INTRODUCTION

On 26 February 2012 official communiqué issued by the FARC-EP secretariat (leadership) surprised Colombia. It announced that the guerrilla group would give up kidnapping as one of its strategies of war. The FARC-EP, or Colombian Revolutionary Armed Forces - People’s Army, made that decision after almost twenty years of systematic kidnapping and prolonged captivity of hostages. It can be seen as an important prerequisite for the announcement of a general agreement for the termination of the Conflict and the construction of a stable and lasting peace that was signed by representatives of the Colombian government and the FARC-EP on August 26, 2012 in Havana, Cuba. After four years of negotiations both sides announced the final agreement on August 24, 2016. In that document, “the FARC-EP renounces its political project to take power by means of weapons, and agree to follow the rules of Colombian democracy to pursue its political objectives. The Government, for its part, renounces to impose on the rebels the punishments defined by the penal law for their politics crimes, related to other, and is trade by a set of sanctions that allows the members of the FARC-Ep to act legally in political” (^{Melo, 2016}, 1). However, in October of that year a referendum to ratify it lost at the polls. 50.2% of voters rejected FARC-EP final agreement while 49.8% voted in favor. Afterward, the government negotiation team and the FARC-EP signed a revised peace deal on November 24 and sent it to Congress for ratification which was achieved on November 29-30, 2016. At least 500 community leaders and social activists have been killed in diverse regions of Colombia after the signing of the peace agreement.

My guess is that the FARC-EP belatedly realized the strong effects of the political long-duration kidnapping on the preferences of its middle and low-ranking members, those who were in charge of handling hostages. This came as a result of the problems created by direct daily contact between FARC-EP operatives and hostages for long periods in a context of intense military pressure, leading to a widening gap between the preferences and beliefs of leaders and those of middle and lower ranking members of the organization.

Unlike Castillo and Balbinotto (2011) in which the organization of the FARC-EP is described and the effects of the kidnapping on its structure are shown, this paper is mainly based on two different interpretative approaches from economic theory. The first was taken from ^{Phillips and Pohl (2013)}, who saw the preference problem as a change in risk for operatives. The second approach came from ^{Shapiro (2013)}, who presented the problem as a divergence of preferences both underlying and induced. In short, the approaches of Phillips and Pohl (2013) and Shapiro (2013) discuss why some FARC-EP operatives’ interests may have deviated from those of leadership’s in the Armed Illegal Organization-henceforth AIO- in spite of combatants’ preferences were considered unchangeable. This deviation might be related to a shift in the operatives’ preferences due to changes in their relationship with the leadership, as its new decisions as the long duration kidnapping were not welcomed by some operatives. It might also be due to the emergence of external factors or new information related to the new living conditions imposed by the kidnapping, which could be perceived in a different way by some operatives. It is clear that if operatives had preferences that were essential to their membership in the AIO, any preference change as a result of a new setting, would involve a self-assessment of their role in the organization and could result in a decision against the leadership’s interests.

From 1970 until 2010, this group carried out all types of kidnappings in Colombia (^{Rubio, 2003}), and was directly responsible for 37% of all the cases reported during this period (Center of Historical Memory, 2013). In the 1990s, it seemed that the kidnapping meant not only economic gains but also that the organization was increasingly efficient in the capture and subsequent management of hostages. “The FARC-EP had made a huge business and an important political asset of the kidnapping of all types of citizens: wealthy or poor, common or V.I.P, Colombian or foreigner, civilian or military and police individuals” (^{Topel, 2009}, p. 30).

As noted above, I use the case of the political kidnapping committed by the FARC-EP¹ in the 2000s and its aftermath for that organization. My hypothesis is that this illegal organization’s decision to kidnap politicians and militaries in order to demand from the government the release of its prisoners from Colombian jails and a practical tool for effective dialogue (^{Rubio, 2003}), led to serious tensions between the leadership’s and lower ranking members’ preferences.

Perhaps the positive outcomes obtained by the FARC-EP in the past prompted the organization to change its strategy and decide to kidnap for political rather than economic reasons to advance its specific political ends. That meant long-lasting abductions because of the refusal of the Colombian government to negotiate with the armed group and increasing costs for the organization. In an interview with the FARC-EP commander in chief, Rodrigo Londoño, alias Timochenko, recognizes that the kidnapping was a wrong approach:

“It (kidnapping) was a method of financing which was justified at the time. We needed it to finance ourselves and we found that instrument that really was not the more humane. We attempted to correct it at the time of Belisario (Colombian president). But as the process did not develop as it was planned so we returned to the confrontation. But there was a moment that we said: that must stop” (Revista Semana, January 1^st 2016).

Data on the number of registered kidnappings include cases of individuals who remained in the Colombian jungle for up to thirteen years. In fact, the negative consequences for the FARC-EP of the transition from kidnapping for ransom to the kidnapping of political figures and agents in the service of the State and its long duration were evident for even the FARC-EP´s most seasoned member Martín Sombra². He stated:

“You have to recognize there was disintegration of Fronts due to the bad policy of the FARC-EP as kidnappings, killing innocent people, the ambition, the money. At this time there is a rethinking of the organization, but it is already difficult to correct it” (Revista Semana, March 1^st 2008).

In fact, one of the most visible consequences of the kidnapping was, first, that a high number of operatives deserted the organization during the same period of political kidnapping, -a phenomenon that had previously been sporadic and isolated-. Second, a high political cost, not only nationwide but also internationally, derived from the cruel captivity of hostages for many years in subhuman conditions and the death of various high-profile hostages-including the execution of several local parliamentary members-. Perhaps all these reasons together led the FARC-EP to announce the end of kidnapping.

This paper is a preliminary contribution to the studies of the impact of the leadership’s decisions in a hierarchical organization such as the FARC-EP over the performance of its some operatives. I use two different approaches from economic theory: i) A risk theory and ii) A preference theory to understand changes in the perception of risk, and divergent preference, both underlying and induced, respectively.

Given the difficulty of collecting reliable data on operative and leadership preferences within covert and illegal organizations, I relied on evidence coming from obtained evidence to develop my analysis from reports of people who were kidnapped by the FARC-EP for political purposes (Araújo, 2008; ^{Betancourt, 2010}; Gonsalves, Stansell and Howes, 2009; ^{López, 2011}; ^{Pinchao, 2009}; ^{Samper, 2013}) and their relationship with their FARC-EP jailers. These accounts gave us a sense of how the FARC-EP operatives experienced kidnappings.

WHAT ARE TERRORIST PREFERENCES?

There has been a recent emergence of a wide array of literature on terrorist preferences. Strategic models have been developed in which terrorist organizations are assumed to be unique operatives that make rational decisions (Abrahams, 2008), or predict how and why organizations such as Al Qaeda select targets within the US (Libick, Chalk and Sisson, 2007), and studies have sought to identify terrorists’ objectives (Keeney and Winterfeldt, 2009). ^{Borum (2004)} studies the behavior of terrorists from the perspective of psychology. Nanbaldov (2013) distinguishes between two types of terrorists: old terrorists (before the end of the Cold War), whose behavior can be explained by the rational choice theory, and new ones (after the end of the Cold War), who represent a substantial departure from rationality. In the same line of research, ^{van Um (2009)} discusses concepts of terrorist rationality and of political rationality for the analysis of terrorism.

Frey and Luechinger (2003) use the rational choice approach to devise a wider set of anti-terrorism policies. They lay out an alternative view to the conventional idea that the incentive to undertake terrorist acts is lower when the expected punishment is greater. Based on the assumption that terrorists are rational operatives, their contribution shows that an effective way to deter terrorism is to raise its opportunity costs.

Although I assume the behavior of the operatives as politically rational (^{van Um, 2009}), my focus is on the study of illegal organizations made up of a leadership and operatives, and specifically in studying operatives as individuals who have different goals from those of their leaders (^{Shapiro, 2007}; 2008; 2012; 2013; ^{Castillo and Balbinotto, 2017}).

In covert organizations with a strong hierarchical structure, such as terrorist groups, in which the leadership makes decisions without prior consultation and operatives accept them without discussion from a restricted set of alternatives, are typically divided by differences of opinion about how to conduct war (^{Shapiro, 2008}). In this setting, the secret and illegal nature of an organization can deepen the problem of interest divergence between the organization leadership and its operatives, enabling the latter to opportunistically take advantage of the situation to act as they prefer, rather than as their leaders would like (Shapiro, 2013, p. 26). In response to this situation, leaders may have well-developed mechanisms or institutions to alleviate the effects on their operatives. For example, leaders often monitor operatives. However, doing so can be costly beyond a certain threshold because it increases the risk of negatively affecting everyone in the group.

To illustrate this concept, I analyze the case of the FARC-EP and the behavior of some of its operatives. This organization faced a high desertion rate of operatives during 2002-2010-data never before reported (Table 1)-after having been characterized by strong internal cohesion and shared values (^{Matta, 1999}; ^{Offstein, 2003}; ^{Pécaut and González, 1997}). Table 1 displays the demobilizations suffered by the FARC-EP from 2002-2009, right at the time it had the highest number of hostages. In terms of military rank and position within the organizations, PAHD (the Program for Humanitarian Attention to the Demobilized) uses five categories to classify those who are already demobilized. In the private category were those members directly involved in the custody of hostages, and in the militia category were those responsible for bringing supplies to the hostage camp.

Table 1 Positions within the FARC-EP 

Source: Arias, Herrera and Prieto (2010)

For many years, the FARC-EP was able to control and maintain the discipline of the group despite occasional cases of desertion that did not affect the organization’s structure. The FARC-EP considered its operatives uniformly motivated and willing to sacrifice for the cause and believed that they would perform any task they were assigned without additional compensation.

These quotes nicely illustrate how control over the FARC-EP fronts was delivered:

“The FARC-EP’s Secretariat (leadership) maintains a tight control over the Frentes (Fronts) and over individual operatives who are in charge of financial affairs. The rationale behind this is evident: The only relatively serious splits that the FARC-EP has suffered in its long history come from people who have abandoned the organization with a handful of dollars. Typically, their following has been from tiny to negligible. FARC-EP leaders are highly aware that a luxurious life style and the enjoyment of pantagruelic (sic) rents can not only undermine the organization’s cherished unity, but also slacken its combativeness. Thus, strong bureaucratic and normative constraints are imposed over the militants, especially those who are more exposed to temptation. But this brings us to the general frame of the institutions developed by the organization, that constitute the immediate set of incentives and constraints for its members” (Gutiérrez, 1999, p. 11)

The following quote addresses control over operatives’ lives and its impact over hostages:

“The FARC-EP had control over every aspect of the guerrillas’ lives -including what passed for romantic relationships. Though we saw a lot of promiscuity and swapping of mates...” “…Because they had so little command of their own lives and made so few choices for themselves, we were just about the only things that they could actually control. Even though they were never able to control us completely, the need to assert themselves over us had a lot to do with their cruel and arbitrary treatment. Knowing this didn’t justify their actions, of course, but it did help explain them...” (Gonsalves, Stansell and Howes, 2009, p. 278).

The rapidly growing desertion rate during the 2000s began to indicate that some operatives were unhappy with the new tasks related to the new political goals of the FARC-EP’s leadership-i.e., political kidnappings and their long duration. The FARC-EP leadership’s theory was that once operatives decided to enter the organization, their identity as individual operatives would be replaced by their identity with the group. Consequently, the FARC-EP leadership thought that they could make any decisions and mandate any types of tasks. They also assumed that the initial motivations and risks associated with membership in the organization would remain constant over time. It seems at the FARC-EP perhaps never thought that the interaction of motivations, risks and rewards would result in some operatives reaching a risk threshold beyond which they would not go as deserting from that AIO. The FARC-EP should have foreseen this possibility.

For a long time, as ^{McCormick (2003)} stated concerning some terrorist groups, the FARC-EP’s leadership did not consider the potentially discordant views of the world but assumed there would be a unitary operative group, defined by a single, stable, and ordered set of preferences, which would be able, with a single mind, to identify, evaluate, and make decisions among competing options. The FARC-EP’s leader assumed that her role was to identify the goals and operating constraints, to assign tasks and then pick the available courses of action that offered the highest expected returns. The operatives’ function would be to accomplish the tasks in the way most favorable to the goals of the FARC-EP leadership.

I am not focused on hostage-taking or kidnapping scenarios in which armed organizations attempt to negotiate with the government to obtain concessions, as presented by ^{Sandler, Tschirhart and Cauley (1983)}, although the effects of failed negotiations between government and an AIO are incorporated into this analysis. Rather, I focus primarily on the reasons why operatives’ preferences eventually departed from the leadership’s interests and the simple causality between the effects of the leadership’s decisions and the high desertion rate of some operatives.

I will outline two different ways of addressing this problem. The first is taken from ^{Phillips and Pohl (2013)}, who see the preference problem as a change in risk for the operatives. The second approach comes from ^{Shapiro (2013)}, who presents the problem as a divergence of preferences both underlying and induced. Although that divergence is not uncommon for organizations that suffer these types of leadership-operative conflicts, in the case of the FARC-EP, the organization was not able to handle them in a timely manner. Its leadership believed that political kidnapping would be a successful action similar to the kidnapping of people for ransom. They thought the political kidnapping strategy would have a relatively higher payoff than its past other criminal actions. The expected benefits included gaining the support of the media, more recognition of its political status, release of its incarcerated members, a demilitarized zone and changes in government policy.

However, the costs for the FARC-EP’s operatives from the kidnappings were significantly greater than the gains. The agency problems for the organization, reflected by an increasing desertion rate, were capitalized upon by the government, which increased military pressure and refused to fully engage in negotiation with the FARC-EP. Through these tactics, the government gained significant political advantage. The end result showed that the FARC-EP had failed in its political kidnapping policy. Two hostages give their views on how the FARC-EP operatives, also known as guerrillas, think about kidnapping:

“Even the guerrillas said, many times, that basically they did not agree with the practice of kidnapping, and they did not share the fact that we were suffering in this situation. Among other things, because kidnapping was contrary to the postulates of Jacobo Arenas, the ideologue of the FARC-EP, who wrote several articles for the FARC-EP’s booklets, in which he manifested his total repudiation of kidnapping as a political practice, and more as a way of financing the war” (Pérez, 2009, p. 178).

“…There was never the potential mass mutiny, but here were several occasions when the guerrilla opened up and said something that explicitly revealed the level of discontent among the ranks… But as far as we could tell, there was widespread questioning of purpose and a dislike for this duty. On the forty-day jaunt after Caribe, we saw that the guerrillas didn’t like the forced marches any more that we did. Now that seed of discontent had blossomed” (Gonsalves, Stansell and Howes, 2009, p. 258).

According to the hostages, it is apparent that some commanders (operatives) disagreed with the practice of kidnapping. These perceptual differences between the leadership and some operatives as to the rationale, process and procedures undoubtedly increased with the passage of time. Martín Sombra, one of the main commanders, discussed his difficulties in taking care of hostages, among them a baby, the son of one of the hostages, who was born in captivity.

“You can imagine: The child crying, some hostages carrying them on a hammock because they were ill, others fucking shit up and military helicopters flying over us... that was very hard to take care of the prisoners!” (Revista Semana, March 1^st, 2008)

As this statement illustrates, during captivity, operatives constantly faced high costs and risks that exceeded the benefits of being in the organization. Although the members had, after all, been recruited to take huge risks and commit dangerous acts to achieve the organization’s goals, long periods of contact with hostages that involved putting the hostages’ and their own lives at risk while awaiting government negotiations could be dispiriting, tested the operatives’ resilience.

DATA ON THE FARC-EP’S POLITICAL KIDNAPPINGS

According to the statistics of Fundación País Libre (Free Country Foundation), in the years between 1996 and 2010, the FARC-EP held 693 abducted persons, including people kidnapped for monetary ransom and others for political reasons. One of the most prominent kidnappings by the group took place on 9 June 2001, when 41 people were kidnapped in the urban area of Valledupar, in the northern Colombian province of Cesar. Additionally, that same month, 17 people were forced to leave the building where they lived, in the morning early hours, in the city of Neiva, Huila, in the mid-west of the country. On 11 April 2002, 11 deputies from Valle’s Assembly were taken from the place where they had their sessions and driven by bus to the Colombian jungle. The Assembly´s chambers were located in downtown Cali, the third most important city in Colombia.

To isolate the effects of political kidnappings from economic ones, this paper focuses on the politicians and high-ranking military hostages who were held by the FARC-EP during 1997-2010 as bargaining chips for the explicit purpose of forcing the government to accept a demilitarized zone for peace talks and some type of prisoner exchange (Table 2).

Table 2 Data on the “exchangeable” list by type of hostage 

Source: ^{Castillo (2014)}.

However, the decision to engage in kidnapping did not produce the expected results. Only in 2001 did the FARC-EP obtain the liberation of 15 guerrilla fighters in exchange for some kidnapped Army officers. Table 3 displays the final resolutions for people abducted by the FARC-EP over the course of almost fourteen years of continuous imprisonment, as characterized by ^{Castillo and Balbinotto (2012)}.

Table 3 Final Resolution for Hostages 

Source: ^{Castillo (2014)}.

Examining how the kidnapping ended for the hostages, the results show that the costs to the FARC-EP exceeded the benefits. Table 3 shows only the year in which the kidnapping occurred and how it ended but no duration. After many costly outcomes-not only political but also social and financial, such as the death, killing, escape and rescue of some hostages-the FARC-EP had to give in and surrender the hostages without any type of prisoner exchange nor demilitarized zones, perhaps to put an end to the high costs that the kidnappings were generating.

RISKS

In behavior models the assumption of risk neutrality leaves us with one less concern. However, in the context of illegal organizations, the types of tasks imposed by the leader upon the operatives could shape the operatives’ risk preferences and exacerbate agency problems inside the organization.

More specifically in the case of the FARC-EP, ^{Gutiérrez (2008)} showed that when operatives were assigned combat tasks, extortive kidnapping and armed assaults-all activities with short-term results-the rate of desertion was low. However, when the FARC-EP turned to political kidnapping, the number of deserters coincidently grew. Records of this phenomenon date only from 2002 due in part to the Colombian government-led program of reintegration for guerrillas. However, if the operatives’ interests were aligned with the leader’s interests, why did some operatives change their minds?

I will use ^{Phillips and Pohl’s (2013)} analysis to give an analytical explanation for the FARC-EP’s decision to stop the kidnapping. Phillips and Pohl explore the ways in which concession or punishment schemes alter the risk-reward trade-off and how they affect the terrorists’ expected payoffs. Their analysis also identifies important relationships between risk preferences and the nature of concessions to, or punishments for, a terrorist organization by a government. Although this paper will focus on the operatives’ preferences shaped by the government’s refusal to negotiate, as did Phillips and Pohl (2013), it also includes the impact of mandatory tasks on operatives’ preferences under the leader-operative relationship.

It is assumed that the thrill of combat is the main motivation for those who join and remain in an AIO, based on ^{Gutiérrez’s study (2008)} of FARC-EP combatants. Gutiérrez describes the FARC-EP’s members as good fighters despite the harsh difficulties faced by those who choose to belong to the organization:

“This gives us a general picture of the organization-individual gap in the Colombia war. Take the FARC-EP, with its strong links to criminals. Its non-paid members (18-20,000) are participating in a conflict in which they have a fair probability of getting killed. They do not benefit from looting. Becoming rich is not a realistic perspective, and this is common knowledge. The organization severely intervenes in all the domains of their life. The FARC-EP’s time horizons are long, because, very wisely, it has refused to offer a more or less precise notion of when victory, or the termination of war, will arrive- its patience is proverbial, and a powerful tool in peace bargaining. This is “metaindividualistic” patience indeed, a life time might not be sufficient to attain the collective goals (however we describe them). No extraordinary income (or ordinary, for that), thus, no family life, and no credible expectation for escaping war. No ethnic or religious glue, either, nor a big doctrinarian build up. Despite this, The FARC-EP’s members generally fight with great verve. There are exceptions, but as a rule their behavior in combat exhibits both skill and motivations against opponents endowed with better technical means. When on the defensive, they do not fall apart, and only on the margins does the group suffer defections” (^{Gutiérrez, 2008} p.14).

Until recently, the FARC-EP appeared to have strong internal cohesion, with operatives’ interests seemed to be aligned with those of the leadership. However, the cohesion appears to have weakened with the adoption of political kidnappings, supporting my position that combat could be more attractive than other activities to the operatives. In this spirit, operatives joining the FARC-EP showed a preference ordering and inclination for combat over any criminal activity. As ^{Spanier (2015)} showed for terrorist organizations, in the case of our study, the combat may appear as a sensitive task that requires great preparation while guarding hostages may be comparable to a non-sensitive and demeaning task, which yields comparatively fewer gains and does not require an individual high effort. Indeed, the liberation of a hostage depended more on the negotiation between the Colombian Government and FARC-EP leaders than on the will, efforts and decision-making capacity of its operatives. It was a task for which success did not depend at all on operatives´ effort and effectiveness.

If one compares how operatives might assess combat and kidnapping, it is possible to imagine some differences between the two. The results of the kidnapping activity were uncertain and depended on the political will of the government, while the combat action depended more on individual efforts. In the context of kidnapping, a greater delay in government decision-making for an agreement, maintaining strong military pressure against the organization, implied higher costs for the operatives who were responsible for hostages. That is, the operatives faced more uncertainty when taking care of hostages than when they were mere combatants.

As time passed, the Army’s presence in the FARC-EP territories trying to rescue kidnapped people entailed more uncertainty and insecurity for guerrilla operatives, particularly those who were forced to move all the hostages constantly and urgently to different locations in the jungle and the countryside. Doing so put not only their lives at risk but also the lives of the hostages. While in combat, the operatives experienced other realities that depended upon their own efforts.

Indeed, the results of any clash between the FARC-EP’s operatives and the Army were immediate on the ground. Before each confrontation, operatives could estimate the payoff that they might obtain given their performance. Those payoffs could include inflicting fatalities on the enemy, gaining more territory, seizing weapons and ammunition from the enemy, or winning a promotion within the organization (Phillip and Pohl, 2013), or the simple and sweet thrill of combat. Although the expected payment scheme altered the risk-reward tradeoff in both cases, it is clear with respect to long-duration kidnapping-unlike with combat-that with time, the uncertainty for operatives grew and the likelihood of obtaining some reward decreased, partly due to the fact that the political kidnapping greatly increased the chances of being killed or captured for operatives. Because of the operatives’ responsibility for the custody of hostages, their potential for combat activity was reduced. In fact, operatives with combat skills who were responsible for the hostages were often forced to flee conflict zones to prevent putting hostages’ lives at serious risk.

^{Betancourt (2010)} describes one of several situations that she experienced in captivity, which shows the constant risk of being caught or killed by the Army for both hostages and abductors.

“At two o’clock in the morning, I was violently awoken by one of the guards shaking me and shouting… ‘Get up, bitch! Do you want to get killed?’…Military planes were flying very low over the camp. The guerrillas were grabbing their backpacks and running away, leaving everything behind them. The night was pitch black, you couldn’t see a thing except the silhouettes of the airplanes you could sense above the trees… They only made the guard bleat all the louder ‘Leave everything! They’re going to bomb us, don’t you get it?’” (^{Betancourt, 2010}, p. 141).

In another camp, one of the hostages relates:

“At first, I really didn’t feel anything. All of a sudden, we felt the Army’s helicopters fired several shots. They almost were flying over the treetops. -This way! Pick up what you can! Let’s run away! This way! -The guards shouted. All the hostages and guards went into a panic. The rain, as always, hit us day and night. The guerrillas were very afraid and we walked at excessive speeds. We did not stop even for a moment.” (^{Samper, 2013}, p. 17).

These forced escapes were costly given the loss of camps that had been built and the necessity of finding another secure location and building a new camp. Guards and hostages lived in constant fear of being attacked by the Army. For guards, indeed, the practice of kidnapping actually exacerbated preference divergence because operatives that guarded hostages, faced dramatically greater risk than leaders.

These types of situations undermined the operatives’ morale, making their relationships with hostages tenser and causing them to subject hostages to acts of cruelty. One of the hostages writes in his book:

“One has to get used to the chains, to be barefoot in the camps, not to use toilet paper and a thousand other things. But what affected me most was the humiliation, to which I never got accustomed” (^{López, 2011}, p. 42).

The emergence of these problems would likely change the payoffs that operatives expected from this type of activity. The enforcement of the new tasks might have affected how operatives saw the tradeoff between costs and payoffs. Only if the payoffs from the organization had been responsive to those changes-that is, if the leadership were willing to increase payoffs-would it have been possible to alter the operatives’ risk preferences in a manner that made them more risk seeking. That is, the operatives were willing to accept riskier actions involving more effort if the payoffs were commensurate with the increased risks, but in the FARC-EP the payoff adjustment is not immediate or it may never happen at all.

However, because there was little flexibility in the leader’s compensation system, the disutilities generated by the new tasks were not quickly mitigated by new compensation systems, which led to growing discontent among operatives. This is the perception of a hostage of his abductors:

“The face of the guerrillas in general is morose and melancholy. They are sad and their faces also reflect a tremendous amount of anger and hatred which has been accumulated for years. Overall, their expression is of dismay or disappointment. They don’t look like people who are involved in an enjoyable activity or job. On the contrary, most of them seem to be resigned. This is understandable since for many of them to be a FARC-EP’s member has been a goal in life, not a conviction but a unique way of life that promises three meals a day in a context of high risk, constant danger and yet, with the monotony and routine tediousness, and the unhealthy jungle” (Pérez, 2009, p. 176).

EFFECTS OF NEW TASKS ON THE BEHAVIOR OF OPERATIVES

As with any armed organization all of its operatives face some expected payoff scheme characterized by a trade-off between risk and reward for all the mandatory activities of the organization. I assume that in facing a new task different from combat imposed by the FARC-EP’s leadership, its operatives conducted a new assessment of the compensation package made so far by the organization. This evaluation depended upon several factors, such as operatives’ position in the AIO and the degree of risk aversion associated with the new activity. In the end, their assessment was reflected in how they performed their activity, that is, how much effort they dedicated to it. As ^{Phillips and Pohl (2013)} explained in the case of terrorists, in our case, it can be stated with some justification that some operatives must also have faced a convex compensation schedule that exhibited the following relation: the riskier the activity was, the greater the expected reward was.

However, could greater incentives encourage an operative to take more risks? According to Ross (2004), little is known about the derived risk preferences of operatives given common types of incentive structures. Therefore, I will take some first steps toward such an analysis by identifying conditions under which some operatives were willing to take less risky actions from their point of view, even if these actions conflicted with the organization’s objectives. For example, our interest is in showing why some operatives chose to desert rather than wait for the outcome in the kidnapping case.

I will assume a setting wherein illegal organization incentives tend to be inelastic in relation to the risks and in which incentives for different actions do not move in concert over time. Moreover, in the case of the FARC-EP, some mandatory tasks, such as the long guarding of hostages, led some operatives to make decisions that they considered less risky at the time such as leaving the organization. Perhaps they became aware that the government’s refusal to negotiate with the FARC-EP increased the risks and decreased the expected rewards of staying in the organization, particularly for those operatives who took care of hostages.

More specifically, what are the effects on the operative’s future decisions if there is a change in assigned tasks? How did the FARC-EP operatives evaluate those new tasks? Did the AIO know how its decisions could affect an operative’s risk preferences?

The problem of a shift in mandated tasks for operatives in an illegal setting must be approached by treating the effects of this change as altering the operatives’ preferences of staying in the organization. If a compensation plan for an individual operative entering the organization was characterized by a particular trade-off between risks and rewards related to specific tasks, than an alteration to such tasks could modify the expected compensation in response to riskier actions involved in the political kidnapping.

In the past, the FARC-EP had carried out a series of successful kidnappings in which they obtained what they demanded without additional costs to the organization. Adopting the political kidnapping strategy, the FARC-EP’s secretariat was betting on a high probability that the Colombian government would fulfill its demands because the kidnapping of high-profile people placed enormous pressure on it that no state could resist. For the FARC-EP, the political kidnapping had a higher expected payoff because it considered the hostages “exchange money” to achieve its political purposes. As a result of this pressure, it expected the Colombian government to provide it concessions, such as the release of its incarcerated members or a demilitarized zone, to alleviate the crisis produced by the hostage-taking. The FARC-EP secretariat expected that taking more risk would generate higher payoffs and more concessions from the government.

However, based on the premise that concessions encouraged future kidnappings, Uribe’s government was adamant about not negotiating. Figures 1 and 2 show the impact of the long period of captivity and the government’s no-concessions policy on the existing payoff system for some FARC-EP operatives. As Phillips and Pohl asserted in the case of terrorists, “the concessions schedule does not immediately add to or subtract from the terrorists’ wealth. The concessions are associated with risk terrorist actions and those actions must be perpetrated in order for concessions to become a possibility. Because the expected payoffs schedule to all terrorist actions are uncertain and subject to variability, undertaking a higher risk action does not guarantee a particular actual payoff or its associated concession” (p. 112).

In our study, as in the case of terrorists, the simplest way that government concessions may influence an organization’s payoff system is by adding to or subtracting from it.

Therefore, a concave schedule of concessions that lowers the expected payoffs to the organization past some point yields a negative addition. Figure 1 shows a no-concessions government policy that is negatively valued by the organization past some point of the expected payoff. This point T can be interpreted as a breaking point at which the existing cost of kidnapping exceeded the expected benefits of that activity. I refer to this type of cost as the number of men responsible for guarding many hostages and keeping them healthy and safe while the FARC-EP moved them constantly and urgently around the jungle and countryside to prevent localization by the Army, among other things.

Figure 1 

In Figure 2, the solid line plots the positive relation between expected payoffs and risks before the subtraction of a perceived concessions system. T corresponds to the point at which to negotiate with the government becomes very expensive for the organization and the expected payoff decreases and becomes negatively valued. This alteration is depicted by several dashed lines in Figure 2. For a time of captivity greater than T, the expected payoffs of kidnapping were lower per unit of risk than before.

Figure 2  

As proposed by Phillips and Pohl, we consider the costs incurred by the organization due to a long captivity as a negative concession that increased monotonically in the negative for expected payoff past a prudent period of waiting for an agreement. That is, as time went by, the likelihood of rescue or escape of hostages increased and risks for operatives also larger. By the government’s refusal to negotiate, costs attached to kidnapping become larger than the expected payoff. The effect of long-duration kidnapping on the existing system of payoffs received by some operatives was to reduce their expected payoffs. This effect would make some operatives less risk seeking and more likely to choose desertion, avoiding to be captured or killed by the Army.

Therefore, I assume that new tasks imposed by the leadership without any payoff adjustment made the operatives’ utility function very concave and made the operatives more risk-averse. Those tasks increased the operatives’ aversion to risk and made them less risk seeking because the new activities raised the risk level higher over a threshold they had initially accepted. Although operatives were willing to take risks, the riskiness of these new tasks exceeded the level of risk initially accepted by each of them.

Regarding kidnapping, as I make this analysis, I see a split develop between the risks taken by leaders and operatives. For the latter, the risks were not compatible with their payoffs. That is, it would seem that the payoffs were only commensurate with the risks originally assumed by them in these actions. While on the one hand the leaders waited for an agreement allowing the liberation of the hostages, on the other hand, the operatives might be waiting for a reward from their leaders to mitigate the risks associated with kidnapping. Unknowingly, the FARC-EP leaders assumed that their risk was comparable to that of the operatives; however, in reality, only the operatives faced the greater risk of direct attack by government troops.

UNDERLYING PREFERENCES VERSUS INDUCED PREFERENCES

^{Shapiro (2013)} offers an explanation as to why the preferences of leaders and operatives are not completely aligned. In terrorist organizations, operational terrorists often have different preferences for targets from those of their leaders. As a result, they have different perceptions about not only the political impact of their actions but also how to use violence and spend money. Despite high security costs, leaders generally mitigate these conflicts of interests by exercising greater control or punishing operatives who misbehave. Unlike legal organizations, covert groups can face greater problems when tasks are delegated. However, the benefits of delegation are evident in instances when operatives have better information or unique technical skills concerning a target.

In these cases, leaders may be better off delegating to an operative. The disadvantages of delegation are linked mainly to strategic decisions that require numerous sources of information (^{Bendor, Glazer and Hammond, 2001}; Bendor and Meirowitz, 2004; ^{Lupia, 2001}). In that setting, leaders know more about how to accomplish objectives than their operatives because they have more experience concerning the political impact of certain actions, and they know how to respond in light of the current political setting. To explain such preference divergence between leadership and operatives, ^{Shapiro (2013)} distinguishes between divergence in underlying preferences and divergence in induced preferences. In his research on the terrorist’s dilemma, Shapiro (2013) refers to preference divergence as a result of differences in induced preferences due to different underlying preferences, information, or beliefs (p. 29). Induced preferences are a function of underlying preferences, which are determined, first, by the information operatives receive and, second, by different beliefs about how to respond to the given information. Therefore, individuals who have similar underlying preferences can have very different induced preferences.

I agree with Shapiro’s explanation, which states that measuring preference divergence is more complicated than simply looking at observed levels of conflict. I see as significant the hostages’ reports that note tensions between kidnapped individuals and their jailers and how they were treated during captivity.

The US contractors discuss their marches:

“…Everyone had it bad, including the FARC-EP. Once again we saw the lower-level FARC-EP guerrillas being treated like pack animals. They carried heavy propane cylinders, cook stoves, and large bags of food. They ferried one load ahead, returned, and then set out again with another heavy load.” (Gonsalves, Stansell and Howes, 2009, p. 232).

“…Everyone did the best they could to help the others, but the FARC-EP were suffering as badly as we were and they took out their frustrations on us” (Gonsalves, Stansell and Howes, 2009, p. 233).

Or the presence of tensions between commandants and guards:

“This was just one of several instances we witnessed when the underlying tension between Milton (commandant) and the guards started to boil to the surface. There was a definite crack being exposed and we moved to exploit it as best we could. Like us, a number of the FARC-EP saw Milton for what he was -a simpleton and petty tyrant.” (Gonsalves, Stansell and Howes, 2009, p. 257).

Operatives were directly responsible to their superiors for keeping hostages safe and alive. Their new role could lead to some operatives to have different induced preferences, setting them apart from their leaders. While for the FARC-EP’s leadership the hostages were a cherished bargaining chip, for others in the organization, the hostages represented the possibility of being captured or killed by the Army, or the possibility of being tried and punished by their own organization if they failed. This difference suggests a clue that in these types of tasks, the delegation of certain responsibilities was not beneficial for the entire organization because operatives were less skilled in the hostage custody responsibilities than the leadership believed. As a matter of fact, they were not trained for that type of task. Complicating the issue was the fact that hostages and operatives were sharing the same spaces and dangerous situations, which led to close relationships between some of them. Occasionally, however, the dangerous situations and close living conditions also made the operatives turn cruel and violent. In fact, the leaders had a strong interest in preventing the emergence of close relationships between hostages and operatives. Their preventive strategy was to maintain a high rate of rotation among those who were guarding the hostages. Additionally, they encouraged operatives to be rude and cruel with the hostages while presenting themselves as benevolent and kind.

^{Betancourt (2010)} relates a conversation she had with Joaquín Gómez, one of the FARC-EP’s leaders, when he visited her at the hostage camp:

“I told him everything we’d been enduring at the hands of these often cruel and insensitive men -the constant humiliation, the scorn, the stupid punishment, the harassment, the jealousy, the hatred, the sexism, all the everyday details that poisoned our lives, with the number of things Andrés (camp commander) forbade us to do increasing by the day, the absence of all communication or information, the abuse, the violence, the meanness, the lying.” (^{Betancourt, 2010}, p.166).

Joaquín Gómez responded:

“Don’t worry. I’m watching over you. As long as I’m here, there are things that won’t happen’. I smiled sadly. He was too distant and too high up in the hierarchy to really be able to protect me. He was an inaccessible to me as I was to him because of both the distance and the stubbornness of these subordinates. He knew this.” (^{Betancourt, 2010}, p. 168)

In the previous quote, it is easy to note the differences between the benign treatment dispensed to the hostages by the FARC-EP leaders and the harsh and cruel approach taken by guards. Their behavior was not homogeneous or absolute-that is to say, between those kidnappers who used multiple tactics to coerce their hostages and those who displayed a single or predominant manner of treating hostages in captivity (^{Phillips and Pohl, 2013}).

In ^{Betancourt’s (2010)} and Gonsalves, Stansell and Howes’ (2009) versions, the leadership was always attentive, helpful and willing to respond to hostages’ requests; in contrast, the majority of camp staff tended to be indifferent or have a confrontational position and attitude with the hostages all the time.

US contractors describe a similar perception of their guards:

“…In spite of all this, most of the time we were as thick as thieves. The Mud Camp’s conditions, the cords and harnesses, the severe blow to our hopes of a quick release, all combined to really rub us all raw. Even when those disputes were at their worst, we were becoming close as brothers. We were seeing the guards as even more of an adversary that before. With the cords around our necks and being tied up, we became more dependent on them. We hated that and they hated that. If you had to pee, you needed a guard to come and untie you and take you to the trench. Sometimes they didn’t feel like letting you go, so they wouldn’t, for an adult to have to plead with someone to let you relieve yourself was incredibly demeaning. It seemed to be the FARC-EP’s intent to drag us down as low as they could.” (Gonsalves, Stansell and Howes, 2009, p. 138).

This quote shows that a member’s position within the organizational hierarchy shaped how that operative treated hostages. An operative’s incentives and reasons to mistreat hostages were inversely related to his position within the organization. By definition, the leadership carried more clout within groups than their operatives. Many of the guard expressed their feelings of anger and frustration toward their hostages. These operatives’ behavior can be interpreted as a hidden behavior that was not observed nor controlled by the leadership.

CONCLUSIONS

This paper discussed two different interpretative approaches from economic theory in order to explain the change of combatants’ preferences caused by political kidnapping carried out by the FARC-EP. The political and economic costs of that strategy led the FARC-EP to renounce the kidnapping in February 2011 and then to initiate the peace process with the government in August of the same year, which ended in 2018 with the agreement signed by both parties for a stable and lasting peace.

The first approach was taken from ^{Phillips and Pohl (2013)}, who saw the preference problem as a change in risk for operatives. The second approach came from ^{Shapiro (2013)}, who presented the problem as a divergence of preferences both underlying and induced. It also discussed the reasons behind the change in the behavior of an operative who is initially committed to the cause of the organization and agrees with the leaders on how best to serve the cause.

However, in the case of the FARC-EP, some low-ranking operatives exhibited remarkable changes in their behavior, showing that their preferences and beliefs were not aligned with those of their leaders; this change developed as a result of the leaders’ adoption of the policy of long-term political kidnappings. The leadership did not fully evaluate the effects that this strategy would have upon its operatives. This phenomenon engulfed the organizational structure at the same time that the government’s strategies for stimulating desertion were increasing.

This method can capture operatives’ changes in behavior when they face situations with different levels of risk. To capture the essence of this method precisely, in the case of the FARC-EP, we can therefore think of the operative as if he is making decisions faced with two critical situations: combat and kidnapping. It is assumed that the operative has different attitudes toward the risk depending upon his location. If he prefers combat to kidnapping, he might be more willing to assume a higher level of risk in a confrontation with the Army than in a kidnapping situation. However, if he prefers kidnapping to combat, then he might be more averse to risk due to his new responsibilities. That is, with this example, we illustrate the role of risk attitudes, which is useful to identify the effects of different external situations on operatives’ behavior.

Therefore, the contribution of this paper is showing that different approaches from economic theory may explain why some operatives change their preferences in spite of such preferences were considered unchangeable. Two factors affect scenarios in which operatives make decisions: i) the leadership’s decisions in which operatives don’t take part due to the organization’s top-down decision-making structure, which reduces any space for operatives’ participation, and ii) the external conditions that indirectly depend on the behavior of the organization as a whole. Although operatives identify with their organization’s objectives, some unexpected decisions from their leaders may not be well received. That is, accepting decisions from their leaders depends on each operative’s expectations of the decision’s consequences. As ^{Shapiro (2013)} states, leadership and operatives evaluate differently the consequences of actions by the organization-particularly when the leadership’s decisions have a long-term impact. In our case, the combat action may be perceived by the operative as an action with immediate results and that is affected by how they act. In contrast, kidnapping results depend on outside conditions beyond operatives’ control, such as successful negotiation between organization leadership and the government. The parties face different levels of risk. By the same physical and organizational separation between leadership and operatives, the latter who are in the most direct contact with hostages will tend to build relationships with them. The leadership will not have control over the effects of this interaction on the operatives’ behavior. The leadership also does not know how environmental changes will affect the preferences of operatives.