HOW MAFIAS TAKE ADVANTAGE OF GLOBALIZATION. The Russian Mafia in Italy

Transcription

1 British Journal of Criminology Advance Access published October 24, 2011 doi: /bjc/azr077 BRIT. J. CRIMINOL. (2011) 1 of 19 HOW MAFIAS TAKE ADVANTAGE OF GLOBALIZATION The Russian Mafia in Italy Federico Varese* How do mafias operate across territories? The paper is an in-depth study of the foreign operations of a Russian mafia group. It relies on a unique set of data manually extracted from an extensive police investigation that lasted several years, including a set of phone intercepts over nine months. Using quantitative content analysis and multiple correspondence analyses (homals), the study reconstructs the activities of the group and its organizational structure in Italy. The paper shows that the core activity of the group protecting racketeering remains located in the territory of origin, while the Italian branch was monitoring investments in the legal economy. The structure of the group abroad indicates that a division of labour developed, alongside extensive contacts with local criminals and entrepreneurs. The paper contributes to broader debates on globalization and organized crime, moral panics, the structure of informal groups and the role of women in organized crime. Keywords: transnational organized crime, mafias, money laundering, Russian mafia, content analysis, correspondence analysis Introduction Recent contributions to the study of organized crime suggest that criminal groups are taking advantage of globalization and that they pose a real and present danger to Western societies(shelley 2006: 43; Castells2000; Williams2001; Scherrer 2008, quoted inwoodiwiss and Hobbs 2009). According to this view, criminal groups migrate easily to advanced market economies where they create long-lasting outposts (Shelley 1999; 2006: 43; Castells 2000: 201; Williams 2001: 71). Western governments have often hyped such a threat from transnational organized crime and have responded by creating new agencies such as the Serious and Organized Crime Agency in the United Kingdom (for discussions, see Hobbs 1998; Woodiwiss 2003; Sheptycki 2003; Levi and Maguire 2004; Woodiwiss and Hobbs 2009). The aim of this paper is to evaluate empirically whether organized crime groups are able to move abroad easily and reproduce their territorial control in a different country. Rather than using the somewhat vague notion of transnational organized crime, we limit our analysis to a specific type of criminal organization, namely a mafia, defined as a territorially based criminal organization that attempts to govern territories and markets. 1 The paper draws on an in-depth study of a Russian mafia group that surfaced in Rome in the early 1990s, and reconstructs the group s activities and organizational structure. Our conclusion is that this group did not attempt to reproduce its territorial * Department of Sociology, University of Oxford, Manor Road Building, Manor Road, Oxford, OX1 3UQ, UK; federico.varese@ sociology.ox.ac.uk. 1 Several studies have established that the Sicilian Cosa Nostra, the Ndrangheta, the Camorra, the Italian-American mafia, the Hong Kong Triads, some Russian gruppirovki such as the Solntsevskaya, and the Japanese Yakuza are essentially forms of extra-legal governance and can be collectively referred to as mafias (Sabetti 1984; Gambetta 1993; Chu2000; Milhaupt and West 2000; Varese 2001; Hill 2003). See also Varese Ó The Author Published by Oxford University Press on behalf of the Centre for Crime and Justice Studies (ISTD). All rights reserved. For permissions, please journals.permissions@oup.com

2 VARESE control abroad, and its migration was the product of an unintended consequence, namely a conflict among crime leaders in Russia. While in Italy, this cell operated a money-laundering service for the main group, and succeeded in building both a rudimental organizational structure and extensive links with local actors. After the police arrested the main players in 1997 and expelled them, the group left no traces. The paper ends by criticizing the concept of transnational organized crime and argues for a clearer distinction between isolated episodes of money laundering and long-term transplantation. Such a distinction has implications for policy making. The Movement of Mafia Groups: Two Competing Views There are two competing views on the ability of mafia groups to take advantage of globalization (Varese 2011; Campana 2011). The first perspective claims that criminal groups including mafias take advantage of globalization by forging international alliances and by creating colonies around the world. Manuel Castells argues that a new phenomenon, global crime, has now emerged, that is, the networking of powerful criminal organizations, and their associates, in shared activities throughout the planet (Castells 2000: 166). A myriad of criminal groups, ranging from the Sicilian mafia to the American mafia, the Columbian cartels, the Mexican cartels, the Yakuza, the Chinese Triads, the Nigerian criminal networks, the Jamaican Posses and many more, have come together in a global, diversified network (Castells 2000: 167). International alliances, joint ventures and subcontracting are claimed to be common features of international criminal cooperation (Shelley 1999; 2006; Castells 2000; Williams 2001). Criminal organizations not only cooperate internationally; they are also expanding into new territories. Shelley writes just as multinational corporations establish branches around the world to take advantage of attractive labour or raw material markets, so do illicit businesses (Shelley 2006: 43). A specific hypothesis can be derived from Shelley s argument, namely that, as globalization progresses, criminal groups will increasingly seek to enter attractive labour or raw material markets abroad, putting into action a rational strategy to expand abroad. For Williams also, organized crime now can migrate easily (Williams 2001: 71; see also Shelley 1999). Castells, who also describes this phenomenon, lists a number of locales where well-known mafias have opened outposts, such as Germany for the Sicilian mafia, Galicia for Columbian cartels and the Netherlands for the Chinese Triads (Castells 2000:201).In addition,these authors argue that contemporary forms of transnational organized crime lack strict hierarchies and that actors ties are distributed roughly equally. Many criminal networks are inherently dispersed, with the result that they do not provide obvious centers of gravity or loci (Williams 2001: 71). Second, the lack of centres of gravity makes it harder for law-enforcement agencies to disband such a group (Williams 2001: 71; Castells 2000; see also Shelley 2006). The competing view maintains that mafias are highly localized and stationary. Peter Reuter, while discussing US illegal enterprises (of which mafias are a subset), notes that theytendtobe localinscope (Reuter1985:21;seealsoHobbs1998).InTheSicilianMafia, Diego Gambetta writes that not only did the [Sicilian] mafia grow mainly in Western Sicily, but, with the exception of Catania, it has remained there to this very day. More generally, the mafia, he adds, is a difficult industry to export. Not unlike mining, it is heavily dependent on the local environment (Gambetta 1993: 249, 251). Similarly, Yiu Kong Chu argues that Hong Kong Triads are localized and they are not international illegal entrepreneurs whose wealth and connections may enable them to emigrate to 2

3 HOW MAFIAS TAKE ADVANTAGE OF GLOBALIZATION Western countries (Chu 2000: 130). Peter Hill, the author of The Japanese Mafia, dates the beginning of the foreign activities of yakuza from the 1970s, when the Ishii Susumu branch of the Tokyo-based Inagawa-kai started to organizegambling trips for wealthyjapanese. Yet, it seems that the yakuza s foreign travel was mainly either recreational or focused on investing in real estate; at no time have the yakuza managed to extend their core protective role beyondanativejapanesemarket (Hill2002: 53). BothReuter(1985) andgambetta(1993) suggest that the apparent difficulty faced by the Mafia in relocating or expanding its business reflects the high cost of monitoring agents that such transplants entail, as well as the difficulties of building a reputation and collecting reliable information in the new territory. Secrecy and illegality make it harder for criminal groups to build the complex systems of supervision that exist in legitimate firms. The paper will evaluate these two perspectives by considering an attempt by the Solntsevskaya crime group to open an outpost in Rome in the mid 1990s. We continue by considering the evidence. The next section presents background information on the Solntsevskaya. The third section describes the data, while the fourth section documents the activities of the Solntsevskaya in Rome and the fifth section reconstructs the internal structure of the organization. The last section draws the main conclusions. The Solntsevskaya in Moscow and Rome The Solntsevskaya crime group founded by Sergei Mikhailov in the mid 1980s is named after a Moscow neighbourhood. 2 An FBI report describes the organization as the most powerful Eurasian organized crime group in the world in terms of wealth, influence and financial control (Moscow Times, 22 February 2000). The group runs a protection racket throughout Moscow, termed roof (krysha) in Russian, controlling access to valuable markets (Modestov 1996; Maksimov 1997; Finckenauer and Waring 1998; Varese 2001). Estimates of the size of the group (possibly exaggerated) range from 5,000 to 9,000 members, distributed across ten semi-autonomous brigades (brigady), all operating under the umbrella name of Solntsevskaya. The Russian police claim that the group controls various banks along with around 100 small and medium-sized enterprises. According to several witnesses and police reports, the Solntsevskaya maintains a common fund (obshchak), which is reinvested in the legal economy. Although little is known about the inner workings of the group, former members have claimed that the organization is governed by a council of 12 individuals, who meet regularly in different parts of the world, often disguising their meetings as festive occasions. These individuals have all gained the title of vor-v-zakone (i.e. Boss, literally criminal-who-abides-by-the-code) through an initiation ritual and have spent lengthy periods in prison. Women have no place in the hierarchy and are excluded from top-level decision making as well as the day-to-day running of the groups (Modestov 1996; Maksimov 1997; Varese 2001; Gilinsky 2007). In 1994, squabbling among the leadership led to the murder, in Moscow, of one of the original leaders, Sergei Sil vestr Timofeev. As a consequence, some of his allies began to fear for their lives. One of them, Ivan Yakovlev (fake names are italicized when first mentioned), a member of the ruling council and a vor-v-zakone, decided to move to a town on the Lazio seaside near Rome with part of his crew (in a conversation recorded 2 This description draws in part upon Varese (2011). 3

4 VARESE by the police shortly after they moved to Italy, one of Yakovlev s accomplices says: Even the police says that only after all the followers of Sil vestr (the Silvestrovskie) are killed, will there be peace, Servizio Centrale Operativo 1997b: 31). In a relatively short time, Yakovlev s group forged connections in his new hometown with a vast swathe of the economic elite as well as with several politicians and officials. Yakovlev already knew at least two Russians in Italy. Boris Sergeyev, who was to become central to Yakovlev s operation in Italy, was a businessperson with connections to the Solntsevskaya group that dealt in oil products originating in Chechnya. Sergeyev had been involved in some money-laundering operations while living in Vienna. Another collaborator of Yakovlev s, Oleg Bogdanovich, had relocated to the same seaside Lazio town slightly earlier, in A small-time crook, Bogdanovich spoke Italian and would become a trusted gofer of the boss (Commissione Parlamentare Antimafia 1997: 18 19). One of the Russians who moved to Italy with Yakovlev was his closest and most trusted associate, Fedya Sidelnikov, who had previously been the bodyguard of Sil vestr Timofeev. Sidelnikov s wife, Raisa Artamova, became responsible for managing the common fund ( the pot ) in Italy for the group (Servizio Centrale Operativo 1997a: 180; 1997c : 37; Commissione Parlamentare Antimafia 1997: 20). Back in Moscow, Yakovlev s main crew included Konstantin Rostovtsev, a former KGB officer who was fluent in Italian and had close contacts with the Italian embassy in Moscow as well as with the political elite in Russia. Rostovtsev had been collaborating with Yakovlev for eight years; he was involved in arranging visas and passports as well as trips to Moscow by Yakovlev s Italian contacts. Sergei Vlasik, also based in Russia, was a businessperson who had worked for Yakovlev as the clean face of the organization and was a financial wizard. He was later killed (Servizio Centrale Operativo 1997a: 81 96). A key local helper was Giorgio Facchini, a former low-level official at the Rome headquarters of the Italian Communist Party, who was married to a Russian, Alina Kamenskaya. At the time of his employment in the Communist Party bureaucracy, Facchini organized trips and cultural exchanges between Italy and the USSR. After being dismissed by the party in the mid 1980s, he continued to be involved in tourism throughout the 1990s. Another critical player in Yakovlev s Italian crew was Giovanni Pepe, Facchini s assistant. Finally, mention should be made of Riccardo Brenno, the former deputy chair of a state-owned company (he had been appointed by the Cristian Democratic Party). Brenno emerged as the most high-profile Italian involved with the group a man whose connections extended to the highest echelons of the business and political elite. Key players are listed in the Appendix. In the next section, we describe the data we have collected on the group and the methods used. Data, Methods and Validity In 1995, the Italian police started an undercover investigation of the Solntsevskaya cell in Rome, which lasted three years ( ) and resulted in a three-volume report of some 2,000 pages that was later used during the trial. 3 The document contains hundreds of 3 While we were working on an unrelated project on Russian organized crime, a journalist alerted us to this investigation and gave us the relevant material. It should be noted that the case collapsed in court because it transpired at the trial that the police did not have the correct authorization to collect evidence, especially wiretapping evidence. This technicality prevented the prosecutors from using the material that had been collected, and the court case was dismissed. The Russian protagonists of this story were nevertheless expelled from Italy and later resurfaced in Moscow. Thus, the reader must remember that the people discussed here have not been convicted of any crime in Italy. In order to protect their identity, all names have been changed. Some place names have also been changed. 4

5 HOW MAFIAS TAKE ADVANTAGE OF GLOBALIZATION pages of transcripts of telephone conversations between Italian and Russian criminals, details of international money transfers and financial operations involving banks in London, Budapest, Vienna, New York and Rome, information about property speculations in Italy, and individual files on some 45 people. The files on individuals contain information concerning his or her various aliases, passport details, residency, travel, ownership of real estate and information collected from other police. These people include several members of parliament in both Italy and Russia, political leaders in Russia, businessmen, members of the armed forces and embassy officials. Regarding the transcripts of conversations, 164 individuals are involved, of which more than 70 per cent are men (N = 115), and the overwhelming majority were born either in the former Soviet Union (65 per cent, N = 107) or in Italy (30 per cent, N = 50). Additional nationalities are Polish (two) and Korean (two), while the nationality of three individuals is unknown. Place of birth may, however, be a poor indication of where the actors were actually operating at the time of this investigation. We could reconstruct this type of information for 161 out of 164 actors. These data confirm that most of them operated in either Rome (47.6 per cent) or Moscow (20.5 per cent). The remaining 32 per cent were based mainly in other cities in Italy and Russia. The median year of birth is 1957 (39 years old in 1996). This suggests that the group was made up of rather experienced individuals an age profile extremely different from, for example, that of the much younger Neapolitan Camorra, and more akin to the older Sicilian mafia (Gambetta 1993: 104; Catino 1997). In order to identify the activities and the organizational structure of the group in Rome, we undertook a content analysis of the conversations contained in the file (for similar analyses, see Natarajan 2000; 2006; Morselli 2009; Campana 2011). We developed a coding scheme with four items ( tasks ): economic investments (EI) are conversations on business investments carried out in the legal and illegal economy; group management (GM) pertains to the day-to-day management of the group, including the remuneration of the workforce and the activities of monitoring, intimidation and punishment of the group s members; protection activities (PA) refers to efforts to control markets and offer criminal protection; and resource acquisition (RA) refers to discussions about the acquisition of specific input resources to run the group, such as information, equipment or labour. 4 Since the conversations are informal, and jargon is often used, we decided to read and code each conversation manually. We assigned a task to each conversation. When a conversation contained discussions of more than one task, we chose the most talked about task as measured by the number of words. 5 In addition, for each individual, we coded how many conversations s/he devotes to which task. Thus, we have two measures: an overall number of conversations per task and an individual-level measure of how many conversations were devoted to which task for each actor. One can then compare the number of conversations devoted to a given task by an actor with the overall number of conversations devoted to it by all actors. In order to reduce coder bias, we undertook intercoder reliability tests, which resulted in a very high and statistically significant 4 A typology first advanced by Dunning (1993), and subsequently modified and expanded by Zaheer and Manrakhan (2001), focused on resource-seeking, market-seeking and efficiency-seeking motivations. 5 Eighty-one conversations (10.7 per cent of the total) contained conversations on more than one task. We have coded these segments of conversations as well, in order to capture the data that we miss by coding only the main task per conversation. The second procedures yielded a percentage distribution that is very similar to those reported in Table 1 EI: 52.2 per cent; GM: 19.8 per cent; RA: 22.1 per cent; PA 5.6 per cent. 5

6 VARESE correlation coefficient between the two coders. 6 This degree of concordance enhances our confidence in the reliability of the results. While studying the files, we decided to use only the conversations from January 1996 to October 1996 (N = 758). This is because wire tapping started in earnest in January (only two conversations were recorded in 1995). 7 The end point is late October 1996, a few weeks after Sergeyev was killed. At that point in time, the Italian police wrapped up the investigation and sent the results to the prosecutor. The main players were subsequently arrested. We also coded quantitative information on the group. From the text, we have extracted information on some 250 individuals. For each individual, we coded information on 20 variables, including first name, last name, nickname(s), date of birth/death, city and country of birth, gender, nationality, date of entry into Italy, place of residency, language(s) spoken and criminal records. Potential threats to validity As famously pointed out by Reuter and Rubinstein, collecting information about people who are consciously involved in illegal activities is hard (Reuter and Rubinstein 1978: 57). Scholars of organized crime have tried their best using different methods of data collection. Some immerse themselves in a field and engage in direct observation and interaction with the people they study (e.g. Hobbs 1988; Jankowski 1991; Adler 1993; Kaminski 2004). Beside the fact that this method is risky (Jankowski 1991: 11), it is next to impossible to penetrate the high echelon of crime groups (Natarajan 2000). Others have conducted retrospective interviews with traffickers confined to prison (e.g. Reuter and Haaga 1989). This strategy of data collection also has drawbacks; for example, interviewees may have a strong incentive to minimize or overstate their roles. Furthermore, imprisoned high-profile criminals who agree to be interviewed are unlikely to be representative of the broader population under study (Natarajan 2000: 75). Wiretap data are an additional source of data (Reuter 1994). In several jurisdictions (e.g. the United States, Canada, Germany, France, Italy, Holland and Sweden), wiretaps can be introduced as evidence in trials. Since they are to be used as evidence, the conversations are transcribed by officers and made available to the prosecutor. Wiretapped conversations capture what actors talk about in their day-to-day business life, including conversations involving lower-level and upper-level actors, both within the gang as well as in society at large (politicians, corrupt officials, businesspeople). Scholars have started to utilize this source systematically (Baker and Faulkner 1993; Finckenauer and Waring 1998; Natarajan 2000; 2006; Morselli 2009; Campana 2011). Yet, issues of validity do arise. There are at least four potential threats to the validity of phone wiretaps as a source of data for reconstructing the activity and structure of a criminal network (for a broader discussion, see Campana and Varese 2011; Varese 2011). First, it might be the case that 6 Two coders read a sample of the conversations (N = 20), identified the tasks and checked the level of inter-coder reliability. The correlation coefficient between the two coders resulted in 0.95, p = Most of the conversations (97.5 per cent) took place without an interpreter between the two speakers. 7 The prosecutor who started the inquiry informed us that a tip had been received about the presence of the Russian boss, Yakovlev, in Italy from American prosecutors, who were interrogating a high-profile member of the Solntsevskaya arrested in Italy in March 1995 and handed over to the American authorities. In December 1995, phone lines and cellphone numbers were still being added for monitoring purposes and the police recorded very few conversations (Prosecutor in the case against [Yakovlev] et al., Rome, March 2007). 6

7 HOW MAFIAS TAKE ADVANTAGE OF GLOBALIZATION those who are intercepted are aware of the surveillance, and so talk strategically on the phone in an attempt to mislead the police. If this were the case, we would expect them not to talk about serious crimes that they intend to carry out. A second threat to validity is that the suspects targeted for wiretapping may not be representative of the network as a whole (Natarajan 2000: 293): in other words, people who might be central to the network may not appear central in the data simply because their phones were not targeted. A related threat is the extent to which the conversations included in the file constitute a representative sample. This reflects a general issue often raised in the study of informal groups, namely that we cannot be sure where the boundary of the network lies (Lauman et al. 1983; Scott 2000: 53 62). We address these risks in turn. First, in our data, the protagonists certainly discuss incriminating matters on the phone, including the use of violence and torture, and the details of murders that they plot and subsequently carry out. If they talked strategically, we would expect them to reveal less and to omit information that could result in a lengthy prison sentence. Furthermore, the police itself validated the content of some conversations by tailing several individuals and checking their movements, both within Italy and in trips to and from Italy. Second, by 1996, the police were consistently listening in on 24 phone lines (cellphones and terrestrial lines). A total of 19 people used such lines (see the Appendix for a full list). The police came to the conclusions that these 19 individuals were the key players in the Italian-based group. Thus, this paper can offer a picture of the activities carried out in Moscow only in so far as they are reflected in the conversations of the Italian branch. In any case, the focus of this work is the activities of the group in Italy. After reading the data, we have no reasons to believe that some key players based in Italy were missed by the police phone surveillance operation. Third, are the recorded conversations to which we gained access a representative sample of the total? In order to answer this question, we carried out background interviews with the police officer in charge of the investigation and the public prosecutor in the case (Prosecutor in the case against [Yakovlev] et al., Rome, March 2007; Police Officer in the case against [Yakovlev] et al., Rome, March 2007). They confirmed that some parts of the conversations were deleted, as they did not refer to breaches of Italian law. Hence, a selection was made by the police, according to their interpretation of Italian law. However, it should be noted that such a definition would still have represented a wide net. For example, discussions of crimes committed in another country still qualify as a breach of Italian law: if a suspect discusses arranging to kidnap another person in Russia, he would still be prosecuted under Italian law. Fourth, the network under investigation had activities in Russia that were not captured by the police wiretaps. As mentioned above, our analysis is limited to the activities while in Italy, although there are enough conversations with interlocutors based in Russia to reconstruct a significant set of activities of the group in both countries. In conclusion, we are confident that this rather large set of conversations is a reliable proxy for the activities of the group in Italy and offers us a unique insight into their behaviour. The Activities of the Group in Italy What did the Russians do in Rome? Table 1 presents the data of the content analysis of conversations recorded by the police over the selected period. Slightly more than half of conversations (51.7 per cent) concerned economic investments (EI). Group management 7

8 VARESE TABLE 1 Distribution of tasks discussed in conversations Tasks Conversations Frequency Percentage Economic investments (EI) Group management (GM) Resource acquisition (RA) Protection activities (PA) Total (GM) and resource acquisition (RA) are the next most frequently discussed (22.8 and 18.9 per cent, respectively), while discussions of protection activities (PA) are relatively infrequent (6.6 per cent). These data can be interpreted as follows: the average actor devotes these percentages of her/his conversations to the four tasks. Since we know how many conversations each actor actually devoted to which task, we can assess whether s/he devotes more or less conversations to a given task than the average actor. Economic investments (EI) Many conversations recorded by the police were devoted to identifying investment opportunities, mainly in Italy, and exporting goods to Russia ( for a broader discussion, see Varese 2011). The source of the money to finance these ventures was Russia. The money arrived in Italy via an extensive network of individuals travelling from Russia to Italy carrying cash or through bank transfers to various companies all based at the same address in Rome that had been set up by Yakovlev s associates in Italy (Servizio Centrale Operativo 1997a: 85). The money from Russia was invested in a diverse array of import export ventures. The police file records no fewer than 100 business deals ranging from purchasing sparkling wine to buying frozen meat, wood, furniture, wheat, steel, pharmaceutical products, Olivetti computers, food products (including fish), Armani suits, works of art, gold, helicopters and antennas (Servizio Centrale Operativo, 1997b: 98). The common feature of these ventures is that the money used to pay Italian suppliers came from criminal activities outside Italy, in breach of Italian law (e.g. Servizio Centrale Operativo 1997a: 132). From several conversations, it transpires that the money being imported illegally into Italy, and more generally at the disposal of Yakovlev for investment in the legal economy, was not solely the fruit of the criminal activities of his own crew. For instance, in one conversation, Yakovlev refers to other Russian criminal groups as having contributed money to buy a fishery (Servizio Centrale Operativo 1997a: 163). Similarly, it emerges that some money entrusted to a Russian businessman close to the group, Vlasik, to invest in Korea, belonged to at least three different groups (Servizio Centrale Operativo 1997c: 40 1). Yakovlev devotes almost 50 per cent of his conversations to investments in the economy and Sergeyev (Yakovlev s business partner, who was later killed) devoted 67 per cent of his conversations to this task. In contrast, Sidelnikov, Yakovlev s enforcer, hardly speaks about this topic. Among the Italians, Pepe, the local helper, devotes 62 per cent of his conversations to this task, while Brenno a high-profile Italian businessman who works for the group devotes 97 per cent. Another indicator of whether the activity was directed at Italy or Russia is the country of origin and country of destination of the call. Seventy-five per cent of the calls originated in Italy, while Russia is the country of origin for 16 per cent of the calls. Italy receives 65.5 per cent of the calls, Russia 23.5 per cent. This suggests that 8

9 HOW MAFIAS TAKE ADVANTAGE OF GLOBALIZATION most of the talking is done within Italy, with some significant conversations conducted to and from Russia, since goods were eventually going to reach that country. These data suggest a rudimentary division of labour within the group, with some actors, such as Brenno, almost exclusively devoted to helping the group fulfil the aim of finding investment opportunities, while the time of others, as shown below, is more evenly spread across the various tasks. As we shall also see below, a third sub-group seems focused exclusively on protection activities and group management. Group management (GM) Any crime boss has the task of maintaining internal order, ensuring that his underlings do not disobey or report him to the authorities and, more generally, that members of his crew do not cheat on him. The mere threat of violence should be sufficient to ensure compliance. This is the case with Ferrari, a Roman petty criminal, who had misbehaved and even threatened to report Yakovlev to the police unless Yakovlev repaid him some money that he owed. Yakovlev s wife, Marina, directly threatened Ferrari, whose wife is Russian, for disobeying Yakovlev ( we will chop your head off ). Ferrari s wife calls back the boss s wife: Marco [Ferrari] begs you for forgiveness, he understood that he was wrong. It should not be Ivan that calls for Marco but he himself should show up and be ready for work, because Ivan is bigger than him, he has been in prison for 15 years. (Servizio Centrale Operativo 1997a: 99) Ferrari was then eased out of the organization rather than physically punished, possibly because any punishment would have attracted the attention of the Italian police (Servizio Centrale Operativo 1997a: ). One would imagine that the ability of such a powerful organization as the Solntsevskaya to inspire fear and secure compliance would be sufficient to ensure obedience from its accomplices (Gambetta 1993). Yet, mafia-type criminal organizations might have to use more violence than one would expect (Pistone 1997). Vlasik, a Russian who had worked for Yakovlev as the clean face of the organization and a financial wizard, had been entrusted by Yakovlev with a considerable sum to invest in a deal in Korea (Servizio Centrale Operativo 1997c: 40). Although Vlasik must have known the risk he was taking, he embezzled the funds and tried to disappear. He was eventually found in Korea and escorted back to Italy, where he was held in captivity and tortured for some weeks (Servizio Centrale Operativo 1997a: 25 6, 164). Later, Vlasik was put on a plane to Moscow accompanied by a jailer. While in Russia, he managed to escape and gave himself up to the police. The punishment was only delayed, however, and Vlasik was later killed (Servizio Centrale Operativo 1997a: 24, 163; Commissione Parlamentare Antimafia 1997: 19). 8 The group was also involved in a second murder. A conflict developed between Yakovlev and Sergeyev. Allegedly, Sergeyev had also stolen money from the group. He was murdered in Moscow in 1996 (Servizio Centrale Operativo 1997a: 52, 35; see also Commissione Parlamentare Antimafia 1997: 19). This second murder had a severe 8 At first, Yakovlev was hesitant about killing Vlasik. In a conversation with Yakovlev, Sidelnikov (one of Yakovlev s deputies) said he wanted Vlasik killed right away, but Yakovlev demurred: No, better talk to him, tell him that he should think twice before deciding to cut loose from his own friends. Sidelnikov agreed. Yakovlev had also planned to kidnap and torture Rostovtsev s secretary (Servizio Centrale Operativo 1997a: 26). 9

10 VARESE destabilizing effect on the group, as it attracted significant police attention. It should be noted, though, that severe acts of violence were conducted in Russia rather than in Italy. The overall percentage of conversations dealing with Group Management is 23. The boss and the two Russian enforcers are those mostly associated with this task: Yakovlev, Sidelnikov and Sidorov devote to this task, respectively, 28, 44 and 45 per cent of their conversations. Most remarkably, it transpires that women were directly involved in group management, contrary to the norm in Russia. For instance, Yakovlev s wife, Marina (who devotes 53.3 per cent of her conversations to this task), directly threatens Ferrari, whose wife is also Russian, for disobeying Yakovlev. Alina Kamenskaya, the wife of Italian helper Facchini, on several occasions advocates appropriate punishments for those who have misbehaved towards Yakovlev (e.g. Servizio Centrale Operativo 1997a: 53; 38 per cent of her conversations deal with group management). Artamova, the keeper of the common fund, is also very involved in conversations on group management(62 per cent of her conversations). Two women around Sergeyev, who will be killed in 1996, spend a considerable amount of time talking about this task: they are Kostina (58 per cent of her conversations) who had entered into a fake marriage with him and worked as his interpreter and overall helper in Italy and Sergeyev s real wife, Nadia (33.3 per cent of her conversations). On the whole, women, Russian enforcers and the boss seem to talk about group management more intensively. Severe acts of violence were rarely carried out in Italy and murders were committed only in Russia. Resource acquisition (RA) Acquisition of resources for an ordinary firm could refer to the acquisition of inputs such as labour, intangibles such as knowledge, innovation capabilities, and management or organizational skills. For illegal firms, resource seeking refers to intelligence and information, labour and technical equipment such as arms, hideaways and spying technology. Shelley (2006) argues that mafia groups migrate to acquire valuable input resources to run the organization back home. In this case, the resources discussed in the conversations that we studied were rather instrumental in running the group in Italy, not as input resources for the group back in Russia. The group spent a great deal of time trying to open bank accounts, acquire Italian citizenship through fictitious marriages and on seeking to obtain forged driving licences and guns. In order to obtain some of these resources, they offered bribes to local officials (Servizio Centrale Operativo 1997a: 120, 89, 51 2). Overall, almost 19 per cent of conversations were devoted to resource acquisition. The individuals most involved in conversations about resource acquisition were the Italian helpers and Bogdanovich, a close collaborator of Yakovlev who spoke Italian and had been living nearby Rome since (Bogdanovich, Facchini, Facchini s Russian-speaking wife and Pepe devote, respectively, 55, 52, 45 and 23 per cent of their conversations to this task). The keeper of the common fund speaks mainly about group management (62 per cent) but her second most talked-about task is resource acquisition (31 per cent). Yakovlev speaks about this topic relatively infrequently (11.2 per cent of his conversations are on this task), while the percentage of Sergeyev s conversations devoted to this task is the same as the overall number (19 per cent). More than 80 per cent of conversations that have resource acquisition as their main task are conducted within Italy, and only 16 per cent between Italy and Russia. 10

11 HOW MAFIAS TAKE ADVANTAGE OF GLOBALIZATION The key finding is that the resources discussed were used to run the group in Italy, rather than as inputs for operations in Russia. As for economic investments, a rudimentary division of labour emerges, with the Italian fixers and a Russian go-between involved in acquiring resources, while the enforcers (Sidelnikov, Sidorov and some women) and the businessmen (such as Brenno) are not involved in this task as much. As one would expect, mostoftheconversationsarewithinitalybecausetheresourcestobeacquiredareinitaly. Protection activities (PA) Since protection racketeering is the core business of the Solntsevskaya, if their motivation for migrating to Italy was to expand their core activity, then we would expect the group to engage in protection activity in the new location. On the contrary, none of the recorded conversations refers to protection activities in Italy. Rather, they offer an insight into racketeering in Moscow and, to a lesser extent, in other cities such as Prague. Often, conversations about this task are prompted by a delay in payments or interference from other racketeers. For instance, the owner of a shop in Moscow is threatened by Yakovlev if he does not pay US$25,000 by the following Sunday. In another case, a Moscow wholesale dealer of food products who is under the protection of Yakovlev (to whom he pays US$200,000 a month) is being harassed by some officials who have requested a sum of one million Russian roubles. In a similar situation, in July 1996, the boss receives a call from Prague in which the caller explains that he has been harassed by certain gypsies who were demanding protection money from him. (The caller says: I had problems with those gypsies, who have beaten up a friend of mine. They asked me who my vor [protector] is and I gave your name. They said that you should be here, Servizio Centrale Operativo 1997a: 188 9). The two actors who are most involved in discussing protection activities are Yakovlev and Sidelnikov, well above the 7 per cent average for the group (Yakovlev and Sidelnikov devote respectively 14 and 22 per cent of their conversations to this task). Also, most of the calls are between Italy and Russia (83 per cent) and only one call is within Italy, suggesting that the activity in question is confined to Russia and the Russians in Rome need to call home to manage this aspect of their criminal activities. Figure 1 presents the tasks discussed in all conversations over almost a year: from January 1996, when the Italian police started systematic phone intercepts, to October 1996, when Sergeyev was killed. On the vertical axis, we plot the number of conversations and, on the horizontal axis, the months of the year. The only correlation that is statistically significant is between Resource Acquisition and Economic Investments (r = 0.758, p = 0.011). That is, discussions of investment opportunities in the economy co-vary with discussions of acquisition of resources, suggesting a relationship between the two sets of tasks. This further confirms that discussions on resource acquisition are closely connected to discussions about money laundering and illegal investments in the Italian economy. In contrast, references to protection racketeering are not cyclical and are unrelated to other tasks. As mentioned above, racketeering was carried out in Russia by members of the group a task unrelated to their presence in Italy (This discussion is also in Varese 2011: 83). Structure of the Group The analysis of the conversations has suggested that the individuals involved in the group undertook different roles within the organization. Some are more involved in acquiring resources for the group (Bogdanovich, Facchini, Facchini s Russian-speaking wife and 11

12 VARESE FIG. 1 Tasks discussed in conversations, January October 1996 Pepe), while others work as enforcers and discuss internal matters and punishments (Sidelnikov, Sidorov and some women). A third group consists of businessmen (such as Brenno), while the boss Yakovlev seems to be busy with several tasks except resource acquisition. This section explores systematically the internal division of labour in the group, namely who does what. This analysis will help us to evaluate empirically whether contemporary forms of transnational organized crime lack strict hierarchies and centres of gravity, as suggested by Williams (2001: 71; see also Castells 2000; Shelley 2006), and corroborate the preliminary findings presented in the previous section. A clear division of labour emerges from the multiple correspondence analysis (Homals) we have undertaken of the relationship between some key players and the tasks (Figure 2) and between principal task and sex, nationality (place of birth), place of residence and language spoken (Figure3). 9 Multiplecorrespondence analysis is a model-free techniqueused to establish the strength between categorical variables and represent them in a geometrical space (Gifi 1990). Despite a fairly long history dating back to the 1930s (Richardson and Kuder 1933; Hirschfeld 1935; Clausen 1998), correspondence analysis has been rediscovered in the early 1960s by French scholars (Benzécri 1973) and used by, among others, the renowned French sociologist Pierre Bourdieu (1979) and more recently by French-American criminologist Bertrand Harcourt (2006: 26 30; see also Mohr 1998). This method of data analysis allows us to represent the relationships between features of the conversations (such as the task discussed) and features of the actors (such as their identity, gender, place of residence). It transforms rows and columns of a data matrix into points of a two-dimensional Euclidean space (Greenacre 1984: 54) and it places categories that are associated close to each other. No pre-existing model is tested, but rather patterns emerge inductively. While the technique is able to display patterns of relations in a space, it does not provide correlation coefficients between specific variables. However, 9 The eigenvalues for Figures 1 and 2 are, respectively, and 0.773, and

13 HOW MAFIAS TAKE ADVANTAGE OF GLOBALIZATION FIG. 2 Correspondence Analysis of Actor and Main Task*,* Main Tasks: EI = economic investments; GM = group management; PA = protection activities; RA = resource acquisition. the existence and the strength of a relationship between two variables can be established by drawing a line from the points of interest to the origin of the principal axes. For instance, in Figure 2, a line can be drawn from the point representing the boss s wife, Yakovleva, and the origin, and another line between the task group management and the origin. The angle between the two lines (from Yakovleva and the origin and from group managementandtheorigin) isacute;thus,therelationshipbetweenthetwovariablesispositive(the scalar product is positive). We can then report that Yakovleva is significantly involved in the task of managing the group, relative to all other actors in the same space. If the angle were a 90-degree angle, it would mean that there is no relationship (the scalar product between the two vectors is zero); if the angle were obtuse, the relationship would be negative (the scalar product is negative). The longer the line between the points and the origin, the stronger the relationship (Greenacre and Hastie 1987; Gifi 1990). Figure 2 represents the relationship between some key actors and their tasks. In the topright-handquadrant,wefindgroupmanagement(gm)andprotectionactivities (PA) both tasks requiring the use of violence, and some key Russians, such as the boss, his two right-hand men (Sidorov and Sidelnikov) and the boss s wife (Yakovleva). In the top left-hand quadrant, the businessmen of the group are clustered (Brenno, Ashin, Manin, Sergeyev), while the task associated with the acquisition of resources in Italy is in the bottom right-hand corner, where the local Italian helpers (Facchini), their wives (Kamenskaya) and the Russians who speak Italian are clustered (Bogdanovich, Rostovtsev). Artamova, who keeps the common fund, is in between two tasks, acquiring resources and group management. Pepe is also half-way between helping the group 13

14 VARESE FIG. 3 Correspondence analysis of gender, nationality, residence, language spoken and main task acquire resources and discussing investments in the economy. The boss is in a central position, slightly closer to group management and protection activities in Russia, but not far from discussions about investments in the legal economy. In sum, Figure 2 represents the internal division of labour of the group and is derived empirically. In our view, it is significant that women are involved in managing the internal discipline; and that the boss has a central role, contrary to the prediction of Williams (2001); the presence of a rudimentary structure, with those who have access to violence (or can influence the use of violence, such as the boss s wife) having a distinctively different job from those who deal with investments and resource acquisition. This systematic analysis confirms the suggestive findings of the simple content analysis reported above. Figure 3 aims to map general features of the entire group and the tasks. In the top right-hand corner, we find Italians who are highly related to investments in the economy; resource acquisition is highly related to being resident in Italy, and speaking both Italian and Russian; while the tasks that involve the possible use of violence (group management and protection activities) are associated with Russians, Russian residents, males and women. Once again, we see a clustering of only a subsection of the group near the activities that involve the use of violence. Conclusions The paper has shown that the Solntsevskaya did not move its protection racket to Italy a finding that is consistent with the predictions of Reuter (1985) and Gambetta (1993). Its core business remained highly dependent on the territory of origin. Yet, the group did expand abroad: why and to do what? The hypothesis that criminal groups globalize 14

15 HOW MAFIAS TAKE ADVANTAGE OF GLOBALIZATION because they are searching for attractive labour or raw material markets (Shelley 2006) is not confirmed by this case study. More generally, many scholars assume that a mafia group, like an ordinary company, rationally decides to open a branch in a distant territory that is directly dependent on the homeland (Shelley 2006; Campana 2011). This paper finds a different mechanism of mafia diffusion: a mafioso is in a new territory because he escapes to it. Once there, he explores local opportunities. He does not receive a salary from back home; thus, the analogy with a firm opening a branch abroad is misleading. The homeland may do no more than bestow a seal of approval on the existence of the outpost, which has a great deal of autonomy. The outpost abroad is not a branch of a corporation, but rather a semiautonomous unit allied to the homeland. Contrary to what Sterling (1995) and Castells (2000) maintain, we find no evidence of international alliances with other criminal groups, but a vast network of local contacts was forged and maintained in the new locale. The paper contributes to the debate regarding the role of women in organized crime (for this debate, see, e.g. Sibert 1994; Principato and Dino 1997; Pizzini-Gambetta 1999; Schneider 2006; Ingrascì 2007; Fiandaca 2007). Women s involvement changes from home country to new territory. In Russia, women would not hold prominent official roles nor be consulted when important decisions are made (Gilinsky 2007: 230 1). Abroad, the picture is very different. The only official role in the group is given to a woman, Artamova, who controls the group s budget. In Italy, women do not use violence directly, but do discuss the use of violence and the appropriate level of punishment, and are entrusted with funds, acquiring informal and formal authority. We contend that women are allowed to play a greater role in the new setting because they can be trusted and trustworthy individuals are in short supply in the new setting. Thus, kin relations operate as a functional substitute for trust within informal and criminal organizations, especially while the group is trying to get established in a new territory. Mafia groups may change their activities across territories (Campana 2011). In Moscow, this group sought to govern territory and markets exclusively, and entered into direct competition with the state as a supplier of protection. Violence and threats were most often directed at individuals beyond the membership and associates of the group. In Italy, the group took on a different function. The recognition that mafia groups may engage in different crimes while abroad has conceptual and policy implications. First, the concept of transnational organized crime is too broad a concept to capture these differences. A more focused concept is that of mafia, which applies to crime groups that attempt to govern territories and markets. We have classified their activities as four tasks, one of which is investing profits in the legal and illegal economy. One policy implication of such an approach is that fighting money laundering abroad is unlikely to threaten the survival of a mafia group in its original territory. In addition, money laundering on its own will have a limited impact in a large economy and should not be billed as a fully fledged mafia invasion (Woodiwiss and Hobbs 2009). On the other hand, laws that make it difficult to trace illegally obtained funds will strengthen the mafia in its territory of origin One such law (Decreto Legge 194/2009) has been passed recently by the government led by Silvio Berlusconi in Italy: the law allows for individuals who illegally channelled money abroad to be allowed to repatriate the funds, after paying a small fee. The key provision is that such individuals can retain their anonymity and do not need to declare how they obtained the money in the first instance (Tabellini 2009). Thus, money obtained through organized crime activities and invested abroad can be cleaned and reintroduced in the country of origin, further strengthening the mafia at home. 15