Counting Lives in a Theater of Terror - an Analysis of Media-oriented Hostage Takings in Iraq, Afghanistan, Pakistan and Saudi Arabia

Counting Lives in a Theater of Terror - an Analysis of Media-oriented Hostage Takings in Iraq, Afghanistan, Pakistan and Saudi Arabia

by Judith Tinnes


This article summarizes key findings of a quantitative and qualitative analysis of media-oriented hostage takings involving local people and foreigners in Iraq, Afghanistan, Pakistan and Saudi Arabia during a five-year period (01-01-2004 to 12-31-2008). Based on a long-term tracking of Islamist online publications, the chosen approach did not only allow to identify static values, but also granted insights into developments over time. Characteristic patterns regarding responsible organizations, nationalities of victims, distribution channels for hostage media, etc. could in this way be identified.  The kidnappers' organizational affiliation turned out to be the most significant variable. It decisively affected other factors like number of abductions, likelihood of fatal outcome or duration of a kidnapping. As a consequence, the highest attention should be given to  this parameter when dealing with hostage situations. Another key finding is that insurgents have been increasingly refraining from broadcasting visual representations of extreme violence. This article is based on the German-language dissertation J. Tinnes. “Internetnutzung islamistischer Terror- und Insurgentengruppen unter besonderer Berücksichtigung von medialen Geiselnahmen im Irak, Afghanistan, Pakistan und Saudi-Arabien” -  in English: “Internet Usage of Islamist Terrorist and Insurgent Groups with Special Regard for Media-oriented Hostage Takings in Iraq, Afghanistan, Pakistan, and Saudi Arabia”).[1] In the following I shall summarize key findings in order to make them accessible to non-German speaking readers.


In the 21st century, the Internet has established itself as a leading medium for the distribution of terrorist messages. As a communication strategy, terrorism is dependent on a dissemination of its deeds by the mass media to be able to unfold its full psychological effect. [2] Jihadists and their sympathizers have recognized that the Internet – due to its power to realize a more even playing ground than hierarchical one-way media – is an excellent medium for a new form of warfare: Electronic Jihad (E-Jihad). This is especially true for terrorist hostage takers who are using the stage of the world to play their theater of terror. [3] The choreography of their kidnapping operations fulfills all needs of high-drama theater as noted by Gabriel Weimann and Conrad Winn: “script preparation, cast selection, sets, props, role playing, and minute-by-minute stage management” [4] The strategy of kidnapping is particularly well suited for getting an extraordinary media attention. Most hostage takings last for a prolonged period during which they unfold a dramatic potential. While bombings, suicide attacks, sniper operations and other terrorist strategies are often anti-climatic and are usually only reported when the events have already come to an end, the media coverage of acts of hostage taking normally starts at a moment when the terrorist activity is still ongoing and the result of the kidnapping drama is open-ended. The uncertainty about the fate of a hostage creates suspense – the key ingredient for a sensational infotainment story likely to capture the attention of large audiences. This explains why kidnappings have become established as a mass-phenomenon in conflict zones with high terrorist activity such as Iraq or the Afghanistan-Pakistan (AfPak) region.


In my dissertation, I have conducted a quantitative and qualitative analysis of media-oriented hostage takings of local people and foreigners in four countries. My working definition is derived from Crelinsten and Szabo [5]:

A hostage taking is a form of extortion, during which a militant organization seizes one or more persons and holds them at an undisclosed location to achieve political, financial, or psychological goals.

By media-oriented hostage takings I understand kidnappings during which media publications (videos, images, audio statements) were released. Media-oriented hostage takings tend to get a high level of attention from mass media and play an often decisive role.

Minwoo Yun  has observed that“….this area of study has suffered a shortage of quantitative data and corresponding analyses. Thus, most studies and articles on this topic have been descriptive and narrative.“ [6] I therefore tied to arrive at concrete numbers and percentage values for media-oriented kidnappings. By using numerical results, characteristic patterns (e.g. responsible groups, nationalities of victims, forms of used violence) could be visualized, from which actionable knowledge for decision-makers involved in future acts of hostage-taking might be derived. Practical information on the identity and nationality of kidnappers and their victims as well as ideology, aims, modus operandi and demands of terrorist groups can assist policy-makers confronted with a kidnapping crisis.

Due to the high number of kidnapping cases worldwide, this analysis was limited to a five-year period (January 1, 2004 to December 31, 2008). Geographically, it was confined to Iraq, Afghanistan, Pakistan and Saudi Arabia. The year 2004 was selected as entry point because it marked the beginning of a kidnapping wave in Iraq; many of these were very media-oriented hostage takings. Afghanistan and Pakistan were chosen because similar abduction waves emerged. The insurgents from the AfPak region often used reports on the nefarious activities of their Iraqi brethren as a blue print. Saudi Arabia was included in my sample for reasons of contrasting: while, on the one hand, the situation in the Saudi Kingdom showed clear parallels to the kidnapping wave in Iraq (existence of an Al-Qaeda branch, kidnapping and attack wave in 2004), Saudi Arabia is a country with more stable political structures – hence insurgents could not gain ground to the same extent they managed to do in the other three nations examined here. A special focus of my analysis was on non-Western victims since they are often disregarded or ignored by Western mainstream media. My analysis was also meant to counteract this selective tendency and offer an analysis that is as inclusive and integrated as possible in the current data situation.

The basis of the data survey was a long-term tracking of Islamist Internet presences – which I have been conducting since early 2006 – as well as a retrospective survey for the years 2004 and 2005, using media reports. During three years, relevant Islamist online presences were searched for new publications on a daily basis and all relevant material was archived. In addition, secondary sources like online news outlets, websites of experts and institutions in the terrorism research field as well as topic-related discussion forums like Clearinghouse Infovlad and Terroristmedia were used. The information gathered was sorted, aggregated and condensed in the form of data tables. [7]

For the years 2004/2005, only hostage takings with at least one fatality could be taken into account due to the fact that the daily long-term tracking started only in 2006; events in earlier years had to be researched retrospectively based on media reports. A practical consideration was that incidents during which people are killed are rather more often reported by the mainstream media than those without any fatality. [8] As most Western mainstream media tend to disregard or ignore non-Western hostages – especially if there are no fatalities – a retrospective analysis exclusively based on secondary sources would have produced only partial results. Hence, the 2004/2005 data survey for the analysis was restricted to those newsworthy hostage takings involving fatalities. Because of the different survey approaches, I split the data in two different sets to allow for greater accuracy. The first set (Corpus 1 / below: C1) contains hostage cases with fatalities from January 1, 2004 to December 31, 2008. The second set (Corpus 2 / below: C2) comprises not only kidnappings with fatalities (in which itoverlaps with C1) but also abductions with other possible outcomes (hostages freed, fates of captives unknown) for the period January 1, 2006, to December 31, 2008.

The assembled data were analyzed with the help of simple descriptive statistical methods. What follows are my key findings.


Number of Hostage Takings and Hostages

Between January 1, 2004 and December 31, 2008, a total of 178 kidnappings with fatalities was recorded (C1). All told, during those incidents, 545 persons were seized. In the period from January 1, 2006 and December 31, 2008, there were 107 abductions (involving 373 persons), in which the victims were eventually either killed or freed; in some cases their fate remains unknown (C2). On average, there were three (3.0) kidnappings every month (C1 and C2). Figure 1 shows that the monthly number of hostage takings during the analyzed period was fluctuating. Experts have come to the conclusion that the abduction rate underlies rhythms and cycles. Such dynamics can not only be observed with kidnappings but also occur with other forms of terrorist activities. [9] Conducting attacks at a symbolically ‘loaded’ point in time is a popular terrorist strategy - such a conduct maximizes the newsworthiness of an attack. Experts have attributed the abrupt start of the kidnapping wave in Iraq in April 2004 to the U.S. military engagement in the unrest city of Falluja, which started April 4th. Reacting to the military operation, insurgents began to kidnap foreigners to dissuade the country of origin of the hostages from joining the ‘coalition of the willing' or – if they already belonged to it – to raise their troop levels. The abduction wave culminated in a rush of kidnappings in the month of October. A catalyst for this development could have been the holy month of Ramadan, during which Jihadists tend to increase their attacks. Yet with regard to hostage takings, this assumption is problematic; no clear correlation between the abduction rate and the holy month of Ramadan could be observed in the following years.


Motivated by high media attention, many armed groups so adopted hostage-taking as a strategy, triggering a wave of kidnappings. Thomas Hegghammer called it

a classic case of 'contagion', i.e. terrorist actors learning from one another [...] Put simply, the introduction of the abduction tactic by certain groups at the outbreak of the Falluja crisis most likely inspired other insurgents groups and militias to adopt the same tactic, thus creating a snowball effect or 'epidemic' [10]

One example for a correlation between the abduction rate and symbolic events are the parliamentary elections in Iraq on January 30, 2005. In the weeks prior to the elections, insurgents launched a series of attacks against voters, polling stations, election workers and candidates [11] to avoid the establishment of a democracy – a political system which, according to their purist interpretation of Islam, runs counter to proper Islamic ruling because of its principle of man-made law (instead of God-given law). Three of the abducted persons shown in January 2005 hostage videos had been directly involved in the election process. Ten of the hostages executed on-camera during that month were killed in broad daylight on busy streets. Observers regarded this brazen action as a carefully crafted strategy of psychological warfare aimed at demonstrating the perpetrators' power and, by implication, the weakness of the Iraqi government.

In 2008, with a total of only two (C1), respectively six (C2) abductions, the kidnapping rate reached its lowest level to date. On average, 0.2 (C1), respectively, 0.5 (C2) abductions occurred per month. This positive development can be mainly attributed to the improvement of the security situation in Iraq.

Location of Hostage Takings

The bulk of recorded kidnappings took place in Iraq. 167 (93.8%) of the incidents recorded in C1 and 91 (85.1%) of the cases in C2 happened in Mesopotamia. The other examined countries played only a secondary role:

  • Afghanistan: C1: 5 kidnappings (2.8%) / C2: 8 kidnappings (7.5%)
  • Pakistan: C1:  3 kidnappings (1.7%) / C2: 6 kidnappings (5.6%)
  • AfPak frontier region [12]: C1: 2 kidnappings (1.1%) / C2: 2 kidnappings (1.9%)
  • Saudi Arabia: C1: 1 kidnapping (0.6%) / C2: 1 kidnapping (0.9%)

Hence, only 5.6% of the abductions in C1 and 15.0% in C2 took place in the AfPak region. In Saudi Arabia, only a single hostage taking – belonging to both data sets – was recorded. This low number can be attributed to the stable security situation in the Kingdom, where successful counter-terrorism operations by the security apparatus prevented the insurgents from gaining a foothold.

If one looks at the geographic location of acts of hostage taking, one can see that, on the one hand, Afghanistan and Pakistan have begun to play a growing role in  kidnappings (see figure 2). This is due to the facts that the Afghan and Pakistani Taliban movements have been gaining momentum and are taking over the media tactics of their Iraqi brethren. On the other hand, it is important to point out that the abduction activities in those nations have never reached the volume of Iraq – at least so far.



While in the years 2004, 2005, and 2006, more than 95% of the recorded abductions took place in Mesopotamia (C1), Iraq has lost in significance a as kidnapping location since 2007. In that year, its percentage share in the abduction activities dropped to 88.0% in C1, respectively, 84.1% in C2. In 2008, Iraq lost its leading role for the first time since the beginning of the kidnapping wave. In C1, its share in total kidnappings lowered to 50.0%, in C2 even to 16.7%. Figure 2 makes clear, that the drop of the ‘kidnapping industry’ in Iraq, once “a signature of post-war Iraqi terrorism” [13] was the decisive factor for the general decline of acts of hostage taking in 2008. The main reason for this development can be found in the improvement of the security situation in Mesopotamia (particularly the weakening of radical Islamist groups like Ansar al-Islam and Al-Qaeda in Iraq. These two groups were – as we will see below – most often responsible for this type of crime).

Fate and Gender of Hostages

Most media-oriented kidnappings ended with the death of at least one hostage. In 93 (86.9%) out of 107 kidnappings in C2, one or several abductees lost their lives. In only nine acts of  hostage taking (8.41%), all captured victims regained their freedom; in the remaining five (4.7%) cases, the fate of at least one captive was left unknown. 273 (73.2%) of the 373 hostages were killed, 79 (21.2%) were freed while the fate of 21 others (5.6%) remains unknown. The low number of unknown outcomes indicates that abductors seem to think that it is more useful to inform the world about the ending of a kidnapping operation than to generate psychological pressure by creating suspense. The main cause of death for hostages was execution at the hands of  their kidnappers. All 273 fatalities in C2 [14] were caused by execution. Abductions with fatalities (C1), were – evidently – characterized by an even higher percentage of killed victims: 456 (83.7%) abductees lost their lives, 70 (12.8%) were freed, the fate of 19 (3.5%) remained unknown. Shooting was the most prevalent execution method: 267 (59.2%) of the 451 executed hostages (C1) were shot by their captors, 72 (16.0%) were beheaded (though this killing method gained by far the biggest media attention), three of the victims (0.7%) were doused with kerosene and burned alive.

If one analyzes the gender of abductees, one can see that men were much more often kidnapped than women. Less than 8 percent of the abduction victims (C1: 5.1%, C2: 7.8%) were female, whereas the bulk of abductees was male (C1: 94.9%, C2: 92.2%). On top of that, the death rate of male hostages was much higher than that of female victims. More than three-fourths of the kidnapped men (C1: 86.3%, C2: 77.6%) lost their lives. In contrast, “only” 35.7% of the women recorded in C1 and 20.7% in C2 were killed. This noticeable difference can be most likely attributed to moral, cultural and religious restraints. The abduction and killing of women is considered un-Islamic in the Muslim world and has thus been proven counterproductive for a successful media campaign by insurgents. During a mass kidnapping of 23 South Koreans in 2007, the Afghan Taliban were heavily criticized because 16 of their victims were female. According to media reports, the Taliban leadership was “unhappy at the kidnapping of women by their fighters”. [15]

Nationality of  Victims

If we combine both data sets, state residents from not fewer than 31 nations became victims of abductions. Western mainstream media reports have suggested that inhabitants of countries that provided troops for the military operations in Iraq and Afghanistan were mainly affected by kidnappings. However, an examination of media-oriented hostage takings offers a different picture:

  • First, citizens from the three countries in armed conflict  - Iraq, Afghanistan and Pakistan - were the main targets of abductors. More than three-fourths (C1: 72.1%, C2: 76.4%) of the victims came from these three conflict-ridden countries.
  • Second, citizens from countries who had no military presence in Iraq were often targeted. Nepalese, for example, became the fourth frequent target of kidnappers in C1.
  • Third, residents from countries who provided a high amount of troops for the military engagement in Iraq and/or Afghanistan (for example, the Netherlands or Georgia) were not targeted at all.

What speaks against the hypothesis that there is a heightened victimization of coalition nationals is the fact that the death rate of captives from coalition countries is markedly lower than that of hostages from non-coalition nations. While 56.0% (C1), respectively, 27.8% (C2) of the abductees whose origin was a coalition country lost their lives, a much higher number of hostages from non-coalition nations was killed (C1: 97.7%, C2: 69.2%).

If we categorize hostages as foreigners and non-foreigners and examine the victimization patterns year-by-year, one notes that the targeting patterns have shifted to locals since 2005. In the figure below, a ‘foreign’ refers to a person kidnapped in a nation other than his or her home country. Whereas in 2004, the number of foreign and non-foreign hostages was roughly equal (foreigners: 51.0%, non-foreigners: 49.0%), since 2005, the amount of local hostages has increased (C1). Figure 3 shows that the imbalance between foreigners and non-foreigners was especially great in 2007 and 2008: In 2007, the number of local residents was 5.6 (C1), respectively, 4.7 (C2) times higher than that of the non-domestic captives, in 2008, it was even 16.5 (C1), respectively, 6.3 (C2) times higher.


The increasing amount of domestic abduction victims signalizes a change of strategy by insurgents/terrorists - their targeting patterns have shifted from the ‘far enemy’ to the ‘near enemy’. The strengthening  of the security services in  Iraq and Afghanistan as well as Pakistan's supporting role in the war on terrorism have created a situation where “hunting down armed 'collaborators' has become one of the armed oppositions' primary concerns“ [16]. The militants' harsh crack-down against local residents is also mirrored by the death rate of their victims: While 68.0% of the foreigners in C1 and 38.1% in C2 were killed, a significantly higher amount of locals lost their lives (C1: 89.6%, C2: 83.7%). The high death rate of domestic kidnap victims indicates that such kidnappings of non-foreigners primary serve the purpose of psychological warfare - local inhabitants of conflict zones are dissuaded from supporting the ‘apostates’ and ‘crusaders’.

Profession of Hostages

Around half of the abductees worked in the security sector (C1: 47.5%, C2: 50.9%). Here, persons with a military or paramilitary background were most often targeted (C1: 25.1%, C2: 28.7%). Civilians who worked for the government were the second most victimized professional category (C1: 17.4%, C2: 19.3%). The bulk of them (roughly 95% in both data sets) performed security tasks (police, building protection etc.). Private security contractors were the fourth (C1: 5.7%), respectively, fifth most (4.0%) targeted occupational group. There are two obvious reasons for the above-average victimization of security professionals:

  • First, security forces are fighting directly against the insurgents and are therefore considered enemies by them.
  • Second, most of the abduction victims who worked in the security sector were assigned to tasks that – on a more or less regular basis – required direct enemy contact  that exposed them to a higher abduction risk.

The higher percentage of security professionals in C2 is a further sign for a strategy changing towards the ‘near enemy’ as most of the victims from this occupational group were non-foreigners. Not only the targeting patterns but also the death rate of abductees  reflect the fact that members of  security organizations and government employees are treated by the insurgents as enemies. More than four-fifths (C1: 87.6%, C2: 83.1%) of the hostages linked to a military or paramilitary entity, lost their lives. Yet the death rate of government employees was even higher. All abductees in C1 and 98.6% in C2 were killed during their captivity. On the other hand, the death rate of professionals belonging to the private security sector firms was markedly lower (C1: 74.2%, C2: 73.3%). Among other factors, this might have been caused by the fact that many of them were foreigners – a status which made them more interesting for financially motivated kidnappers.

In an earlier study on kidnappings in Iraq (covering the period April to August 2004), Thomas Hegghammer had noticed a “clear dominance of drivers and manual workers”  [17] among the victims. This study reaches similar findings: In C1, drivers were the third, workers the fourth most often targeted professional group. However, in C2, their significance decreased (drivers: from 8.4% in C1 to 3.2% in C2; workers: from 6.8% in C1 to 1.61% in C2). These findings which can also be observed in regards to other professional groups were probably not only caused by the different approaches of both data sets, but might also be a further indicator for a heightened victimization of the ‘near enemy’: While Iraqi abductors in 2004 eagerly tried to deter foreign nations from providing helpers for the reconstruction, since 2005, their interests have shifted to the ‘near enemy’, especially the domestic security forces.

Organizational Membership of Abductors’

As most kidnapping incidents in our four countries study took place in Iraq, it is not surprising that the bulk of hostage takings in both samples (C1: 94.0%, C2: 85.0%) was conducted by militant organizations active there. Insurgent groups claiming abductions can be dichotomously categorized in regularly and singularly active organizations. The first category consists of established militant groups which regularly claim attacks over a longer period of time. Their modus operandi is not restricted to hostage takings but rather encompasses a much broader spectrum of different terrorist and/or guerrilla tactics. The second category comprises organizations who surface only once or a few times. Usually, they claim just a single operation – mostly a kidnapping – and are not heard of again. Only in exceptional cases, they conduct one or two more abductions. Experts suppose that many of those mayflies, in fact, are brigades of established organizations who do not want to use their ‘company name’ for strategic reasons. Most radical Islamist groups, for example, avoid openly demanding ransom money because asking for material goods is deemed un-Islamic in their salafist interpretation of Islam. In C1, 14 militant organizations who claimed kidnappings belonged to the first category, 16 to the latter. In C2, 12 insurgent groups could be assigned to the first category, 11 to the second. All organizations belonging to the category of singularly active organizations were based in Iraq.

Both data sets contained abductions in which the captors did not reveal any organizational affiliation. The low frequency of such cases (C1: 1.6%, C2: 3.7%) indicates that hostage takers are anxious to operate under a group name (even if it is not a common one like in the case of singularly active organizations). This conduct is probably part of the perpetrators' media strategy: As terrorism is a communication strategy, a terrorist attack is generally only successful if it becomes known who has conducted it (except in certain ‘false flag’ operations). Besides, the perpetrators' identification heightens the possibility that a terrorist attack is newsworthy because the mainstream media are more willing to report on an act of terrorism if the militants' affiliation is known. [18] In the category of the regularly active groups, the bulk of abductions (C1: 79.8%, C2: 70.0%) was claimed by only four militant organizations – all based in Iraq: Ansar al-Islam (AAI), Al-Qaeda in Iraq (AQI), Jaish Abu Bakr al-Siddiq al-Salafi (JAB) and Islamic Army in Iraq (IAI). The fact, that the three first mentioned groups are radical Islamist organizations, implicates that media-oriented kidnappings are the strategy of choice for groups with a radical Islamist ideology. The national Islamist IAI, had once been the organization of the national Islamist camp whose ideology was similar to AQI's. In recent years, growing rifts between radical and national Islamist organizations have led to a situation where formerly co-operating groups have grown apart in terms of ideology. This development was also mirrored in their kidnapping activities: While the IAI claimed a total of 6 kidnappings with fatalities in the years 2004 and 2005, it assumed responsibility for only one abduction per year in 2006 and 2007. It did not claim a single kidnapping in 2008.

Among the four Iraqi groups primary responsible for kidnapping activities, AAI and AQI dominated. Together they claimed 71.6% of the abductions in C1 and 64.5% of those in C2. Albeit AQI got the biggest media attention, it was in fact AAI that was responsible for most of the media-oriented abductions (C1: AAI: 43.7% vs. AQI: 27.9%; C2: AAI: 41.1% vs AQI: 23.4%). A comparison between C1 and C2 reveals a rise of kidnappings by the Afghan and Pakistani Taliban (C1: 5.5% vs. C2: 12.2%). The increased percentage cannot be explained by the differences between C1 and C2 alone, but can certainly also be attributed to the strengthening of the Afghan and Pakistani Taliban which triggered a growing number of abductions in both countries. All the other regularly active groups had only a marginal share in kidnapping activities (C1: 3.8%, C2: 3.7%). The singularly active groups also played only a secondary role. Roughly one-tenth of the recorded kidnappings (C1: 9.3%, C2: 10.3%) were claimed by them.

The organizational affiliation of the captors did not only decisively influence the frequency but also the outcome of abductions (see Figure 4).


For example, nearly all kidnappings (C1: 93.8%, C2: 93.3%), which were conducted by AAI, AQI, JAB and IAI ended with the death of all abductees. If we relate the death rate to the high activity level of these groups, it becomes evident why abductions in Iraq are characterized by a very high ratio of fatalities. People who fell in the hands of the Taliban had a markedly higher chance of survival than captives of AAI, AQI, JAB and IAI: 50.0% of the Taliban kidnappings in C1 and 30.8% in C2 ended with the death of all abductees. The differing fatalities ratio of both organizational categories is probably rooted in the different demand strategies of the two groups: While AAI, AQI and IAI made disproportionate demands (like a complete withdrawal of troops from the country), the Taliban bargained for more ‘reasonable’ concessions (like a prisoner exchange). For Iraqi radical Islamists, making demands was often only a matter of form whereas their true motivation was exercising psychological pressure. JAB did not raise any demands at all – which implies that their hostage takings were exclusively driven by psychological warfare considerations. On the other hand, the Taliban did not only try to influence their target audience psychologically, but also aimed at extorting concrete concessions (especially when kidnappings involved foreigners).

If one analyzes the influence of the organizational membership, two more aspects are striking:

  • The method of killing: For example, AAI, AQI, JAB, IAI and the Afghan and Pakistani Taliban claimed 91.7% of all hostage decapitations. AQI beheaded most captives (30.6%) closely followed by AAI (27.8%). It appears that decapitations are the strategy of choice for radical Islamist organizations while groups with a more moderate Islamist or secular ideology refrain from this archaic and cruel execution method.
  • The duration of kidnappings: For instance, Taliban kidnappings lasted markedly longer than abductions by AAI, AQI, JAB and IAI (C1: 35.7 days vs. 9.2 days, C2: 39.1 days vs. 6.8 days). This is a further indicator that the Taliban are more anxious to extort reasonable concessions than their Iraqi brethren whose main motivation consists in psychological warfare and who are therefore less interested in time-consuming negotiations.


With regards to demands, is has to be pointed out that their analysis was more problematic than that of most other data categories; more than three-fourths of the recorded demands  could not be classified unambiguously. In 82.6% of the cases in C1, the captors did not publicly make any demands for the release of their victims. In C2, the perpetrators abstained in 75.7% from demanding any concessions. Theoretically, abductors can refrain from raising demands for two reasons (however, merely on the basis of open source information, it cannot be said which one is more true):

  • The captors are not interested in negotiations: This is the typical setting of psychological-warfare motivated hostage takings. Those are lacking the element of demands which is otherwise stereotypically associated with ‘classical’ hostage situations. Instead of extorting a concrete entity (for example, a government), the captors try to influence a collective (the public) – i.e. a larger, less-specified target group; they want to influence that group psychologically to achieve a certain result such as increase public pressure to withdraw from coalition warfare. In the absence of negotiations, the hostage does not function as a ‘bargaining chip’.
  • The demands are conveyed in secret: many countries officially adhere to a strict no-ransom policy (i.e. they do not make any concessions to kidnappers). However, in spite of the official ‘no negotiations, no concessions’ declaratory policy, a number of  hard-line governments have engaged in several unofficial deals. In order to avoid  undermining the credibility of a country's official policy, such deals are usually made in secret and denied in public. 

If we analyze hostage takings with regards to the nature of demands, we notice a significant dominance of cases with solely political demands (C1: 90.3%, C2: 76.9%). Cases, during which the captors brought forward exclusively financial demands, were either absent (C1) or  played a subordinate role (C2: 7.7%). Hostage takings with a combination of political and financial demands were rare too (C1: 9.7%, C2: 15.4%). These findings indicate that abductors apparently refrain from explicitly demanding money. However, it is safe to  assume that the actual percentage of financially motivated kidnappings is higher - the nature of public demands does not always reveal the kidnappers' true motivations. For example, hostage takers with financial aims can pretend political demands to raise the pressure, or politically motivated kidnappers accept a ransom payment if they realize that their extortion attempt will not lead to accepting their political demands.

Among the political demands, the most frequent single demand was a prisoner release (C1: 38.1%, C2: 39.7%). The second most frequently made demand was for the withdrawal of troops from a town, a region, or a country (C1: 23.8%, C2, 15.9%). With regards to financial demands, kidnappers either requested a ransom (C1: 3.2%, C2: 9.5%) or asked for a compensation money for human and material losses caused by military operations (C1: 1.6%). However, the latter demand was only raised once.

Hostage and Execution Videos

Hostage videos are the most prevalent type of publications in media-oriented hostage takings (C1: 78.8%, C2: 72.1%). Other types of exposure (i.e. images of the captives or their ID-cards as well as audio statements) played only a secondary role (C1: 21.2%, C2: 28.0%). In 130 (73.0%) of the 178 hostage takings recorded in C1, the captors broadcasted video footage. During those kidnappings, 187 hostage videos were released – an average of 1.5 videos per abduction. In 71 (66.4%) of the 107 kidnappings recorded in C2, the captors published video footage. In those cases, 99 hostage videos were released – on average 1.4 videos per kidnapping. The average run-time of the tapes was 6.3 minutes in C1, and 8.0 minutes in C2, respectively. The longer average run-time in C2 was influenced by the fact, that since 2006, captors have released longer lasting compilation videos lasting between 24 and 63 minutes.

A sub-classification of the recorded hostage videos reveals that execution videos and ‘classical’ hostage videos (i.e. tapes which show the captive alive serving as a proof-of-life or as a mouth piece for publicizing demands) were the most frequent type of hostage footage. 52.4% of the videos recorded in C1 were execution videos, 36.4% were ‘classical’ hostage videos. Conversely, in C2, ‘classical’ hostage videos had the biggest share in video publications (51.5%), whereas execution videos here made up only the second-largest sub-category (36.4%). While the differing methodical approaches (only kidnappings with fatalities vs. abductions with all types of outcome) certainly contributed to the quantitative differences between C1 and C2, those disparities can also be attributed to a general decline of execution videos due to changing ‘PR’ considerations of the kidnappers. Other types of hostage videos (for example, post-mortem tapes showing captives after execution) played only a marginal role. Together, they made up only around ten percent of all video publications (C1 and C2). The popularity of ‘classical’ hostage videos and execution footage can be attributed to several advantages perceived by the kidnappers:

  • Credibility: Hostage videos serve to provide a proof-of-life and evidence that a group is indeed responsible for a kidnapping and holding the victim in captivity. Execution videos are a macabre perversion of proof-of-life as they prove that the victim was indeed killed by the captors.
  • Emotionalization: Hostage and execution videos enable the audience to get a picture of a captive's physical and psychological situation. As the footage documents the victim's fear and agony, it is also used to create compassion in one or more target groups, who can thereby be induced to apply political pressure (e.g. appeals to government) and thereby  influence the political process (reduce sympathy for an acting government, calls for a troop withdrawal from a conflict zone, etc.)
  • Newsworthiness: Hostage footage tends to get extraordinary media coverage. Kidnappings involve life-and-death situations where the outcome is in the balance and sectors of the public identify either with the victim or (more rarely) with the terrorist. The combination of “human interest”, conflict and drama often proves irresistible for media, especially commercial ones, wishing to attract and hold audiences.
  • Psychological warfare: Hostage videos demonstrate the power of the abductors to exert total control on their defenseless captives; by displaying the helplessness of the  victim they enhance public fear of the perpetrators thus adding weight to their threats and demands. Execution videos are the embodiment of a ‘theater of terror’ and create maximum shock value. However, they also create feelings of revenge.

If we sub-categorize execution videos, we can see that videos involving killing by gunfire were the most frequent type of execution footage (54.8%). Beheading videos accounted for 42.3%. Beheadings are more frequently media-transmitted than shootings: While only 16.0% of all hostage executions were beheadings, decapitation videos made up 42.3% of the recorded execution tapes. On the other hand, the percentage of shootings was higher than that of shooting videos (59.2% vs. 54.8%). Nevertheless, the controversial effect of execution footage (especially beheading scenes) has caused the insurgents to rethink the use of such cruel footage as an instrument in their war of ‘nerves’. The fear of losing part of their  constituency is evident from internal and public directives by Jihadist ideologues (for example, Ayman al-Zawahiri). There have also been heated discussions in Jihadist forums on the Internet. As a result, there has been a sharp decrease of execution videos.


With regards to decapitation footage, changes in the public release patterns could already be observed in the period between summer 2004 and summer 2005. Figure 5 visualizes that during the first Iraqi kidnapping peak in autumn 2004, kidnappers mainly published beheading videos. However, in the course of the next abduction peak in early 2005, this distribution almost inverted: shooting videos became the most prevalent type of execution videos.

Publication Channel for Hostage-related Media Products

If we have a look at the publication outlets kidnappers used for the distribution of their footage, a clear preference for the Internet can be seen. Around three-fourths of the recorded hostage media were distributed online (C1: 75.4%, C2: 73.5%), whereas only a quarter was disseminated via traditional media channels (for example, the Qatar broadcasting company Al-Jazeera). The reasons for the popularity of the Internet are obvious. They are rooted in several advantages for both militants and their sympathizers: first, and foremost, the versatile applications (propaganda, publicity, psychological warfare, fund-raising, recruitment, etc.) but also in more general advantages (amongst them: independence from time and location, multimedia environment, independence from editorial control or censorship by official media, possibility of quick and anonymous communication). Together, they make the Internet an ideal platform for terrorist threat- and violence-based communications.

Figure 6 makes clear that the Internet did not gain its leading role right from the beginning of the Iraqi kidnapping campaign; it reached its full potential only with some delay. However,  after a long-term preference for the use of  the Internet we see that traditional media re-gained some of their initial significance in 2008. This might have to do with the kidnappers' organizational affiliation. The highly active Iraqi radical Islamist groups (AAI, AQI, JAB and IAI) made strongest use of the Internet. 89.5% of their media-oriented publications recorded in C1 were disseminated online; in C2, these four Iraqi groups even used the Internet to the exclusion of other channels. In contrast, the Afghan and Pakistani Taliban passed the bulk of their media footage to traditional media outlets (for example, the Pakistani news station GEO TV). This can be partly attributed to the fact that the Taliban's media skills ARE still behind those of Iraqi insurgent groups – even though it has made considerable progress in recent years due to the influence of their more Internet-savvy sister organizations (especially Al-Qaeda Central and AQI).



A year-by-year comparison of online and traditional publications reveals three characteristic patterns, apparently determined by organizational specifics:

  • 2004: Gradual unfolding of the AAI/AQI/IAI [19] media campaign / maneuvering between online and traditional publications: The three most Internet-savvy Iraqi insurgent groups, responsible for the bulk of media-oriented hostage takings, did not started their media campaigns at the same time. Only by August 2004, all three organizations regularly published hostage media. Besides, in their ‘experimental phase', AAI, AQI and IAI disseminated their materials through different channels (probably to examine the media effect). Execution videos were mainly distributed online, while ‘classical’ hostage tapes were often passed on to traditional media outlets. For example, 83.3% of AQI's execution videos were posted to the Internet, while only 30.0% of their non-bloody footage was released online. In doing so, the kidnappers apparently tried to circumvent any censorship of their cruel footage by the mainstream media.
  • 2005/2006: Commitment to the Internet / predominance of online publications: Since February 2005, AAI, AQI, IAI (and since then also JAB) relocated their media activities exclusively to the Internet. The trend to online publications is probably rooted in the evolution of the insurgents' media campaign strategy. Apparently, the militants came to the conclusion that the Internet is a better instrument for the distribution of terrorist materials than the traditional media.
  • 2007/2008: Revival of traditional distribution channels: Increasing kidnapping activities by Afghan and Pakistani insurgents triggered a growing number of hostage media portrayals from the AfPak region. As the AfPak insurgents (especially the Taliban) are less Internet-savvy than their Iraqi counterparts, the traditional media gained back prominence at the expense of the Internet.

The findings indicate that generalizations about  Internet usage by terrorists are tricky as practices are decisively influenced by group-specific features.


By utilizing quantitative methods that allow the identification of concrete numbers and the calculation of percentage shares of media-oriented acts of hostage taking – a terrorist strategy that, with a few exceptions, had only be discussed on the basis of intuitive estimates until now – this study has (at least partially) closed an existing research gap. On the one hand, the analysis provided information about quantitative aspects of the kidnapping activities. For example, the total amount of media-oriented hostage takings (C1: 178, C2: 107), abducted persons (C1: 545, C2: 373), and published hostage videos (C1: 187, C2: 99) could be determined. The long-term tracking of Islamist online presences did not only allow the identification of static values (i.e. total values), it also made possible an analysis of developments over time. For example, it could be shown that in 2008, the number of media-oriented hostage takings has markedly decreased (drop to 0,2 (C1), respectively, 0,5 (C2) kidnappings per month – while the total monthly average was 2.97 kidnappings).

On the other hand, the study offered qualitative insights into the strategy of media-oriented hostage takings. So it became clear that Iraq was the main conflict theatre for abductions (more than 85.0% of the kidnappings recorded in both datasets happened in Mesopotamia), and that Afghanistan and Pakistan have begun to play a growing – though not yet comparable – role. The kidnappers' organizational membership turned out to be the core parameter in the analysis. It decisively affects other factors such as the amount of kidnappings, the likelihood of fatalities, the duration of hostage crises, or the choice of publication channels for hostage-related media footage. When assessing and responding to hostage situations, counter-terrorist  policy-makers should therefore give top priority to organizational factors of terrorist groups.

A holistic assessment of the findings of the study makes it necessary to take into account the restrictions of the methodology utilized in this study. First, I have to stress the incompleteness of body of data in database  C1. The restriction to hostage takings deemed newsworthy by official and public media (i.e. those with fatalities) stood in the way of an integral view of the phenomenon. A second restriction, which also relates to C2, is the limitation on media-oriented hostage takings. Since there are bound to exist discrepancies between the insurgents' PR-oriented propaganda campaign and their actual needs on the physical battlefield, only an additional analysis of non-media-oriented kidnappings – and a comparison between media-oriented and non-media-oriented hostage takings – could guarantee a holistic examination of the activities. In Iraq and Afghanistan, a kidnapping-industry has emerged in recent years; likewise in Pakistan, we can observe an increase in hostage takings. Most of these abductions occur in the shadow zone beyond the scope of the news value system that determines coverage by the mainstream media. A comparison of media-oriented (psychological warfare, political motivation) and non-media-oriented (mainly criminal) abductions would show us in all probability a much larger proportion of financially motivated abductions. This hypothesis is supported by reports on child abductions in Iraq: while in none of the media-oriented kidnappings children were targeted, time and again one can find reports on non-media-oriented, financially motivated abductions, in which Iraqi children were targeted. [20]

A third shortcoming of this study is its limitation to Iraq, Afghanistan, Pakistan and Saudi Arabia. To get a fuller picture of media-oriented hostage takings, one should expand the analysis to other regions and also explore supra-regional developments. In addition, a comparison  between hostage takings where the location of the victim is not known, and hostage situations where the hostages are holed up in a known location with the hostage takers might also be warranted. Last but not least, the role and quality of official and public media-coverage and the Internet in the contagion of  kidnappings and hostage takings ought to be explored.

About the author: Judith Tinnes studied Information Science and New German Literature and Linguistics at Saarland University (Germany) where she obtained a Magister degree in 2006. In 2010, she earned a PhD in Information Science at the same institution. Since 2006, she has been conducting extensive research on the Internet usage of Islamist terrorist and insurgent groups by monitoring Islamist websites.


[1] The full text of the dissertation can be found at: [accessed:  10-29-2010]

[2] See Cornelia Glück, ‘ETA und die Medien‘. In: Sonja Glaab (Ed.) (2007). Medien und Terrorismus – Auf den Spuren einer symbiotischen Beziehung. (Wissenschaft & Sicherheit, Vol. 3). Berlin: Berliner Wissenschaftsverlag, p. 22.

[3] The term ‚theater of terror‘ was introduced by Brian Michael Jenkins (1974). Terrorism and Kidnapping., p. 3. [accessed:  10-29-2010]

[4] Gabriel Weimann and Conrad Winn (1994). The theater of terror: mass media and international terrorism. New York: Longman, p.52.

[5] Ronald D. Crelinsten and Denis Szabo defined hostage taking as “extortion to achieve political, monetary, or psychological goals.“ – Ronald D.  Crelinsten and Denis Szabo (1979). Hostage-taking. Lexington: Heath, p. ix.

[6] Minwoo Yun (2007).’Hostage Taking and Kidnapping in Terrorism: Predicting the Fate of a Hostage’. Professional Issues in Criminal Justice, Vol. 2, Issue. 1, p. 23.

[7] The author will provide the tables to interested readers upon request

[8] See, for example, David Th. Schiller (2007): 'When it bleeds, it leads the headlines....' Ein Essay zum Thema 'Medien und Terrorismus' aus journalistischer und sicherheitspolitischer Perspektive. In: Sonja Glaab (Ed.) (2007). Medien und Terrorismus – Auf den Spuren einer symbiotischen Beziehung. (Wissenschaft & Sicherheit, Vol. 3). Berlin: Berliner Wissenschaftsverlag, p. 101, or Gabriel Weimann and Conrad Winn (1994), op. cit. , p. 128.

[9]  See, for example, Michael Eisenstadt and Jeffrey White (2006).‘Assessing Iraq's Sunni Arab Insurgency’. Military Review, Vol. 86, Issue 3, p. 41; Norman Antokol and Mayer Nudell (1990). ‘No one a neutral: political hostage-taking in the modern world’. Medina: Alpha Publications of Ohio, p. 163, or Thomas Hegghammer. (2004).’The Iraqi Hostage Crisis: abductions in Iraq, April-August 2004’. FFI/RAPPORT-2004/03105., p. 15. [accessed: 10-29-2010]

[10] Idem, p. 9.

[11] See Michael Eisenstadt and Jeffrey White (2006), op. cit., p. 40.

[12] Due to a lack of background information, an exact classification by country was not possible.

[13] Kasun Ubayasiri (2004).’Virtual hostage dramas and real politics’. Ejournalist, Vol. 4, Issue. 2, p. 1. [accessed: 10-29-2010]

[14] In one case, the kidnappers stated that their victim, a British national, had committed suicide. But later it was revealed that, in fact, the man was killed by his captors.

[15] Emal Pashtunyar (2007):.’Taliban Leadership Angry Over Kidnapping Women’.Korea Times, 16 August,  2007; [accessed: 10-29-2010, link expired]

[16] International Crisis Group (2006). ‘In Their Own Words: Reading the Iraqi Insurgency’. Middle East Report N°50. 15 February 2006;, p. 24. [accessed: 10-29-2010]

[17] Thomas Hegghammer, op. cit., p. 20. [accessed: 10-29-2010]

[18] Gabriel Weimann and Conrad Winn, op. cit.,p. 128.

[19] JAB claimed only two hostage takings in 2004; as its influence on overall kidnapping activities was only marginal, it was not considered in this examination.

[20] See, for example: Gerry J. Gilmore (2008). ‘Terror Tactics Backfire on Al Qaeda in Iraq, Colonel Says’. (American Forces Press Service, February 06, 2008). or AFP (2009): ‘Iraqi boy rescued from abductors after 70 days’. (Google News, June 06, 2009). [accessed: 10-29-2010]


Creative Commons License
This work is licensed under a Creative Commons Attribution 3.0 License.


Perspectives on Terrorism is  a journal of the Terrorism Research Initiative and the Center for Terrorism and Security Studies

ISSN  2334-3745 (Online)

Disclaimer, Terms and Conditions