Left wingers like to claim that there is no point killing terrorist leaders because it either doesn't affect the organisation or makes them even stronger as it pisses off members (and the populace from which they draw their support). Naturally the supposed solution is to give in to the terrorists' demands.
But when you look into the literature, you find support for the claim that killing terrorist leaders can be effective:
"Very few groups have radicalized after a crisis in leadership. Indeed, of the thirty-one cases of leadership crisis that we examined, only one group became more radical. In that case, the radicalization occurred after the arrest of the group’s leader. This finding differs from what we expected. Our second hypothesis was that assassinations would most likely lead groups to become more radical. Making a martyr out of a leader, we assumed, would lead to a strengthening of resolve and the subsequent radicalization of the group. Our findings provide no evidence to suggest that our hypothesis is correct. In fact, a group appears much more likely to disband or become less radical after the assassination of a leader. This suggests that the psychological impact of the assassination of a leader is not sufficient to spark a radical transformation in a group’s ideology. An additional explanation for our finding is that the assassination of a leader might often be accompanied by widespread attacks on the movement that disrupt much more than the top leadership, precipitating the failure of the group. Another possibility is that the assassination of a leader often greatly disrupts the movement, possibly causing a loss of funding, internecine fighting, and/or the loss of the leader’s unifying and charismatic personality.
A related finding is that the assassination of a leader is more likely to cause a group to fail or disband than is an arrest of the leader. This may be because a group is more likely to remain intact if members believe that their leader will eventually return to power. Alternatively, as in the case of Action Directe in France, leaders may be communicating with active members of the group from their prison cells, thereby perpetuating their influence (Dartnell 1995).
Table 2 also reveals that the natural death of a leader normally results in “no change” to the movement... However, since we have only two cases of natural death in our sample set, it is difficult to draw general conclusions from these cases.
Looking at the aggregate indicators, our analysis suggests that about half of the movements (slightly more than fifty percent) continue with “business as usual” after a crisis in leadership. At first glance, this suggests that something other than the leader is critical for group cohesion... the results might also reflect a tendency among leaders to make arrangements in case they might be incapacitated. If a group is highly institutionalized and has clear lines of succession, then the loss of a leader would presumably be less likely to cause major changes in its direction...
The arrest of a spiritual leader, like Rizal, contrasted with the arrest of a logistical leader, like those who dominated the PKI, might be more likely to induce a group to become more radical...
A group that experiences internal disputes is likely to split into factions immediately following a crisis, but will also be more likely to survive it. This type of group may not reunite in its pre-schism form, but its strongest faction or factions will be able to continue the movement in coexistence with competing organizations...
Groups that factionalize but lack experience with dispute resolution are ill-equipped to deal with factionalization successfully. However, groups that have allowed some level of debate and dissent are inherently more adaptable in the face of internal challenges and are thus more likely to emerge from a schism. This type of group may not reunite in its pre-schism form, but its strongest faction or factions will be able to continue the movement in coexistence with competing organizations...
Organizations with a hierarchical command structure might be better suited to deal with setbacks than groups with a more decentralized command structure...
Our findings also reflect the importance of religious ideology in a group’s evolution... Their membership may have decreased in number and moderated, but a clearly defined membership still existed... Religious beliefs may be among the strongest sources of cohesive attachment for a group. While certain doctrines may change, the fundamental claims inherent in religion tend to persist in the face of obstacles that may stymie other ideologically-based, but non-religious, movements, such as Marxist and nationalist movements...
We also found that groups that rely highly or exclusively on a leader’s teachings for their doctrine display a universal trend toward survival...
Our first hypothesis—that a group that fails to fulfill a stated goal will suffer internal disputes but ultimately survive under a revised objective—is largely supported by the data. Only one out of four cases of failure to achieve an important goal ended in the failure of the group. Each of the three that survived had a complex, institutionalized, hierarchical group structure; the one that failed had a more decentralized and informal structure. Thus, groups with more centralized, hierarchical structures may be better equipped to deal with operational failure than informally structured groups.
Furthermore, the data confirmed that groups that radicalize as they develop but subsequently face a crisis (either of leadership or of goal failure) tend to continue under a revised objective. Specifically, the movements with these characteristics in our sample tended to revert to a condition more closely resembling their original, relatively moderate state.
Our second hypothesis—that when the leader of a group is killed by an external force the group will become radicalized—was invalidated by the evidence. In twelve of the eighteen relevant cases, the group continued with little change or actually deradicalized after the killing of a leader; in the other six, the group disbanded. Not a single case emerged where a group radicalized after its leader was killed, suggesting that the psychological impact of a killing does not in and of itself incite a radical transformation in a group’s ideology. Historically, it is also common for the killing of a leader to be accompanied by broader attacks on the movement that disturb the rest of the membership, making the group’s failure more likely.
Our third hypothesis—that the arrest of the leader will not significantly alter the ideology or operations of the group in the long term—was confirmed by our study. Seven out of eleven movements that lost a leader permanently or temporarily to arrest were able to continue without great change. This confirmation implies that an arrest has a different impact on a group than a killing. As an imprisoned leader is still alive, the group may hold onto the hope that the leader will one day be freed and will return to the movement. Thus the group may be less motivated to generate great change. Some leaders have been able to communicate with their groups from prison, via coded messages or lawyers, thereby providing their group with a feeling of continuity and security that would diminish the likelihood of the group undertaking radical change.
Finally, our fourth hypothesis—that the death of a leader from natural causes will not have much of an impact on the group—was supported by the data. We found only two cases in which the leader died of natural causes, but both of those groups experienced smooth transitions and no major changes of direction after the deaths. Importantly, the deaths of the leaders of both groups were anticipated due to their sickness and old age. While it is difficult to draw general conclusions from a sample of two cases, we speculate that groups that have prior knowledge of an impending natural death would tend to take steps to prepare for the death and construct a clear line of succession...
One important finding is that the way in which a leader is neutralized matters. Our research suggests that a movement will react differently during a crisis of leadership depending on whether the leader is killed, arrested, or dies of natural causes. As noted earlier, movements that witness the killing of a leader appear to be more likely to fail than movements in which the leader is arrested. In addition, movements in which the leader dies of natural causes appear to be the most resilient to a crisis in leadership."
--- TARGETING THE LEADERSHIP OF TERRORIST AND INSURGENT MOVEMENTS: HISTORICAL LESSONS FOR CONTEMPORARY POLICY MAKERS (2004) / Lisa Langdon, Alexander J. Sarapu, and Matthew Wells (Journal of Public and International Affairs)
"Some scholars argue that targeting the group's leadership reduces its operational capability by eliminating its most highly skilled members and forcing the group to divert valuable time and limited resources to protect its leaders. Decapitation tactics are also intended to disrupt the terrorist group's organizational routine and deter others from assuming power. Scholars have credited these tactics with creating intra-organizational turmoil and even organizational collapse, most notably, the demise of the Kurdistan People's Party and the Shining Path following the arrests of their leaders...
I argue that leadership decapitation significantly increases the mortality rate of terrorist groups, even after controlling for other factors. Using an original database - the largest and most comprehensive of its kind - I analyzed the effects of leadership decapitation on the mortality rate of 207 terrorist groups from 1970 to 2008. The analysis differs from previous quantitative studies because it evaluates the effects of decapitation on the duration of terrorist groups as opposed to the number, frequency, or lethality of attacks after a group experiences leadership decapitation...
I then use concepts from leadership studies, organizational ecology, and terrorism to provide a theoretical explanation for why terrorist groups are particularly susceptible to decapitation tactics. I argue that terrorist groups have unique organizational characteristics that amplify the importance of their top leaders and make leadership succession more difficult...
Although several scholars have evaluated the effectiveness of decapitation tactics, few have done so systematically. The vast majority of analyses rely on case studies to support a specific conclusion. Others examine the effectiveness of decapitation tactics within a particular country, of which Israel seems to be the most popular. Although these country- and region-specific case studies help policymakers and scholars understand more about this controversial tactic, the findings from these studies cannot be generalized across all terrorist groups.
Three primary works, however, have tried to systematically test decapitation's effectiveness across multiple groups and over longer periods of time, but all focus solely on the relatively short-term effects of this tactic or feature small-n datasets. Lisa Langdon and her colleagues examined nineteen guerrilla, terrorist, religious, and revolutionary groups from 1750 to 2004 that each boasted more than 100 members. They concluded that "the leadership of a group can generally change or be seriously challenged without threatening the group's survival." Langdon and her team, however, based their findings on an extremely small sample that was ill-suited to deriving statistically significant results. Moreover, their study attempted to explain variation in the effectiveness of decapitation across several types of organizations with little in common over a period of more than 250 years.
Aaron Mannes found mixed results in his study. In analyzing the change in the frequency of attacks before and after a terrorist leader was killed or captured, Mannes also relied on a small sample. The study, which examined terrorist groups with more than 100 members, contained only seventy-one groups and sixty decapitation strikes.22 Additionally, most of Mannes's results were not statistically significant.
Jenna Jordan has made by far the most comprehensive attempt to test the effectiveness of leadership decapitation.23 Jordan concluded that decapitation strategies not only are ineffective but may be counterproductive. She found that instead of causing organizational collapse, leadership decapitation often extends the survival of groups that would have otherwise dissolved. Jordan's dependent variable was whether or not the group survived more than two years after experiencing decapitation. Although I agree that organizational survival is a better metric than the number, frequency, or lethality of attacks, Jordan set the standard for evaluating counterterrorism policies too high. A time horizon of two years is a reasonable period to evaluate public policy, but imposing arbitrary time horizons when trying to accurately evaluate leadership decapitation and its effects on terrorist groups may not be useful, especially if the effects persist beyond two years.25 In addition, given the near unanimous agreement in the field that no "silver bullet" solutions exist in counterterrorism, I argue that scholars should not rely on "silver bullet" metrics - for example, whether a group experiences organizational collapse within two years after leadership decapitation - to evaluate counterterrorism policies. Examining the short-term effects of these policies is important, but policymakers should consider their long-term effects as well. Imagine if doctors and patients disregarded chemotherapy and radiation treatments, two of the most popular and successful regimens for treating many types of cancer, because of their painfully debilitating side effects in the short term. This article is an attempt to fill the void by providing a long-range analysis for policymakers to consider when making decisions concerning counterterrorism policy.
For leadership decapitation to be an effective counterterrorism policy, two conditions must be met. First, terrorist group leaders need to be important to the overall success of the organization. If they are not, there is no reason to expect that organizational performance will suffer in their absence. Second, leadership succession must be difficult. If leaders are easy to replace, the benefits of targeting high-ranking leaders may not be worth the costs.
Several scholars have concluded that targeted assassinations are ineffective for ending insurgencies, disbanding drug cartels, and changing state behavior. The conclusions from these analyses cast doubt on the likelihood that leadership decapitation can work against terrorist groups. I argue, however, that terrorist groups are different: they have unique organizational characteristics that increase the influence of their leaders and exacerbate the difficulties associated with leadership succession.
The conventional wisdom suggests that leaders significantly affect organizational performance, but finding quantifiable proof of this causal relationship is surprisingly difficult. When evidence of a causal link exists, it is often weaker than expected...
The selection bias that makes evaluating leadership influence in economic firms difficult is not a factor in analyzing terrorist group leaders.
Second, the institutional constraints that limit the influence of leaders in economic firms and legitimate political organizations do not affect terrorist leaders... Unless the group is state sponsored, terrorist leaders do not answer to a superior or a board of directors. They are not as worried about perceptions of legitimacy or morality from those other than the populations from which they recruit or are trying to influence.
Third, scholars argue that leaders of economic firms can typically affect only a few of the variables that determine organizational performance. Terrorist leaders, however, can wield enormous power and influence over all aspects of their organizations, from their structure and identity to the pace and scale of group activities...
Replacing terrorist group leaders is more difficult than replacing leaders in other organizations... I argue that leadership succession is especially difficult for terrorist groups because they are violent, clandestine, and values-based organizations...
The clandestine nature of terrorist groups also increases their dependency on leaders; complicates leadership succession; and negatively affects organizational learning, performance, culture, and decisionmaking. To maintain operational security and avoid detection from outsiders, leaders of terrorist organizations have a disincentive to institutionalize their operations, making leadership succession difficult. There are two distinct logics behind this disinclination. First, leaders in terrorist organizations do not want to codify how they operate, because doing so makes them more susceptible to state infiltration. Bureaucratization may enhance organizational learning, performance, and efficiency, but it may also provide the state with the knowledge necessary to destroy the organization. Some terrorist groups do have formal hierarchies, but not all members are likely to understand them. Individual cells often maintain independence from one another so that captured individuals or even cells cannot compromise the entire group. This lack of formalization and institutionalization increases the level of uncertainty, which in turn complicates leadership succession and produces organizational instability. This characteristic holds true for all organizations, including legitimate organizations such as state governments following the assassination of the head of state, but its consequences are more significant for terrorist groups.
The second reason terrorist leaders are disinclined to institutionalize their organizations may be more selfish and more personal. Not only do terrorist leaders fear being captured or killed by the state or rival groups, but they also worry about being removed from power by their own group. Similar to leadership succession in other illicit, violent, and clandestine organizations, replacing terrorist group leaders often relies on Hobbesian principles rather than on institutionalized processes. It is common for terrorist leaders to suffer from paranoia, a personality disorder worsened by a clandestine existence that can produce "burn syndrome," or a "pervasive fear that other people know what they're doing." For example, believing that his group was plotting against him, Sabri al-Banna (aka Abu Nidal), head of the Abu Nidal Organization, ordered the murder on a single night of 170 followers whom he suspected were traitors. Abimael Guzmán, leader of the Shining Path in Peru, was so paranoid about being ousted in a coup that he "surrounded himself with female lieutenants but readied none to command in his absence." Because terrorist leaders know that they live and die by the sword, they hesitate to provide subordinates with the knowledge and skills to run the organization in their place. This disinclination to institutionalize not only centralizes power in the hands of the terrorist group's leader, but it injects an air of uncertainty when a top leader is removed, complicating the ability of a successor to understand and run the organization effectively...
Values-based organizations such as religious cults, social clubs, and terrorist groups have greater difficulty replacing their leaders than do profit-based organizations, including drug cartels. Three reasons explain why. First, values-based organizations require their leaders to possess unique skill sets that not every leader has, namely, the ability to provide transformational leadership. Finding successors with these requisite skill sets is not easy. Second, leadership succession is less difficult in profit-based organizations because the monetary incentives of holding power are usually sufficient to attract a steady stream of successors, even when leading involves tremendous risk. The incentives for holding power in values-based organizations can be more complex and more abstract. Third, articulating the vision, mission, and strategy of values-based organizations can be especially difficult when these elements are created from scratch and are hard to conceptualize.
Leaders of terrorist groups must possess a unique set of skills to attract and maintain membership. In his seminal work on leadership, James Burns draws a distinction between transactional and transformational leadership... Transformational leadership... goes beyond personal self-interest by appealing to the values and emotions of followers. Transformational leaders, therefore, seek to create significant change in the behavior and belief systems of their followers, often encouraging personal sacrifice to achieve goals that benefit the team, group, or organization...
My dataset consists of 207 terrorist groups from sixty-five countries active from 1970 to 2008. Among quantitative analyses that have leadership decapitation and terrorist groups, it is the largest dataset It includes 204 observations in which the leader or leaders were either killed or captured. Additionally, I recorded 95 other incidents in which the leader or leaders (1) were expelled from their group, (2) died of natural causes or in an accident, (3) voluntarily resigned from their leadership position, or (4) accepted a cease-fire agreement with the government and formally entered the political process. In total, the dataset contains 299 observations of leadership change...
Previous analyses of leadership decapitation, almost all of which focus on short-term consequences, present a bleak picture of the effectiveness of this counterterrorism tactic. The findings from this study tell a different story...
Regardless of how I conceptualized the effect of decapitation, terrorist groups that experienced the loss of a leader had higher mortality rates than those that did not. Depending on how I modeled the effect of decapitation, terrorist groups were 3.6 to 6.7 times more likely to end than those that did not experience decapitation.
The variables representing group size, state regime type, and organizational structure were statistically insignificant. The results also show that ideology did not affect the group's mortality rate. Right-wing groups were more than four times as likely to end following leadership decapitation when compared to nationalist/separatist groups, but this becomes less interesting when one considers that there were only six right-wing groups in the dataset, four of which ended following decapitation. Terrorist groups with allies are up to 52 percent less likely to end than groups without them, and 39 percent less likely to end if they are competing with rival terrorist groups. The one state-level control that was highly statistically significant throughout all of the models was GDP per capita, a proxy for state counterterrorism capacity. An increase in the log of GDP per capita resulted in a 47 to 53 percent increase in the mortality rates for terrorist groups...
The more time a state requires to remove a terrorist leader, the less impact leadership decapitation will have on the group's mortality rate. As figure 3 shows, the effect of decapitation on a terrorist group's survival rate is cut approximately in half after ten years. At approximately twenty years, decapitation may have no effect at all. The most important finding from this graph, however, is that time matters when decapitating a terrorist group leader.
I estimated three models to determine how group size affected terrorist group mortality rates. Earlier I hypothesized that smaller groups should have higher mortality rates than larger groups following decapitation, because larger groups have more resources and thus more capacity to endure. The findings from the size models, however, show otherwise...
In all three models, size is not an important variable in explaining organizational decline in terrorist groups...
I employed a similar progression to determine how groups with different ideologies affect terrorist group mortality rates...
Unlike the base model, where right-wing groups were the only ideological type to be statistically different from nationalist groups, the religious groups in model 9 are statistically different from their nationalist-group counterparts. Compared to nationalist groups, religious groups are almost 77 percent less likely to suffer organizational death. When a religious group suffers the loss of its leader, however, it is almost five times more likely to end than are nationalist groups. Of the fifty-three religious groups in the dataset, nineteen have ended, including sixteen that ended after the government killed or captured their leaders. Of the thirty-four religious groups still active, twenty have experienced decapitation...
I examined the effect of the method of decapitation on the mortality rate of terrorist groups in three separate Cox models. The results show that all three methods I identify-killing the leader, capturing the leader, and capturing then killing the leader-significantly increased the terrorist group mortality rate...
Some scholars may argue that instead of measuring the effect of leadership decapitation on terrorist group mortality, the main explanatory variable, Exp. Decap, measures "bad" groups or "bad" leaders that needlessly put themselves in jeopardy. In other words, "bad" groups and "bad" leaders get selected out of the system, but this is not necessarily evidence to suggest that decapitation is to blame for the group's demise. To control for this potential endogeneity problem, I included a dummy variable for groups whose leaders die while in command for reasons having nothing to do with state efforts. These include leaders who have died of natural causes or who were killed in some other random way, such as in a car or plane accident. If the hazard ratio for this variable is statistically significant and greater than "1," then this reduces the chances that my analysis suffers from an omitted variable bias or an endogeneity problem.
In model 15, the variable for natural causes drops out because the effect of decapitation is "turned on" only for the year in which decapitation occurred and because there are no instances in which a group ended in the same year as a result of this form of leadership turnover (16 observations). In model 17, however, I changed the specification for the effect of decapitation so that the effect is "left on" for the duration of the terrorist group's life cycle. Here the variable for natural causes is statistically significant at the 10 percent level (p-value = 0.07) and greater than "1," indicating that groups that lose their leaders from an illness or an accident are 2.5 times likelier to end than groups that do not lose their leaders in a similar fashion. Given that several of these cases include leaders who lost long battles with chronic diseases such as cancer (i.e., cases that allowed the group to diligently prepare and plan for the day when their leader died), this is an impressive finding. Because transportation accidents and illnesses are random events, are unassociated with leadership, and could conceivably affect any group, this finding suggests the absence of an endogeneity problem...
Contrary to findings in other studies, I found that religious terrorist groups were less resilient and easier to destroy than nationalist groups following leadership decapitation. Although religious groups appear to be 80 percent less likely to end than nationalist groups based on ideology alone, they were almost five times as likely to end than nationalist groups after experiencing leadership decapitation. I believe this is because of the important role leaders of religious terrorist groups play in framing and interpreting organizational goals and strategies."
--- Targeting Top Terrorists: How Leadership Decapitation Contributes to Counterterrorism (2012) / Bryan C. Price (International Security)
Yeo (2019) qualitatively examines the cases of Hamas, Hezbollah, the Abu Sayyaf Group and Jemaah Islamiyah and it seems that pure decapitation (killing the leader) has mixed effectiveness, but killing multiple leaders is rather effective.