoner=s Dilemma. however. is not ideal. What if the South=s information about the North=s defendants. based on the fragmentary information collected though its own human sources. As a result.from human sources and from electronic surveillance . This equilibrium. No further elaboration is given. Nothing ventured. Each country breaks even (except for the de minimis cost of a long-distance phone call). when combined with the more fragmentary North intelligence. 2339B.@ 9 For this. there would be an equilibrium with no payout on either side. could have been used 9 See 18 U. Northern intelligence telephones the South=s intelligence service and says that is has classified intelligence related to the individuals recently charged with credit card fraud. This means that the scenario will take the extensive form. This is something that South suspects. it does not part with this information lightly. In South. If the dialogue ended at that point. let=s recognize the reality that the two countries can communicate. Advised of the news from North. '' 2339A. who in turn is speaking by phone to Al Qaida leadership in Afghanistan. News of the credit card fraud prosecution reaches personnel within the country of North=s intelligence service. to share with our prosecutors. _______________ That is easy enough. the intelligence collectors have not received the benefit of a consumer for which they yearn. the first exchange starts with a phone call from South. How the North intelligence service has obtained this information . nothing lost. Breinholt May 2006 . _______________ South: We want your intelligence.

a yes or no answer. the Southerners have better diplomats eager to get into the same. Let=s assume that South=s decision will not be informed by the preferences of Northern intelligence.: Don=t Use It Scenario 2 Scenario 3 North Don=t Share Scenario 1 Scenario 1 In Scenario 1. Let=s say that. probably the worst situation. truly sad state of affairs. Placed within the Stag Hunt or Prisoner Dilemma analysis. This means that North=s reaction is limited to a binary choice.to thwart a particular terrorist plot. To avoid this. if they are fortunate enough to receive it. unrealistically) that the Southern negotiators are not authorized to give assurance how the Southern prosecutors will use the Northern intelligence. or not. Future historians will have a heyday. They will likely become professional witnesses. optically. two allied contiguous nations committed to fight international terrorism. Breinholt May 2006 -9- . If this occurs. North uses it in criminal prosecution. The intelligence shared. The intelligence is shared. Scenario 1 is. and that North has no way to control how the South will use this Northern intelligence once it is disseminated. in hopes that the Northen intelligence officials to reconsider their rejection. North does not use it. there is no payout or costs for either player (other than the costs which would be assessed if the failure to share information is later determined to be part of a failure to prevent a particular terrorist attack).: Use It South. Let us assume (again. with hundreds of innocent civilians dead in both South and North? Personnel involved in the decision to not share intelligence that could have disrupted the attack will have quite a bit of explaining to do. these scenarios translate into this 2 x 2 matrix: North: Share Intelligence South. South and North. cannot reach an agreement to share information that is of joint interest to each of them. International negotiations begin. the countries should try to negotiate conditions to avoid Scenario 1. there are four possible scenarios arising from this point: Scenario 1: Scenario 2: Scenario 3: South persists in saying no. Counterterrorism Cooperation & Game Theory J. However. With two players each facing a binary choice. before various commissions and legislative bodies in both countries. South will be faced with a unilateral choice: either use the intelligence to supersede the criminal charges. when the initial request is rejected. Perhaps the diplomats repeat the South=s President=s statement that you are either for or against us. What if the plot results in a sensational attack.

other than the benefit to its intelligence service of having found a consumer who considered their intelligence without actually making use of it. another benefit. because there was no mechanism for obtaining additional assistance from North. For example. Once the intelligence was shared. Because North was not (by the assumed conditions of the game) involved in the decision of how and whether the Southern prosecutors would use their information in judicial proceedings. This means that the cost to the Northerners is small. The discussion of Scenarios 2 and 3 suggest additional conditions that would maximize the prospect of an optimal resolution: a third round of play in which the players can negotiate further conditions that will justify their decision to cooperate. What about the benefits to North? There are none. the Northern intelligence may have been shared in raw form. making it hearsay and not properly authenticatable under the Southern evidentiary rules. and permit the Southern prosecutors to obtain additional assistance from their Northern colleagues so that the Counterterrorism Cooperation & Game Theory J. North held its breath. there would undoubtedly be frustration at possessing intelligence it could not use. a prospect that may involve cross-examination of the witness and risk unpalatable amount of disclosure of national security secrets. Scenario 2 earns in a payout of intelligence that it uses in a criminal prosecution.Consider Scenario 2. where South makes the unilateral decision to use the information. although there might not be a cost. Given the foregoing assumptions and the state of play and the unknown factors that prevent a full analysis of Scenarios 2 and 3. It may arise from the fact that there was no mechanism for obtaining North=s assistance to place the intelligence into a form that would be admissible in a Southern judicial proceeding. and the goodwill it engendered with its southern neighbor. However. which benefits North. Whether Scenario 2 is a good situation for North will depend on whether the benefits exceeded the costs. because the South=s decision to use it is made without regard to this North=s concerns or equities. if one does not count the damage to the future willingness of North to provide this type of intelligence in light of the resulting damage to its sources and methods. Now consider Scenario 3. For South. the conditions should permit North to maintain a certain amount of control over how the Southern prosecutors use the Northern intelligence even after it is disseminated. What is also unclear is what went into South=s decision not to use the intelligence. and any drying up of sources and methods would be limited to that portion which it would otherwise have treated as NOFORN (unclear from these facts). the cost to the North of this decision was likely the permanent loss of the sources and methods that gave rise to this intelligence. What are the respective costs and payouts to each side? The Northern government has assisted in a Southern terrorism case. The Northern intelligence service has found a worthy consumer for its products. it had no way to impact the bottom-line assessment of Scenario 2. For South. The Northern intelligence is shared but it is not used by South. since the sharing of the intelligence did not result in its publication. Breinholt May 2006 -10- . There is not much of a cost. unless the Northern intelligence service was willing to provide a Northern witness. not knowing what the ultimate costs would be because it would not be involved in the decision by South would use the information.

they will explain to you how they plan to present it in court. They will also try to minimize the uncertainty in what the court will require to authenticate and admit that information. by seeking an advance court ruling on its admissibility. How about this? You give us this information. By the time of the trial. But what if these plans break down when it comes to the trial? Let=s say that any witness we provide to help you introduce this intelligence is subject to cross-examination? How will we protect against the prospect that our witness may be forced by an Southern judge to testify about Northern state secrets? Counterterrorism Cooperation & Game Theory J. If we fail to cooperate and terrorists strike. to share with our prosecutors. How adamant are you about not giving us that intelligence? After all. but your use of it will be contingent on our ability to prevent you from using it if and when it becomes clear that such use will involve an unpalatable amount of disclosure of our intelligence collection capabilities. Breinholt May 2006 North: South: North: South: North: -11- . How will that work? We will let you examine the intelligence. O. The negotiation would look like this: _______________ South: North: South: We want your intelligence. O.K. we will all be clear on what is required. Let=s discuss this further. each of our nations face a common enemy. Our prosecutors will review it.our sources and methods .K. If they decide that it justifies the addition of terrorism charges to our indictment.intelligence can be effectively exploited in a judicial proceeding. They will walk your lawyers through their specific plans and strategy. the media and historians on both sides of the border will be dancing on our graves. in the fight against international terrorism. you=re either with us or against us. but your prosecutors can=t use it unless you convince us that it will be done in such a way that our national security interests . How about this? You can see this information. and this is an important opportunity.will be maintained. O. No.K. Our President says that.

we have used game theory to construct the optimal infrastructure to reach the most mutually beneficial equilibrium. What little risk remains is offset by the ability of North to make the unilateral Counterterrorism Cooperation & Game Theory J. Does that work for you? Voila. I think we have a deal. Breinholt May 2006 -12- . in your judgment. ORCON assures that Scenario 3 will be reached only if it is more mutually beneficial than Scenario 2. since this depends on factors that are beyond the control of the players at the time of the agreement. we avoid Scenario 1 entirely.South: How about this? Even after the start of the trial. and considering the rational moves by each. This way. you maintain the right to control that intelligence consistent with your national security interests. including the right to force us to pull that information from the criminal proceeding if it appears. By changing the dynamic to allow for communication in the context of the sharing of counterterrorism intelligence between two allied countries. From there. which is based on North=s assessment of the costs (where any uncertainty has been minimized by pre-trial litigation). _______________ North: This is the equilibrium that is reached through negotiation. The optimal equilibrium will be reach in this situation if South is willing to grant ORCON assurances to North as a condition of examining its intelligence. This is close to an existing intelligence concept known as Aoriginator-controlled@ information (ORCON). Here is why: if South grants ORCON assurances. Intelligence labeled in this way cannot be used or disseminated without the consent of the originator. South) willingness to concede to the providing country (North) the unfettered right to maintain control over the requesting country=s use of the intelligence. The intelligence is shared. even if our prosecutors have to suffer the most extreme judicial sanction the dismissal of the prosecution. However. we get to review and possibly use the intelligence you have collected. something that is not permitted by the classical formulation of the Stag Hunt and the Prisoner=s Dilemma. whether the ultimate resolution will be Scenario 2 or Scenario 3 is not known. Getting to Scenario 3 requires a willingness of North to give permission for South to use the information in the judicial proceeding. There was a key concession that led to this settlement: the requesting country=s (here. that our court is permitting an unpalatable amount of disclosure. as a condition of it being permitted to review it in the first instance. and we each gain the joint benefit of fighting terrorism through bilateral cooperation. you will maintain an unfettered right to prevent us from introducing your intelligence. We will honor our commitment to you.

S. This means that it can exercise its ORCON rights at its own discretion. v. Ahmed Ressam. 10 Although the Ressam case received more attention. Ressam.Supp. Breinholt May 2006 . was ultimately convicted of an Al Qaida plot to blow up Los Angeles International Airport. We have established a set of rationally-negotiated conditions in which everyone wins. when the United States was on a heightened state of alert. 2002) -13- Counterterrorism Cooperation & Game Theory J. Reality: The Charlotte Hizballah Case Shortly before the turn of the Millennium.S.Wash.2d 1252 (W. 10 U.-Canadian in Washington saw something strange about a foreign traveler seeking to enter the U. The negotiations assure this result. and the officer uncovered explosive materials in the trunk of the traveler=s car. on an automobile ferry at the northwest border. there another case on the other side of the country that was perhaps even more significant. If Scenario 3 is reached .the best one for South .S. It involved a group of Hizballah operatives in North Carolina. an American customs official in at the U.it will also necessarily be the best one for North. an Algerian who had spent time in Canada. The traveler=s attempt to physically escape the officer=s questioning was unsuccessful. 221 F. The group was led by Mohamad Youssef Hammoud. The result was aided by assistance from the Canadian government. and came to the attention of the authorities because of an interstate cigarette smuggling operation uncovered by an off-duty sheriff.D. at least in terms of the long-term implications for international cooperation. IV.decision to withdraw consent to use the intelligence. except the terrorists.

Evid. Harb testified that Hammoud had declined to become involved in providing equipment . and carried out by an operative trained by Iran's Revolutionary Guard. 12 The prosecutors also introduced summaries and analysis of conversations captured electronically through surveillance conducted by the Canadian Security Intelligence Service (CSIS). at 327. Hammoud gave him $3. while the North Carolina tax was only 504. 381 F. However. money laundering.Hammoud and his associates would buy large quantities of cigarettes in North Carolina. see Fed. depriving the state of Michigan of $3 million in tax revenues. A number of these CSIS recordings were destroyed pursuant to routine procedures.R. Breinholt May 2006 -14- . 326-7 (4th Cir.which occurred in Canada 0 because he was helping Hizballah in his own way but that. 11 It became the first American terrorist financing case to ever go to trial. see id. At Hammoud=s trial. credit card fraud. which could be used for both civilian and military activities. where he would urge the attendees to donate money to Hizballah. mail fraud. They took advantage of the fact that Michigan imposes a tax of $7. Hammoud was ultimately charged with various immigration violations. when Harb traveled to Lebanon in September 1999. however. The CSIS information showed that Hizballah wired tens of thousands of dollars from Lebanon to Canada for the purchase of the dual use equipment. 2004). Later in the procurement efforts. 13 The CSIS wiretaps showed a procurement operation involving Harb was overseen by Hizballah's Chief of Procurement from Lebanon. During this period. Hammoud would lead weekly prayer services for Shi'a Muslims in Charlotte at his home. Hassan Laqis. and pay fifty cents 11 12 13 United States v. to describe the cigarette smuggling operation and Harb=s efforts to assist Hizballah in obtaining Adual use@ equipment. at 335. and conspiracy to provide material support to Hizballah. At the trial. Hammoud. During pretrial proceedings. the district court ruled that the CSIS summaries were admissible as recorded recollections. the prosecutors introduced the factual portions of some of these summaries (the analysis was redacted from the summaries before submission to the jury). sale of contraband cigarettes. At trial. In addition to the RICO charges. the plotters transported cigarettes valued at roughly $7.50 per carton of cigarettes. and sell them without paying Michigan taxes. Hammoud would then forward the money to Hizballah leaders in Beirut. the prosecutors called a childhood friend of Hammoud named Said Harb. 803(5). Before their arrest on federal racketeering charges. such as global positioning systems. smuggle them to Michigan. summaries and analysis of the conversations were prepared by a CSIS communications analyst shortly after each conversation was recorded. and as public records. Hammoud stipulated to the admissibility of the summaries. Counterterrorism Cooperation & Game Theory J.5 million. Mohamad Dbouk. Id. Id. Hizballah entered into an agreement with its operatives in Canada to purchase equipment with fraudulent credit cards. Rule 803(8).500 for Hizballah.3d 316.

17 16 15 14 Id. the negotiations that started as very formal but were ultimately transported to various Ottawa taverns. What went into that decision? The public answer to that question comes from a book about the Charlotte Hizballah case. Scott Broyles and Martha Rubio.could be useful in convincing Harb that he had no choice but to cooperate.S. electronic surveillance . Lightning Out of Lebanon: Hezbollah Terrorist on American Soil (Ballantine 2005). and Martha Rubio. where the two sides of cops and spies bonded over drinks and cigarettes. Breinholt May 2006 -15- .never before used in any judicial proceeding. an American who had traveled to Vancouver from Charlotte. In 1999. Robert Conrad. it learned that the Hizballah dual use procurement involved Said Harb.-based Hizballah cell.usdoj. Canadian intelligence specifically. It notified the FBI. ASmokescreen for Terrorism. Counterterrorism Cooperation & Game Theory J. Hammoud directly involved the dynamic discussed above. AAn intricate dance thus commenced in the spring of 1999.Canadian negotiations essentially mimicked the negotiations between North and South after the ban on communication in the Stag Hunt and Prisoner Dilemma games were lifted.on the dollar for all items procured.@ 16 The time pressures were driven by the approaching trial date for Hammoud on the cigarette smuggling charges.offered American counterterrorism officials insight into the operations of a U.pdf) Tom Diaz and Barbara Newman. January 2004 (available at: http://www. who was charged with racketeering but not yet terrorist financing. The FBI knew that Harb was tied to Hammoud. at 214.gov/usao/eousa/foia_reading_room/usab5201. 15 It turns out that the U. As Diaz and Newman describe it. the dance turned into a race against time. According to the book. at 207. By the summer of 2000. A rolling American team of FBI investigators and Justice Department lawyers shuttled to Canada to court the Canadians. The American prosecutors realized that Harb would be a valuable prosecution witness if he could be convinced to cooperate against Hammoud. and that the CSIS electronic intercepts .S. Lightning Out of Lebanon by Tom Diaz and Barbara Newman. 17 See D. American or Canadian . 14 Thus. The author of this article was one of the Justice Department lawyers on these trips. which quickly realized the significance of the intelligence. The others were Ken Bell.@ USA BULLETIN. CSIS had wiretaps and other communications intercepts to monitor the activities within Canada of Hizballah operatives. This intelligence was so good that the Americans used it in the criminal prosecution of the cell leader. Id.

in your judgment. Harb received a three and a half year sentence Hammond ultimately received a sentence of over 150 years.The assurances that ultimately led to the deal was referred to as the Asilver bullet:@ in exchange for its granting American prosecutors the right to use the Canadian intelligence.S. which he is currently appealing. Institutionalizing the ASilver Bullet@ It is sometimes said that great cases make bad law. which they did prior to Harb=s February 2001 guilty plea. This way. This was certainly not true of the Charlotte Hizballah case. which represents an anecdotal example of two allied countries coming together in the efforts to fight a common threat. including the right to force us to pull that information from the criminal proceeding if it appears. the solution was Asilver bullet:@ American prosecutors assured CSIS lawyers that. We will honor our commitment to you. They reiterated that. 18 The U. no matter what happened. prosecutors promised to seek a pre-trial ruling from the court in North Carolina on precisely what information must be presented in order to successfully admit the Canadian intelligence under one of the hearsay exceptions under the Federal Rules of Evidence. this was the last promise made by the country seeking permission to use the intelligence. at 216. you maintain the right to control that 18 19 Id. What does it say about bilateral counterterrorism cooperation generally? Can these lessons from Charlotte be institutionalized? In Hammond. CSIS lawyers were assured that they could pull the plug and withdraw the Canadian intelligence from the case at the first suggestion that the presentation of the evidence would imperil Canada=s security. they would undertake certain steps to protect Canadian sources and methods. even if we have to suffer the most extreme judicial sanction . a concession that ultimately ices the deal: South: How about this? Even after the start of the trial. CSIS lawyers would maintain an unfettered right to pull back that intelligence if it appeared during the American judicial proceedings that Canadian national security was about to be compromised.the dismissal of the prosecution. we get to review and possibly use the intelligence you have collected. who was ultimately convicted. Harb was the testifying against Hammoud. Id at 217. -16- Counterterrorism Cooperation & Game Theory J. if they were permitted to use Canadian intelligence in their pursuit of an American criminal prosecution. and creative thinking about how intelligence can be transformed into evidence admissible in a criminal trial. Recall. in the hypothetical dialogue between North and South. Breinholt May 2006 . 19 V. that the court is permitting an unpalatable amount of disclosure. you will maintain an unfettered right to prevent us from introducing your intelligence. A year later.

and it will be deprived of engaging in that Counterterrorism Cooperation & Game Theory J. Breinholt May 2006 -17- . consider the requesting party. For the requesting party. since (1) the cost is a necessary condition to receiving the information. however. Does that work for you? The Asilver bullet@ is essentially the reaffirmation of Aoriginator controlled@ (ORCON) limits on dissemination. since there is a benefit of having pertinent knowledge.receiving information . What about the cost? The cost of granting ORCON assurances is whatever will be incurred in seeking the originator=s permission prior to using the information. To the originator. once we lift the ban on players’ communication that prevents a certain optimization in Stag Hunt and the Prisoner Dilemma and play the games forward. and the costs/benefit from two scenarios: (1) it refuses to grant ORCON assurances. the benefit to the originator will exceed the cost. Under ORCON. the requesting party receives the intelligence. is the second of these results . which makes a decision by the originator to share the information easy.intelligence consistent with your national security. In other words. It is the rational outcome. since it means that the originator has found a consumer. Now.not always superior to the first? It will always be better off receiving the intelligence. the originator of intelligence does not lose control over how it is used merely by disseminating it. Instead. the intelligence will not be forthcoming. In the first situation. the receiving entity agrees. sharing does not mean loss of control. from small (the recipient country is ultimately unable to use it) to large (the intelligence is used to obtain a significant counterterrorism result). even if its value is limited by ORCON controls. to be bound by the originator=s conditions on use. it will remain ignorant of the intelligence. one of the assumed goals of its intelligence apparatus. As shown in Part III. The originator gives up nothing. and (2) it agrees to grant ORCON assurances. with ORCON assurances. The size of the benefit will vary. consider the cost/benefits to each side of the negotiations. the intelligence can be shelved. as a condition of gaining access to the intelligence. In general. and we each gain the joint benefit of fighting terrorism through bilateral cooperation. the cost of dissemination subject to ORCON is close to zero. and to not disseminate it to any third-party absent the originator=s permission. This means that any positive benefit will make that choice worthwile. the negotiation of ORCON conditions will result in the most beneficial equilibrium. This cost. without which there will be no benefit and (2) if the costs of obtaining additional permission from the originator do not exceed the benefit of using the information. but it is subject to originator controls. will always be exceeded by the benefit. since it maintains control over whether and how the intelligence is used by the receiving party even after it is disclosed. There will always be some benefit to sharing intelligence. To illustrate. as opposed to maintaining the unilateral right to decide whether to use the intelligence that is shared. if we agree that the originator=s decision to share the intelligence will depend on ORCON assurances. In the second. If the requesting party refuses to grant ORCON assurances to the originator.

Fortunately. allies are transformed into adversaries. Bilateral counterterrorism cooperation is not a zero-sum game. Breinholt May 2006 . Hence. they illustrate the dangers we simply cannot afford. government in their negotiations with other country=s intelligence services? Binding the U. 20 When they are applied to the situation of international cooperation by allied nations against terrorism. such as in most sporting events. they are able to communicate. the equilibrium will be the result of negotiation. prosecutors. in decision to grant immunity or enter into plea agreement. the adversarial relationship will result in a failure to connect the dots or solve the puzzle. In this situation. unlike the players in the Prisoners Dilemma. In Hammond. Where this is possible. at least if it involves countries with a history of trusting each other. government to certain legal positions is what federal prosecutors so everyday. and to eliminate uncertainty.S. and a mutual understanding of each countries discovery obligations and evidentiary rules that make operations decision truly a collaborative process.later cost/benefit analysis. In a zero-sum game. This should be recognized as the ideal. In these cases. The negotiation depicted in this article shows the fallacy of viewing international counterterrorism cooperation and information-sharing as a binary choice. If each has a piece of a larger puzzle. with each side being rationally self-interested and aware of the other=s moves. a gain for one participant is always at the expense of another. Why should they not be permitted to grant ORCON assurances to foreign intelligence services? The implication of this analysis is that greater and more meaningful counterterrorism cooperation between allied nations is possible through a decision to institutionalize ORCON controls. Conclusion Game theory tools and the Stag Hunt and Prisoner=s Dilemma scenarios show the cost of uncertainty. the equilibrium cannot satisfy both players. These rules can lead to player decisions that mimic what would occur in a zero-sum game. the Asilver bullet@ assurances came from U. 20 -18- Counterterrorism Cooperation & Game Theory J. where the players are unable to communicate. In such a situation. the uncertainty results from the rules of the games. This analysis illustrates the benefits of institutionalize a practice that my not seem intuitive in the absence of game theory analysis: countries should be generous is giving ORCON assurances to their allies if it increases the sharing of intelligence. It is not as simple as A Azero sum game@ is one in which all outcomes involve a sum of all player's payoffs of zero. Should lawyers be authorized to bind the entire U. game theory illustrates a way out of this danger.S.S.

each side maintains its right to protect its sovereign equities while maximizing the opportunity for effective joint counterterrorism operations. Breinholt May 2006 . For the question of whether one country should share its terrorism-related intelligence with another. In the end. This ideal has an added benefit: the recipient country is not left alone when facing the mechanical challenged of using the information. but it remains subject to control by the originator. the hypothetical negotiations suggests an ideal equilibrium.locating and negotiating with the proper foundational witness. I think we have a deal. the ORCON assurances imply a mechanism for the originator=s involvement in the solution . a burden that will not be borne solely by the recipient.proclaiming that you are either with us or against us. Voila. -19- Counterterrorism Cooperation & Game Theory J. Intelligence is shared. which maintains a role in the recipient=s decision-making of how and whether the intelligence is used. if the intelligence may be potentially useful as evidence in a judicial proceeding. and that everyone should share information. Effective cooperation requires conditions to assure that each side is able to maintain its equities in the face of the actions of the other. For example.