The Characteristics of Suicide Terrorists:

An Empirical Analysis of Palestinian Terrorism in Israel

Introduction

During the afternoon of Friday April 16^th, 1993 , Israeli radio reported a terrorist attack in a restaurant near Mechola in the Jordan Valley . In contrast to previous incidents, this time the terrorist, a member of the Hamas organization, parked a car loaded with explosives between two buses, in a parking area close to the restaurant. He chose not to leave the car, but rather to detonate the explosive device while he was still in it. At that time, most ordinary Israeli citizens, and even those responsible for policy-making, could not begin to imagine that this event, which caused the death of two Israelis and wounded another five, was the first manifestation of a dramatic shift in tactics used by the Palestinian terrorist organizations in their struggle with Israel.

Despite the fact that in terms of the number of casualties this event did not differ much from previous terrorist acts perpetrated by Palestinian organizations, it was nevertheless unique. This was the first time that a Palestinian organization initiated a suicide attack against an Israeli target. The event in Mechola signaled the beginning of a long chain of suicide attacks perpetrated by Palestinians against Israeli targets. Actually, over the last decade the suicide attack became one of the most prevalent tactics of the Palestinian struggle.

Suicide terrorism is nothing new in human history; it first reared its ugly head long before the Palestinian-Israeli struggle. This phenomenon was seen among the Jewish Sicaris as early as the 1^st century, among the Moslem Hashishiyun in the 11^th century and among the Asians in the 18^th century (Schweitzer, 2001) . In the late 20^th century suicide terrorist attacks took place in countries such as Lebanon , Egypt , Turkey and Sri Lanka . Since the early 1990s, researchers have shown much interest in this subject, addressing it mainly from two perspectives. Most of them focused on the operational aspects of suicide bombing (Joshi, 1996 ; Kushner, 1996 ; Israeli, 1997 ; Sprinzak, 2000 ; Schweitzer, 2001) , how the attacks are carried out and under what circumstances. Other analysts (Merari, 1990) have employed psychological theories in an effort to explain the Shaheeds’ (The Islamic suicide Terrorist) motivations.

This analysis takes a somewhat different, sociological approach, and seeks to answer the following questions:

First, what are the social characteristics of individuals who kill themselves or attempt to do so in the course of carrying out terrorist attacks on Israeli targets?

Second, how do such ‘suicide terrorists’ differ from other perpetrators of terrorist attacks, but who do not intend to kill themselves in the course of these operations? Do such social attributes as age, marital status, religious background, gender, educational attainments and experience in participating in terrorist attacks, serve to discriminate between the two types of attackers? Or, is the distinction based simply on random considerations? And if suicide terrorists are somehow different, how does this help us understand the terrorist phenomenon in general?

Suicide Terrorism

Prior to addressing the profile of the suicide terrorist, a short description of the phenomenon is required. Contrary to other terrorist attacks, even the most dangerous ones, the suicide attack is an “operational method in which the very act of the attack is dependent upon the death of the perpetrator” (Ganor, 2000: 1). The terrorist is fully aware that if he does not kill himself, the planned attack will not be implemented. Most suicide attacks are carried out by activating explosives worn or carried by the terrorist in the form of a portable explosive charge, or planted in a vehicle driven by the terrorist himself (Ganor, 2000) . The suicide terrorist actually becomes a ‘human time-bomb’. He selects the time and place to execute the explosion in accordance with the prevailing circumstances, in order to cause maximum damage to the target (Schweitzer, 2001) . Actually, suicide terrorism has inherent tactical advantages over ‘conventional’ terrorism: “It is a simple and low-cost operation (requiring no escape routes or complicated rescue operations); it guarantees mass casualties and extensive damage; there is no fear that the interrogated terrorists will surrender important information (because their deaths are certain); and it has an immense impact on the public and the media (due to the overwhelming sense of helplessness)” (Sprinzak, 2000: 66-68) .

Profiling Terrorists

Attempts to profile terrorists and to look into the causes that lead them to perpetrate violent acts were rather marginal compared to the extensive literature written in the field of terrorism (Schmid, 1985: 191-194) . This is especially true with regard to the phenomenon of suicide terrorists.

A scrutiny of the few academic researches profiling terrorists reveals that, despite the limited scope of the literature, there are two distinct approaches. The first proposes a psychological profile of the terrorist while the other approaches the phenomenon from a sociological perspective. While the psychological attempts to profile terrorists seem to be very promising, they are nevertheless subject to criticism in two major areas. First, there is the argument raised by Laqueur that no profile of a terrorist exists. According to his contention, terrorism fluctuates over time and the profile of terrorism changes with circumstances. He further argues that we can be sure that most terrorists are young but their actions and psychological makeup vary according to social and cultural conditions (Laqueur, 1999: 38-40) . White echoes Laqueur by arguing that individual personalities vary so greatly that it seems impossible to produce a single behavioral profile of a terrorist (White, 2002: 26) . The second argument regarding the difficulties of the psychological profiling of terrorists is methodological and focuses on the limited accessibility of the research subjects, a fact which is most prominent in the case of suicide bombers (Merari, 1990: 202-203) . Indeed, even the most comprehensive attempts to look into the psyche of the terrorist often rely on historical analysis (Israeli, 1997) .

Hence, it seems that a more reliable strategy for profiling terrorists would be the sociological one, though this is also not infallible (Hermann, 1997: 132) . The main problem with this approach is also methodological and is derived from the difficulties in attaining data regarding the socio-demographic background of terrorists, due to the fact that this data is mostly held by governmental and security authorities. In this research we hope to overcome this problem by utilizing a dataset which was gathered especially for the purpose of this study.

Can Suicide Terrorists be Profiled?

Following the discussion of the approaches to profiling terrorists, it should be mentioned that many of the studies dealing with the phenomenon from both psychological and sociological perspectives tend to diminish the uniqueness of the characteristics of perpetrators, especially with regard to suicide terrorists. Martha Crenshaw, for example, argues that the mindset of a suicide bomber is no different from that of Tibetan self-immolators, Irish political prisoners ready to die in a hunger strike, or dedicated terrorists worldwide who would prefer to survive after an operation but know their chances negligible (Crenshaw, 2000) . Ariel Merari, who has conducted a study of individuals who committed acts of suicide terrorism, supports this contention. He profiled more than 50 Muslim suicide bombers serving in Hizballah, Amal, and secular pro-Syrian organizations in Lebanon , as well as Hamas and Palestinian Islamic Jihad in Israel . He came to the conclusion that there is no single psychological or demographic profile of suicide terrorists (Merari, 1990) . Recently, he developed an alternative approach to study the phenomenon. According to his contention, “suicide terrorism as we know it is an organized, organizational phenomenon.” He further argues, based on interviews with Palestinian suicide bombers whose explosive charges failed to detonate, and also with their families, that there was little that connected the Shaheeds besides a few demographic characteristics. The average age was twenty-two, similar to that of the suicide bombers in Lebanon , which was twenty-one. The age range in the earlier wave of suicide bombings was 18-38, with only three of the suicide bombers being older than twenty-eight (Moghadam, 2002) .

Ganor, however, offers some more characteristics of these perpetrators. The typical Hamas Shahid, according to his research, is a religious male, unmarried and unemployed, who had a high school education and is between the ages of 18-27 (Ganor, 2000) . He also emphasized that the suicide terrorists do not have to undergo a long socialization process before participating in the suicidal act. According to his contention, the suicide terrorist was not an active member of any terror organization Before his last fatal attack and he had not participated in regular terrorist activities. On the contrary, in most cases he was a young, vulnerable person with strong religious affiliations. These religious feelings had been skillfully used to influence him into participating in terrorist activity. It is not surprising that many of those recruitments take place in mosques and religious schools (Ganor, 2000).

 A similar sociological profile may be found among the Black Tigers who are the suicide battalions of the Liberation Tigers of Tamil Eelam (LTTE) in Sri Lanka . According to Gunaratana, Black Tiger recruits tend to be young, unemployed and unmarried (Gunaratna, 2000) . Joshi, who studied the same group, adds that the age range of the recruits is 14-16 and the ratio between women and men is three women for every two men. Women are preferred as recruits due to the fact that they tend to wear heavier clothes and are not subject to the same kind of movement restrictions and body searches as men (Joshi, 2000) . A slightly different profile is manifested in the case of the suicide bombers belonging to the Kurdistan Workers’ Party (PKK). Only four out of the 15 attacks carried out by the organization (from 1995-1999), were perpetrated by men, the remainder by women. In terms of age, education and socio-economic background, these women did not differ much from suicide terrorists in other places. Their age range was 17-27, they possessed no professional skills and in general they came from poor families (Ergil, 2000) . A study that was conducted among Shi’as in Lebanon reinforces the correlation between low socio-economic status and the readiness to perpetrate suicide acts (Schbley, 2000: 182-183) . Based on the aforementioned assumption that suicide terrorists do have some common socio-demographic traits, the next question we should address is how do suicide terrorists differ from non-suicide terrorists?

The Profile of the Suicide vs. the Non-Suicide Terrorist

One of the earliest attempts to offer a sociological profile on terrorists was made by Russell and Miller (1977) . According to their research, the average terrorist in the 1970s was a single male, aged 22-24, with at least a partial university education, most often in the Humanities. He came from an urban background and an affluent middle or upper class family that enjoyed some social prestige. Most of these terrorists were recruited on university campuses and adhered to Marxist or other revolutionary theories. In the early 1990s these findings were challenged by two different studies (Handler, 1990; Smith and Morgan, 1994) . According to Handler (1990), as a result of theoretical and methodological problems, the findings of Russell and Miller suffered from severe bias. Following Handler (1990) critics, Smith and Morgan (1994) chose to differentiate between the profiles of left and right-wing terrorists. Indeed, some of their findings deviated substantially from Russell and Miller’s. For example, in terms of age, they found the average terrorist to be much older than indicated previously. The average age of the left-wing terrorist, according to their study, was 35 while that of the right-wing counterpart was 39. In terms of gender they found that 73 percent of the left-wing terrorists and 93 percent of the right-wing terrorists were males. The two groups differed dramatically, however, in regard to race and education. Ninety-seven percent of right-wing terrorists, and 29 percent of left-wing, were white. In the field of education, 54 percent of left-wing terrorists held a college degree, as opposed to only 12 percent of the right-wing terrorists. The place of residence of the left-wing terrorists was mostly urban while those of the right-wing came from rural areas.

Neidhart (1992), who conducted a similar study on the population of German terrorists, both left and right-wing, concluded that most of the left-wing terrorists tend to be young; however, right-wing terrorists generally resemble their US counterparts who are somewhat older (Weinberg, 1993; Smith and Morgan, 1994) . With regard to gender, Neidhart’s findings are very close to Smith and Morgan’s. He found out that among the left-wing organizations women had much higher representation than those on the right. The educational background of the German terrorists was also found to be similar to their American counterparts. Neidhart also found out that while 42 percent of the left-wing terrorists held university diplomas, only 10 percent of the right-wing terrorists possessed a similar degree. (Neidhart, 1992: 219) .

Table 1 around here

Table 1 presents the sociological characteristics of both suicide and non-suicide terrorists as featured in the literature. A comparison of the characteristics within the group of suicide terrorists (Hamas, LTTE and PKK) sheds light on some interesting elements which are relevant for the issue of profiling. First, all the suicide terrorists are young (the age range is 14-27). Second, in most cases they are unmarried. Third, most of them are rated low in terms of education, employment and SES status. However, there is a clear distinction between the three cases in terms of gender, religiosity and ideological affiliation.

We contend that these three variables are related to each other. While in the case of Hamas, which is an Islamic religious fundamentalist organization, all the suicide bombers are males, in the other cases the picture is somewhat different. In contrast to Hamas, both the LTTE and the PKK are primarily nationalistic organizations lacking religious characteristics. This allows them to rely mostly on female suicide bombers. Support for this argument may be found in the Palestinian suicide attacks on Israeli targets in early 2002. For the first time since the appearance of the suicide tactic in the realm of the Israeli-Palestinian conflict, four women carried out these acts. They all belonged to the Al-Aksa battalions, the military wing of the nationalist Fatah movement. The religious organizations (Hamas and Islamic Jihad) did not denounce the women’s participation, but none of their attacks was carried out by a woman. 

Research Hypotheses

Thus, we may hypothesize that there are some features which characterize suicide bombers: (H1) suicide bombers will be young; (H2) unmarried; they will have a religious background which includes: (H3) religious education and (H4) religious ideology; (H5) they will come from a low SES background. In the case of suicide bombers who belong to religious fundamentalist groups we hypothesize that the terrorists will mostly be males (H6).

The organizational approach leads us to yet another hypothesis derived from Merari’s contention about the important role of the organizational environment in preparing the terrorist for a suicide mission. Our hypothesis (H7) is that suicide terrorists go through a longer socialization process within the organization than non-suicide terrorists, and thus are expected to have more experience in perpetrating terrorist acts.

Method

Overview of the Sample and the Research Design

In order to address these research hypotheses, we utilized a dataset that was established for the purpose. The dataset was created in two stages. First, we had to locate all the terrorist events (suicide and non-suicide) which had taken place in Israel over the last decade. This was done through a careful scrutiny of the Israeli daily ‘Ha’aretz’ by a group of trained students.[1] Each terrorist event that was located was coded, according to a detailed codebook. In the second stage, we established a quantitative dataset which included characteristics relevant to profiling them, based on the sociological approach. The whole process of locating and coding the data was cross-checked in order to avoid any possibility of bias or loss of data. This methodology, which became relevant over the years in the study of social protest as well as political violence (Hibbs, 1973; Lehman-Wilzig, 1990), was entitled ‘Events Data’. It allows us to acquire longitudinal and reliable information with regard to the phenomenon under study as well as its characteristics and causes (Peters, 1998: 177).

Overview of Analysis

The research population includes 819 terrorists who participated in terrorist acts from April 1993-February 2002; 743 of them are non-suicide and 80 are suicide terrorists.[2]

In order to test our hypothesis we applied two stages of statistical analysis. Our primary analyses were descriptive, containing comparisons of frequencies and means. In order to find significant differences between the two groups (suicide and non-suicide terrorists) we applied ANOVA analysis and Chi-Square measurements. At the second stage we attempted to compose a profile of the suicide terrorists. Thus, a logistic regression model is used to overcome the problem of predicting probabilities of discrete independent variables and to estimate the factors influencing the terrorist’s prospects of belonging to the suicide terrorists’ group.

Measurements

The dependent variable in this study, as mentioned earlier, is the character of the terrorist attack. ‘Suicide attack’ was defined as an act in which the death of the perpetrator was an integral part of the operation and necessary for its accomplishment (Ganor, 2000). As for the independent variables, eight variables representing a wide range of socio-demographic traits were used (Fleming & Stohl, 1988). These were: age, marital status, type of education, ideological affiliation, SES background and previous experience in terrorism acts (Please refer to table 2).

Table 2 around here

Results

Prior to the analysis of the suicide bombers make-up, we would like to present some characteristics of the suicide terrorism phenomenon in Israel . These characteristics may indicate the magnitude and main features of this phenomenon over the last decade.

Figure 1 around here

As mentioned earlier, ever since the first suicide act in April 1993, this tactic has become an integral element in the Palestinian struggle. Figure 1 indicates the emergence of the suicide tactic compared to other terrorist modes of operation. Indeed, in the early 1990s’, suicide events were far less common in comparison to other events, but with time they increased to almost 20 percent of all events in the mid 1990s. A significant decline in the volume of the attacks can be detected in 1999 and the first half of 2000.  However, with the outbreak of the Al-Aksa Intifada there was a renewal in the occurrence of this kind of attack. By early 2001 these suicide attacks constituted 13.4 percent of the total number of Palestinian terrorist acts and at the beginning of 2002 they amounted to almost 10 percent.[3]

Figure 2 around here

Figure 2 indicates the targets of Palestinian suicide attacks. Most of the events took place in urban centers, mainly shopping centers (38%) and public transportation (30%). As for other features, the most common mode of operation of the suicide bombers was detonating explosives carried on their own bodies (77%). Only in 23% percent of the events did the perpetrators activate larger explosive devices inside cars. As expected, the number of casualties, both injured and killed, is far higher in suicide attacks (17.59) compared to non-suicide acts (2.42). 

As for the characteristics of the perpetrators, Table 3 presents a comparison between suicide and non-suicide terrorists based on the above-mentioned independent variables. In order to assess whether the differences between the two groups are significant we applied a variance analysis test (ANOVA) for the interval variables and Chi-Square for the nominal ones.

Table 3 around here

The findings presented in table 3 give an initial support for most of our research hypotheses. A relatively surprising finding is derived from a variance analysis, which we conducted for the age variable. Despite the fact that the significance is only marginal (P=.083) the descriptive findings indicate that generally the suicide bomber (M=24.58, S.D.=7.46) is older than the non-suicide terrorist (M=22.4, S.D.=5.94).

 In more detail, when we divided the cases into three age groups, we found out that while suicide terrorists were under- represented in the youngest and oldest age groups, they dominated the middle one (ages 22-27) (please refer to figure 3). This finding casts doubt on our first hypothesis (H1) and requires further analysis that will be presented later.

Figure 3 around here

With regard to the second hypothesis (H2), here the findings were clearer. Suicide terrorists in general tend to have fewer family ties than non-suicide terrorists. This conclusion is derived from the fact that the percentage of bachelors among the suicide terrorists was 84.2% while among the non-suicide terrorists it was 60%. This finding was found to be highly significant.

The two hypotheses according to which suicide terrorists would have a more religious background (H3 and H4) were confirmed in this analysis. As presented in table 3 the percentage of suicide terrorists among those who were educated in religious schools (82.8%) is far higher compared to that of non-suicide terrorists (36.1%). Moreover, the percentage of suicide terrorists affiliated with religious fundamentalist organizations (88.4%) is higher than that of the non-suicide terrorists (56.9%). Both of the results were found to be significant. Thus it seems that religious affiliation plays a major role in the terrorists’ tendency to perform suicidal acts. 

As expected in (H5) we found a significant difference between the two groups with respect to the SES background. The findings indicate that suicide bombers usually come from less affluent villages and towns than their non-suicide counterparts.

As mentioned earlier the gender of the suicide terrorist is also widely debated in the literature. No significant finding was detected in the comparison between the two groups. However, we should indicate that, unlike other terrorists organizations in the Palestinian scene, women are only marginally represented in terrorist activities (1 percent among suicide terrorists and 3 percent among non-suicide ones). Moreover, women's activities are almost absent in fundamentalist religious groups such as Hamas and the Islamic Jihad (1.3 percent) and only somewhat more visible in the activities of nationalist organizations such as Fatah (2.7 percent).

Our last hypothesis (H7) also gains initial support from the findings. It appears that the suicide act is rarely the first terrorist activity in which these perpetrators have taken part. Only among 20 percent of the suicide terrorists was the suicide attack the first terrorist act they had ever been involved in, in contrast with 35 percent among the non-suicide terrorists. Thus, these findings support our hypothesis, which was derived from Merari’s contention.

Table 4 around here

Our goal in the final stage of the analysis was to introduce a profile of the suicide terrorist. In order to do so we applied a logistic regression model, a procedure often used in profiling groups of people that are categorized by a Specific behavior (Bachman & Peralta, 2002; Hickman & Piquero, 2001). Table 4 present the Wald effect and the part R² of each variable (which indicate the relative contribution of the variable effect on the total R²).

 In general we can see that religious education, which is the strongest variable in the model, and religious affiliation, both strengthen the affinity of the terrorist to choose suicide. (Though the ideology variable b estimate is negative, in the model it represents the tendency of the terrorist to become suicidal).

The second strongest predictor is the terrorist's age. This was found significant and allows us to assume that the suicide terrorist is likely to be older than his non-suicidal counterpart, thus supporting the findings of the descriptive statistics, which indicated that the suicide terrorists tend to be more experienced in terrorist activities.

Finally, the weakest variable in the model was the SES background of the terrorist. However, we can see that, significantly, suicide terrorists tend to be from the lower 1/3 of the SES Scale. We did not find significant differences between terrorists from the middle of the scale and those of the upper 1/3. Therefore we can assume that low SES background plays a major role in the tendency of the terrorist to choose suicide.[4]

 All the variables above were found significant by chi-square and Wald statistic (both are commonly used to test significance of the variables in LRM), thus we can conclude that suicide terrorists tend to be older than non-suicide, from lower SES background and with stronger religious affiliation.

Although there is no widely-accepted direct analog to OLS regression R², we used Hagle and Mitchell’s pseudo-R-Square and found that the percentage of variance explained by the model is (R²= .48). To conclude, these findings support our rationale for using the sociological approach and the assumption that socio-demographic characteristics of the terrorist give a fair indication of his prospects of belonging to a suicide terrorist group.

Discussion of Findings

Our findings challenge some popular assumptions about the ‘careers’ of suicide bombers in the context of the Israeli-Palestinian struggle. Commentaries in the United States and elsewhere have tended to depict the bombers as naive. They are often depicted, prototypically, as vulnerable young men, individuals with virtually no previous involvement in terrorist violence. These unsophisticated young people are recruited or come forward themselves as the result of a surge in religious commitment brought on by prayer and the inspirational sermons of charismatic Imams.  The atmosphere in the Gaza Strip and on the West Bank is such that the would-be martyr’s decision to embark on his journey of destruction is reinforced by the approval of family and community.  In a society where ‘honor’ is among the highest virtues these are indeed powerful social pressures in support of the suicide bomber’s decision.  Further, once selected for an ‘operation’, the naïve young man is surrounded by a team of Hamas or Islamic Jihad handlers who not only provide tactical instructions etc. but insulate him from any potential cross pressures which might weaken his resolve. 

If this is the conventional picture of Palestinian suicide bombers, how do our findings challenge it?  The social profile (see above) suggests we are dealing with a somewhat different cast of characters. Instead of naive young people, it seems that these are veteran terrorists. They are older, on balance, than non-suicidal Palestinian terrorists, at least those that have been apprehended or killed by Israeli authorities. In addition, the act of committing suicide by attempting to kill as many Israelis as possible in the process represents the culmination of a career in terrorism and not a singular, spectacular event.  As a matter of central tendency, the suicide bombers are individuals who are more likely than their more conventional colleagues to have had previous involvements in terrorist violence. They are by no means newcomers to terrorism. Furthermore, so far as we are able to tell, there are no significant changes over time in these characteristics. Currently, the Shaheeds are not any younger or less experienced than those who launched suicide attacks in the past.

In conclusions, if this is the case, where does it leave us in terms of our general understanding of the sociological portrait of suicide bombers?  It is true that the latter are in some respects similar to many of their counterparts in Israeli prisons. And it is also the case, as our findings suggest, that there are meaningful differences. But the fact that leaps to our attention is the resemblance between what appears to be the process of becoming a suicide bomber and the process of becoming a conventional, for want of a better term, terrorist in other places. The literature concerning other protracted campaigns suggests that becoming a terrorist typically does not involve a sudden conversion to or involvement in terrorist violence. Rather, as in Germany, Italy, Spain and much of Latin America, becoming a terrorist frequently involves a series of steps beginning with participation in direct action protest, small scale acts of defiance, clashes with the authorities which sometimes lead to recruitment or, sometimes, formation of a group committed to terrorism. This pattern seems much closer to the reality of organized Palestinian suicide bombing than the more spontaneous process suggested above, (see, for example: Robin, 2002).  To reduce the spectacular and widely publicized actions of suicide bombers to the mundane language of academic sociology, more than anything else we seem to be dealing with instances of ‘role socialization’. But what a role it is!

Bibliography

Bachman, Ronet and Peralta, Robert. 2002. "the relationship between drinking and violence in an adolescent population: does gender matter?" Deviant Behavior 23,1 : 1-20.

Crenshaw, Martha. 2000. Suicide Terrorism in Comparative Perspective. Countering Suicide Terrorism: An International Conference, Herzliya, Israel, International Policy Institute for Counter-Terrorism.

Ergil, Dogu. 2000. Suicide Terrorism in Turkey: The Workers' Party of Kurdistan. Countering Suicide Terrorism: An International Conference, Herzliya, Israel, International Policy Institute for Counter-Terrorism.

Fleming, Peter. Stohl, Michael and shmid alex. 1988. "The Theoretical Utility of Typologies of Terrorism: Lessons and Opportunities" in The Politics of Terrorism. Edited by Stohl, Michael. New York, NY: Dekker : 153-195.

Ganor, Bohaz. 2000. Suicide Terrorism: An Overview. online article- http://www.ict.org.il/articles/articledet.cfm?articleid=128.

Gunaratna, Rohan. 2000. Suicide Terrorism in Sri Lanka and India. Countering Suicide Terrorism: An International Conference, Herzliya, Israel, International Policy Institute for Counter-Terrorism.

Handler, Joel. 1990. "Socioeconomic Profile of an American Terrorist: 1960s and 1970s", Terrorism 13: 195-213.

Hermann, Tamar. 1997. Social Movements and Political Protest - Terror. Tel-Aviv: The Open University Press.

Hibbs, Jr. Douglas, A. 1973. Mass Political Violence: A Cross-National Causal Analysis. New-York: John Wiley.

Hickman, Matthew and Piquero, Alex. 2001. "exploring the relationships between gender, control balance, and deviance", Deviant Behavior 22,4 : 323-351.

Israeli, Raphael. 1997. "Islamikaze and their Significance" Terrorism and Political Violence  9: 96-121.

Joshi, C. L. 2000. "Ultimate Sacrifice…" Far Eastern Economic Review  163(22): 64-67.

Joshi, M. 1996. "On the Razor's Edge: The Liberation Tigers of Tamil  Eelam" Studies in Conflict and Terrorism 19: 19-42.

Kushner, Harvey. W. 1996. "Suicide Bombers: Business as Usual" Studies in Conflict and Terrorism 19: 329-337.

Laqueur, Walter. 1999. The New Terrorism: Fanaticism and the Arms of Mass Destruction. New York : Oxford University Press.

Lehman-Wilzig, Sam N. 1990. Stiff-necked people, bottle-necked system : the evolution and roots of Israeli public protest, 1949-1986. Bloomington: Indiana University Press.

Merari, Ariel. 1990. "The Readiness to Kill and Die".  in Origins of Terrorism, Edited By William Reich, Cambridge: Cambridge University Press.

Moghadam, Adiniu. 2002. "Fletcher Hosts Ariel Merari, Israeli Expert on Suicide Terrorism". Fletcher Ledger 3(8).

Neidhart, Frank. 1992. "Left-Wing Right-Wing Terrorist Groups: A Comparison for the German Case" in International Social Movements Research.  Social Movements and Violence: Participation in Underground Organizations Edited By B. Klanderm and D. D. Porta.  Greenwich: Jai Press, 215-235.

Russell, Charles . A. and Bowman. H. Miller .1977. "Profile of a Terrorist" Terrorism 1: 17-34.

Rubin, Elizabeth. 2002. "The Most Wanted Palestinian" The New York Times Magazine (June 30, 2002) pp. 26-31,42, 51-55.

Schbley, Ayla. H. 2000. "Torn Between God, Family, and Money: The Changing Profile of Lebanon's Religious Terrorists" Studies in Conflict and Terrorism 23: 175-196.

Schmid, Alex. 1985. "Political Terrorism - A Research Guide to Concepts, Theories Data Bases and Literature". Amsterdam: North-Holland Publishing Company.

Schweitzer, Yoram. 2001. Suicide Bombings - The Ultimate Weapon? .http://www.ict.org.il/articles/articledet.cfm?articleid=373

Smith, Brent, L. and Morgan, Kathryn, D. 1994. "Terrorists Right and Left: Empirical Issues in Profiling American Terrorists" Studies in Conflict and Terrorism 17: 39-57.

Sprinzak, Ehud. (2000). "Rational Fanatics" Foreign Policy 120: 66-73.

Peters, Guy, B. 1998. "Comparative Politics: Theory and Methods" New-York: New York University Press.

Wasmund, Klaus. 1986. "The Political Socialization of West German Terrorists in Political Violence and Terror edited by Merkl, Peter. Berkeley: University of California Press: 191-228.

Weinberg, Leonard. 1993. The American Radical Right: Exit, Voice and Violence. Encounters with Contemporary Radical Right. Boulder, Westview Press: 185-203.

Weinberg Leonard and William Eubank. 1987. The Rise and Fall of Italian Terrorism. Boulder, CO: Westview Press: 77-103.

White, Jonathan, R. 2002. Terrorism: an Introduction. Belmont: Wadsworth Thomson Learning.

Table 1 – The Characteristics of Suicide and Non-Suicide Terrorists

Table 2 - Research Variables

^a The data on each region was collected from the Palestinian Central Bureau of Statistics (PCBS). Afterwards, we calculated the SES value of each region by averaging all the items. We found the variable is reliable (α=.831) and factor analysis showed all the items are conceptually bound.   

* About the Following Variables, we did not have sufficient data -

1. Level of Education

2. Status of Employment

3. Criminal Record

Figure 1 - Distribution of Modes of Terrorists Attacks

Figure 2 - Distribution of Suicide Attacks Targets

Table 3 - Descriptive Statistics and analysis of variance for the Research Variables

^a Chi-square test was preformed for the variable and was found significant

^b ANOVA test was preformed and were found Significant.

^c ANOVA test was preformed and were found of marginal significance

Figure 3 - distribution of terrorist's age  

Table 4 - Logistic Regression Model for the research variables

^a For purpose of simplicity and avoidance of ecological fallacy, the SES was consider in this analysis as categorical variable and was categorized into three main groups (1’- the 1/3 areas with the lowest SES and 3’ with the highest).

^b Although the number of observations is low relatively to the (N) of the whole sample, it includes 77% of the suicide terrorist. Furthermore, we have no reason to assume there is any bias between those terrorists that were included in the analysis and those who were not