Differential Association Explaining Jihadi Radicalization in Spain: A Quantitative Study
June 27, 2017
Abstract: Violent radicalization leading to involvement in jihadi terrorism appears to be highly contingent upon two key factors of what has been termed “differential association,” namely contact with radicalizing agents and pre-existing social ties with other radicalized individuals. This empirical study, which examines all those arrested in Spain for jihadi terrorism activities over the four-year period between 2013 and 2016, quantitatively assesses the importance of these two factors and sheds light on why some individuals radicalize while many more with similar demographic and social characteristics, in the same country, do not. The importance of contact with a radicalizing agent points toward the relevance of ideology in the development of jihadi terrorists, while the significance of pre-existing social ties indicates the relevance of communitarian bonds with local networks, which facilitate terrorist radicalization and recruitment.
Between 2013 and 2016, a total of 178 individuals who adhered to salafi-jihadi attitudes and beliefs were arrested in Spain for terrorism-related activities. The authors gathered information on all of them and built a database—the Elcano Database on Jihadists in Spain (EDBJS)a—so as to study empirically, among other topics, their socio-demographic characterization, patterns of terrorist involvement, and radicalization processes before their detention.b The sources were criminal proceedings and related court documents as well as public hearings at the Audiencia Nacional (National Court) in Madrid, the only jurisdiction in Spain dealing with terrorism offenses, in addition to police reports and press releases from the Ministerio del Interior (Ministry of Interior). Interviews with law enforcement experts, as well as to a lesser degree a search of media reporting, contributed to the resulting database. It should be noted that for certain variables, there is much more comprehensive information about some cases than others.
Most detainees are men (87.1%), three-quarters of them were between 18 and 38 years old at the time of detention, and more often married (54.8%) than not. Moroccans (42.7%) and Spaniards (41.5%) are the main nationalities. Around half are second-generation descendants of Muslim immigrants and 40% are first-generation immigrants. Around 10% are converts. Some 70% attended secondary education, over three times the number of those who only attended primary school. At the time of detention, nearly half the detainees were unemployed (25.2%) or had no known occupation (19.7%). A further 25.2% had some degree of employment in the services sector. No less than one-fourth of all them had a previous criminal record as petty criminals.
In the vast majority of known cases, their radicalization started in either 2011 or 2012 as the civil war in Syria unfolded and jihadis made advances in Mali. The average age at the onset of the process was 20.7 for women and 25.9 for men. Nine out of every 10 detainees radicalized partly or fully while inside Spain. However, far from this phenomenon happening uniformly with respect to the size and distribution of the Muslim population across the country, it occurred in clusters.c The four administrative demarcations where detainees radicalized coincide with their residence: the Catalonian province of Barcelona (23.2%), the North African city of Ceuta (22.2%), Madrid and its metropolitan area (19.2%), and Melilla, the other Spanish enclave surrounded by Moroccan territory (12.1%).
Based on the data from the EDBJS database, this article explores the environments and modalities of radicalization for the detainees. Then, informed by differential association theory, the authors assess the social influences relevant to the detainees’ radicalization. Differential association theory is a longstanding criminological framework for understanding deviant behavior. It posits that individuals become criminals because they belong to social circles in which “definitions” favorable to deviant behavior outweigh alternative ideas and in which deviant conducts, unimpeded by counteracting forces, are learned through interaction with other people, primarily communication in small intimate gatherings.d
Adapting differential association theory to jihadi radicalization, this article assesses thereafter, with respect to the detainees, one of the key differential association factors, namely their previous exposure to salafi-jihadi attitudes and beliefs through contact with radicalizing agents. Finally, the authors look at pre-existing social bonds between detainees and other individuals who similarly became involved in jihadi terrorism activities, including as foreign terrorist fighters (FTFs), in order to gain a better understanding of how their violent radicalization processes took place and tended to cluster in the case of Spain.e
Environment and Modalities of Radicalization
An estimated 40.3% of individuals arrested in Spain between 2013 and 2016 for activities related to jihadi terrorism, in which information on their environment of radicalization was available, were radicalized in the context of a mixed, simultaneously online and offline environment (Table 1).
While an online environment of jihadi radicalization was observed in three-quarters of all these cases, 35.3% of detainees radicalized exclusively online.f
Exclusively online radicalization is typically found among individuals who radicalized solitarily as well as among those who radicalized as part of small clutches of people geographically dispersed across the national territory and who never met in person, at least before actual involvement in terrorism-related activities. A further 24.4% detainees appear to have radicalized only offline, a percentage significantly lower than the one corresponding to those who radicalized exclusively online. The examples of offline-only radicalization in the dataset are found among individuals who adhered to salafi-jihadi ideas while in prison, but also among individuals who radicalized within dense local networks of intimate interpersonal relations and where there was no evidence their use of online communication technologies was directly related to their process of radicalization.
Independent of whether their radicalization took place online, offline, or through a combination of both in a mixed environment, the fact is that an overwhelming majority of the individuals arrested in Spain between 2013 and 2016 for activities related to jihadi terrorism embarked on this journey in the company of other people. That was indeed the case for almost nine out of every 10 of them, an estimated 86.9% (Table 2). Conversely, no more than 13.1% of all the detainees—about one out of every 10 of them—radicalized on their own without interacting with others aside from their exposure to jihadi propaganda, making them genuine cases of self-radicalization.
Most of the detainees who radicalized alongside others did so mainly in mixed environments of radicalization—both online as well as offline—and to a lesser extent, exclusively offline (Table 3). Nonetheless, a small but still significant proportion of them radicalized into salafi-jihadism only online. This latter subset of individuals typically established online contact with one or more individuals, at or near the start of the radicalization process, but never got to know each other in-person. This was observed among detainees who communicated intensely via websites or social media and occasionally by phone, but never in person.1
Data confirms that online jihadi radicalization is not limited to individuals who radicalize alone. In fact, only four out of 10 detainees who radicalized solely online involved purely self-radicalization. It should also be noted that among the small set of 14 self-radicalized individuals in the overall database, 13 ended up adhering to salafi-jihadism solely in the context of online exposure to these ideas. The remaining case corresponds to a converted man, 45 years old at the time of his arrest in 2014, whose radicalization derived from meticulously reading many books and other printed materials on Islam in general and particularly on salafi-jihadi doctrine.2
Online radicalization may therefore take place either in the company of others or in solitary, with the data on individuals arrested in Spain between 2013 and 2016 for jihadi activities showing many more were radicalized by communicating with others online than those surfing the internet without those personal communications. Radicalization in the company of others is also by far the most common modality of jihadi radicalization for those detainees whose process took place to some degree offline, either in an offline-only environment or in a mixed, simultaneous online and offline environment.
Critical Contact with Radicalizing Agents
Among detainees who radicalized in the company of others, irrespective of their radicalization environment, the influence of a radicalizing agent was critical. In all cases of radicalization in company—86.9% of all cases in which relevant data is available as indicated above—individuals experienced the influence of at least one radicalizing agent. However, the type of contact they maintained with a radicalizing agent varied. In 73% of the cases in which relevant data was available, this contact involved only in-person, face-to-face interaction (Table 4), while only 17.6% of these cases involved online contact. Finally, 9.4% maintained contact with their radicalizing agent both in person and online.
In well over half of the cases in which in-person contact with agents of radicalization was observed, at least one or more individuals who can be characterized as activists in part or in whole played this role (Table 5). Activists in this context are defined as individuals previously involved in jihadi activities, often as FTFs in a conflict zone but also as former militants who had been imprisoned in Spain or confined elsewhere. Their experiences furnish them a degree of distinction, even of a charismatic kind, among those they indoctrinate, who come to accept them as ideological guides.
One example of detainees in Spain between 2013 and 2016 who acted as in-person radicalizers and who can be described as activists were two individuals trained by al-Qa`ida or the Taliban in Afghanistan prior to 9/11, captured shortly thereafter, and subsequently confined at the U.S. military base in Guantánamo. They were handed over some years later to Spanish authorities and finally acquitted—by the Audiencia Nacional in one case and by the Tribunal Supremo (Supreme Court) in the other. In the latter case, the acquittal occurred despite a previous condemnatory sentence handed down by the Audiencia Nacional.3
Additionally, the individuals who radicalized through contact with others in-person were also in about 20% of cases radicalized in part or in whole by friends, in 20% of cases in part or whole by religious figures, and in about 19% of cases in part or in whole by relatives. (See Table 5). An example of a religious figure who acted as a radicalizing agent in Spain for arrested individuals or those who became foreign fighters in Syria between 2013 and 2016 is Ibrahim Mohammed Hijjo. A man of Palestinian origin, Hijjo was since 2013 a prominent preacher at an Islamic place of worship in the Basque province of Vizcaya.4
In some of the cases, the individual characterized as activist who acted as the radicalizing agent was also a relative of the person he or she radicalized and who became subsequently detained on terrorism grounds. This was the case of detainee Karim Abdesalam Mohamed, also known as “Marquitos,” who was convicted in 2015 after becoming one of the leading figures in the jihadi circles of Ceuta’s Barriada del Príncipe.5 This man was well-known in the area because of his 2006 arrest and prosecution for terrorism crimes and had also radicalized his own nieces and nephews.6
Notably, religious figures were seldom radicalizing agents for detainees who radicalized online without the help of an in-person radicalizing agent. Activists, either FTF or not, were online radicalizing agents in six out of every 10 of cases in which relevant data was available (Table 6). In over half of the cases involving detainees who radicalized in online contact with an agent of radicalization, the radicalizing agents were peers.
One example of a person who radicalized online with the help of FTFs or activists is Samira Yerou, a Moroccan woman residing in the Barcelona area, who was 32 at the time of her arrest on March 2015. Throughout her radicalization process, she maintained contact via the internet, social media, and instant messaging applications with al-Qa`ida and Islamic State militants in Syria and also with salafi clerics based in countries of the Arabian Peninsula as well as Syria and Morocco, including Omar el-Hadouchi in the latter country, as well as jihadi activists based in Austria.7
The Strength of Pre-Existing Social Bonds
In addition to underscoring the significant role played by radicalizing agents for individuals beginning and continuing the radicalization process, data on detainees in Spain from 2013 until 2016 also clearly shows how the process leading to the adoption of salafi-jihadi ideas and subsequent terrorist engagement is associated with pre-existing social ties to others involved in jihadi terrorist activity. Indeed, pre-existing social ties between detainees and other individuals arrested in Spain for terrorist-related activity or who became an FTF during the four-year period under study were found in nearly seven out of every 10 cases (Table 7).
Most of the detainees in the study had long established social ties with at least one other detainee in Spain or with an FTFg prior to beginning their radicalization process. In turn, over eight out of every 10 of these individuals with pre-existing social ties radicalized in either a mixed online/offline environment or in an exclusively offline context (Table 8). Conversely, almost eight of every 10 detainees lacking these social ties experienced their radicalization processes exclusively through the internet.
These social ties, which are common among the detainees in this study, are based on links formed through neighborhood of origin, friendship, and kinship. These three different kinds of interpersonal bonds are not mutually exclusive and may converge. According to the data, in 80% of the cases where these social ties among detainees or between detainees and FTFs are known to have existed, they originated in neighborhood relationships in the individual’s town/area of residence, and in half of these cases the neighborhood relationships formed within a specific district in the town of residence. (Table 9).
For 50% of the detainees who had previous social ties with at least one other detainee in Spain or an FTF, these interpersonal bonds were based in part on friendship. Likewise, those exhibiting social links to other detainees based in part on kinship made up 42.7% of the cases. Such percentages clearly indicate that different types of previously established social bonds intermingle for many individuals, including intimate ties founded on both friendship and kinship. As has been noted, these relationships are not mutually exclusive.
It is worth underscoring the fact that among the subset of detainees having known kinship ties with at least one other detainee in Spain or with a FTF from Spain or Morocco before beginning their radicalization process, these ties most often refer to bonds between siblings (Table 10). The siblings of detainees or FTFs have indeed been identified in 10 of the police operations carried out in Spain between 2013 and 2016 against the threat posed by jihadi terrorism.h
Also notable among those cases with known kinship ties were bonds with in-laws, couples or partners, and even between uncles or aunts and nephews and nieces. Prior kinship ties between detainees or detainees and FTFs based on relationships between fathers or mothers and sons or daughters, and also between cousins, were also present, but were less frequent than the others already mentioned.
A number of key observations can be made about jihadi radicalization in Spain based on this quantitative study of the individuals arrested in the country between 2013 and 2016 for activities related to jihadi terrorism. Contact with a jihadi radicalizing agent such as an activist, a religious authority, or a relative was observed in 86.9% of cases on which information was available. Social bonds, formed long before the initiation of radicalization and typically from within a neighborhood or via more intimate friendship and kinship relations, were detected in 68.7% of the detainees for whom relevant information exists. The authors’ empirical assessment on those individuals over a period of four years, amid an unprecedented wave of jihadi mobilization in Western Europe, indicates that violent radicalization leading to jihadi terrorism involvement is highly contingent upon these two factors of differential association.
Further research on these two differential association factors—contact with a radicalizing agent on the one hand and pre-existing social bonds on the other—can advance understanding of the jihadi radicalization process in general and help explain why some individuals become radicalized while others with similar demographical and social characteristics in similar countries or within the same country do not. The answer in many cases is they just do not happen to be exposed to the influence of radicalizing agents or have previous close ties with other individuals already radicalized or undergoing the process of radicalization. Conversely, the two factors of differential association help explain the existence of uneven pockets of jihadi radicalization across Spain and other Western European countries.
Furthermore, the importance of contact with a radicalizing agent points to the relevance of ideology in the production of violent Islamist extremists and jihadi terrorists.8 Similarly, the importance of pre-existing social bonds emphasizes the relevance of local networks, made out of interpersonal ties and communitarian bonds, which facilitate jihadi radicalization and recruitment.9 Taken together, these two factors indicate that jihadi radicalization leading to terrorism involvement to a large extent is associated with social interactions through which individuals learn about ideas justifying terrorism. This framework of understanding sheds light on how attitudes and beliefs condoning jihadi terrorism are channeled as well as clustered in Spain, other Western European countries, and perhaps further afield.
From a policy perspective, the implications of the empirical findings on these two major interrelated factors of differential association and recent jihadi radicalization in Spain seem straightforward and twofold, at least considering a range of analogous countries in the immediate Western European scenario. Key in preventing the seeding and spread of jihadi radicalization processes is, first of all, to detect radicalizing agents and neutralize their actions by means of law enforcement. Secondly, national programs aimed at preventing these processes should avoid dispersion of institutional efforts and should focus on localized demarcations where jihadi radicalization is known to concentrate. CTC
Fernando Reinares is Director of the Program on Global Terrorism at Elcano Royal Institute, as well as Professor of Political Science and Security Studies at Universidad Rey Juan Carlos, both in Madrid. He is also Adjunct Professor of Security Studies at Georgetown University in Washington, D.C. His latest books include The Evolution of the Global Terrorist Threat – From 9/11 to Osama bin Laden’s Death (edited with Bruce Hoffman) and Al-Qaeda’s Revenge: The 2004 Madrid Train Bombings, both published by Columbia University Press. Follow @F_Reinares
Carola García-Calvo is Senior Analyst on International Terrorism and member of the Program of Global Terrorism at Elcano Royal Institute, as well as Associate Lecturer on Terrorism and Security Studies at Universidad Pontificia de Comillas, both in Madrid. She co-authored, with Fernando Reinares, Estado Islámico en España (Islamic State in Spain), published by Real Instituto Elcano. Follow @Carolagc13
Álvaro Vicente is a research assistant in the Program on Global Terrorism at Elcano Royal Institute in Madrid.
[a] The EDBJS schema is designed to store information, at an individual level of analysis, for about 136 variables (50 variables on socio-demographical traits, 40 variables on radicalization processes, and 46 variables on terrorist involvement).
[b] According to Spain’s Ministry of Interior, 265 individuals were arrested in the country as a result of police operations against jihadi terrorism in the period between 2013 and 2016. However, it is important to note that this study’s EDBJS excludes detainees who were not finally brought before antiterrorist courts as well as those who, even if arrested in the course of law enforcement operations against jihadi terrorism and brought before a judge, had no proven jihadi ideas. This accounts for the apparent discrepancy in the number of cases.
[c] This finding for Spain has also been observed with regard to other Western European countries in the context of the same wave of jihadi mobilization. See, for example, Georg Heil, “The Berlin Attack and the ‘Abu Walaa’ Islamic State Recruitment Network” CTC Sentinel 10:2 (2016): pp. 1-11; Michele Gropi, “The Terror Threat to Italy: How Italian Exceptionalism is Rapidly Diminishing,” CTC Sentinel 10:5 (2016): pp. 20-28; and Pieter Van Oystaeyen, “Belgian Radical Networks and the Road to the Brussels Attacks,” CTC Sentinel 9:6 (2016): pp. 7-12.
[d] On the original formulation of the differential association theory, see Edwin Sutherland, Principles of Criminology (Philadelphia: Lippincott, 1947, fourth edition). Also Edwin Sutherland and Donald R. Cressey, Criminology (Philadelphia: Lippincott, 1978, 10th edition) and Edwin H. Sutherland, “A Sociological Theory of Criminal Behavior,” chapter 12 in Suzette Cote ed., Criminological Theories: Bridging the Past to the Future (Thousand Oaks, CA: Sage, 2002). A recent application of the theory to radicalization and recruitment for armed organizations relying on terrorism tactics can be found in Taylor Armstrong and Jonathan Matusitz, “Hezbollah as a Group Phenomenon: Differential Association Theory,” Journal of Human Behavior in the Social Environment 23:4 (2013): pp. 475-484. The authors are particularly grateful to professor Emilio Lamo de Espinosa, chairman of Elcano Royal Institute, for his insightful advice on differential association theory.
[e] Quantitative findings are offered, throughout the article, in the form of tables. All 10 tables include a distribution of percentages calculated on the basis of the total number of cases on which data is available for single variables (tables 1, 2, and 7) under consideration or statistics derived from the crossing of variables (tables 3 and 8). When disaggregating variables, data similarly refers to the total number of detainees on which information was found for each of the categories or subcategories being measured (tables 4, 5, 6, 9, and 10). Figures for cases on which there was no relevant information are always indicated as missing data.
[f] Entries in EDBJS are coded as pertaining to radicalization “exclusively online” only where there was absence of evidence of any physical contact with a radicalizing agent, with other individuals undergoing the same process or related face-to-face interaction concerning exposure and adoption of salafi-jihadi ideas before a particular individual became involved in terrorism. Entries are coded as “exclusively offline” where there was an absence of evidence of any other environment of radicalization and where there was no evidence that the use of communication technologies directly contributed to radicalization before a particular individual became involved in terrorism. It was sometimes difficult coding entries in the various categories. However, the authors’ rigorously examined each of the cases and substantiated enough information to code categories about the environment of radicalization for as many as two-thirds (119) of the total number of detainees in Spain between 2013 and 2016 for activities related to jihadi terrorism. Data was missing for this variable in the remaining one-third of the cases (59).
[g] All the foreign terrorist fighters in question had either Spanish or Moroccan nationality, and were residents either in Spain or in Morocco.
[h] Second phase of Operation Kibera in December 2014; first and second phase of Operation Chacal in January and March 2015; Operation Jardín Beni in February 2015; Operation against two teenagers in February 2015; Operation Tebas in March 2015; fifth phase of Operation Kibera in July 2015; arrests carried out in Ceuta in February 2016; Operation Sable in April 2016 as well as the detentions carried out in July 2016 in two different operations in Lleida and Girona.
 Audiencia Nacional, Sala de lo Penal, Sección Cuarta, Sentencia 39/2016, pp. 8-64; evidence obtained during the public trial corresponding to Sumario 6/2015 of Juzgado Central de Instrucción número 3, held at Audiencia Nacional, Sala de lo Penal, Sección Cuarta, December 12-14, 2016.
 Audiencia Nacional, Juzgado Central de Instrucción número 1, Diligencias Previas 68/2014, Auto de Prisión Provisional of April 10, 2015, pp. 5-6.
 Ministerio del Interior, Oficina de Comunicación y Relaciones Institucionales, Nota de Prensa of June 16, 2014, p. 3; Nota de Prensa of February 23, 2016, p. 2.
 Audiencia Nacional, Sala de lo Penal, Sección Segunda, Sentencia 38/2016.
 Audiencia Nacional, Sala de lo Penal, Sección Segunda, Sentencia 23/2015, p. 460.
 Sentencia 20/2012, handed down by the Sala de lo Penal, Cuarta Sección, of the Audiencia Nacional; Audiencia Nacional, Sala de lo Penal, Sección Segunda, Sentencia 23/2015, pp. 27, 31, 272, 275, 279, 283, 286, 413, and 419; Ángeles Escrivá, “Detenida la joven ceutí que fichó el Estado Islámico,” El Mundo, January 14, 2015.
 Sentencia 38/2016. See also Carola García-Calvo, “There is No Life Without Jihad and No Jihad Without Hijrah: The Jihadist Mobilization of Women in Spain, 2014-2016,” Madrid: Elcano Royal Institute, ARI 34/2017, pp. 8-9.
 See, in this respect, Mohammed Hafez and Creighton Mullins, “The Radicalization Puzzle: A Theoretical Synthesis of Empirical Approaches to Homegrown Extremism,” Studies in Conflict and Terrorism 38:11 (2015): pp. 966-968. Also, for a discussion on the importance of ideology in processes of radicalization leading to terrorism, Peter Neumann, “The Trouble with Radicalisation,” International Affairs 89:4 (2013): pp. 873–893.
 Quintan Wiktorowicz, Radical Islam Rising: Muslim Extremism in the West (Lanham, MD: Rowman & Littlefield, 2005), p. 22; Mohammed Hafez and Creighton Mullins, “The Radicalization Puzzle: A Theoretical Synthesis of Empirical Approaches to Homegrown Extremism,” Studies in Conflict and Terrorism 38:11 (2015), especially pp. 964-966; also Arie Perliger and Ami Pedahzur, “Social Network Analysis in the Study of Terrorism and Political Violence,” PS: Political Science and Politics 44:1 (2011), pp. 45-50.