Printer Friendly

Dark networks as problems.

Most of the literature on networks and collaboration is quite positive. Collaborative networks are seen as appropriate devices to tackle public management problems and to successfully coordinate political, social, and economic action. Figure 1 portrays a familiar relationship between public organizations and problems. Each organization (O1, O2, O3, and O4) intersects with only part of the problem, and none of the organizations is linked in any kind of a collaborative relationship (Hjern 1992, 4). This figure captures the rationale for public sector networks to collaborate to solve problems such as homelessness and health care delivery that cannot be contained within the boundaries of one organization. Because the problem is bigger than any single organization, collaborating with other organizations is necessary if there is any hope of making progress in alleviating the problem.

[FIGURE 1 OMITTED]

What writers on networks have widely ignored is the nature of the problem. The focus among network scholars, ourselves included, has been to try and figure out how to structure collaboration in a way that leads to problem amelioration. It is usually argued that networks, compared to hierarchies, are a better solution to solving nonroutine, nonstandardized, or even ill-structured (Simon 1973) or wicked (Rittel and Webber 1973) problems. These problems, however, were more or less constructed as socially nonreactive, and if there was political resistance to political measures, it was assumed that it was possible to co-opt those groups or organizations in some way or another.

But what happens if the problem is not an unorganized set of poor or cognitively impaired clients but another network, perhaps engaged in illegal activity? If this is the case, then the problem in figure 1 could be labeled "drugs," "terrorism," or "arms trafficking" and the organizations could be labeled "police," "immigration," "intelligence," or "the military."

While the organizations in figure 1's pentagons attempt to alleviate the problem, inside the problem space is a set of actors (both individuals and organizations) bent on making the problem worse. In the case of arms trafficking, there are arms dealers eager to sell AK-47s to conflict zones all over the world. There are shipping and air transport services willing and able to transport the weapons to warlords who are destabilizing much of West Africa. The warlords trade diamonds for guns, and the arms brokers and transporters in turn launder the diamonds with brokers in Antwerp and deposit the money in Swiss banks.

Viewing the figure this way brings the realization that there is a set of individuals and organizations that constitute a network striving to achieve ends that create collective-action problems for governments all over the world. The actions of those fighting against drug trafficking to tighten border security create incentives for drug lords to build tunnels under the border to try to evade the increased security. In both cases there is the need for collective cooperation if individual goals are to be achieved. It is also possible that the actors in the problem space are linked in a network as well. This network may be strictly a terrorism network, or it may be connected to other illegal networks such as arms smuggling, money laundering, and drug dealing. These connections among illegal networks in the problem space can be critical to the success of their wicked ends.

There may also be connections between the illegal networks, dark networks in our terms, and the legal networks striving to destroy them. In the world of illegal activity, a CIA agent who penetrates a terrorist network is called a double agent. Figure 2 captures the reciprocal relationships among the illegal networks in the problem space and the network of linked legal organizations attempting to control and curtail their activity. It also captures the possibility of interpenetration between illegal and legal networks. Forgetting the requirement to be functionally nonjudgmental in the social sciences, the actors and organizations that cooperate in the problem space are called dark networks, in that their activities are both covert and illegal.

[FIGURE 2 OMITTED]

How do we connect these dark networks to the world economic and political system? Joseph Nye (2002) believes that international relations exist on three different though related levels. The first is the world of military power and might. Here the United States is hegemonic and spends more on its defense than the next twenty nations combined. The second is the world of economic power and influence. Here Europe, America, and the developed countries of Asia all play important roles in a multilateral system. The third level involves
  the multifarious and proliferating nongovernmental activities shaping
   our world: currency
   flows, migration, transnational corporations, NGOs, international
   agencies, cultural
   exchanges, the electronic media, the Internet, and terrorism.
   Non-state actors communicate
   and operate across this terrain virtually unconstrained by government
   interference; and the
   power of any one state, the US included, is readily frustrated and
   neutralized. (Judt 2002, 3)


The third level is the domain of a variety of covert networks--some, such as arms trafficking and drugs, are illegal and instrumental in their pursuit of wealth and power; some, such as Al Qaeda, wish to destroy both the military and economic power of the "infidel" West.

The goal of this article is a theoretical reformulation of the relationship between networks and problems. We hope to make the case that collaboration is not always for laudable purposes. We also want to introduce a strategic contingent perspective to the discussion of the interaction of legal and dark networks. We wish to see in our cases if the terrorist group Al Qaeda, heroin-trafficking networks, and various networks engaged in smuggling arms and diamonds into and out of a failed state anticipate, react to, and adapt to the actions of their foes to disrupt and destroy them. We hope that this research will help policy makers reformulate the problem of illegal activity in network terms. To do this we need to learn a good deal about dark networks. First, are there any commonalities among the structures of dark networks? Do they adopt similar structures to cope with the need for secrecy necessitated by the illegal nature of their operations? Have these dark networks invented new organizational forms? In the past, national liberation movements created new organizational archetypes. The Bolsheviks in Russia, anarchists throughout the West, the French Resistance to the Nazi occupation of France, the Man Mau in Kenya, and the African National Congress in South Africa--all were clandestine organizations that had to evolve unique organizational structures to cope with the continuing effort of various governments to liquidate them.

Second, is there a point where dark networks come together? As depicted in figure 2, are there actual linkages between and among them--weak ties, if you will, that allow them to collaborate across different fields of illegal activity? Must they have unfettered access to territory to be effective? Is it possible to not have a safe haven, like Taliban-controlled Afghanistan, and still maintain effectiveness?

Third, according to intelligence and police officials in many different nations, international criminal and terror activity seems to be organized in network form; what can we learn from studying these networks that might help legitimate states combat them? The activities of dark networks today are a major policy problem, as piracy was in the eighteenth and nineteenth centuries. Piracy ended when Western nations took concerted action against pirates and ceased using them for their own purposes, for example, sponsoring them to ravage the shipping of other nations (Keohane 2002).

In addition to these three questions, there is a major assumption underlying this research. If dark networks are problems for the organizations that try to combat them, then to the extent that this combat is successful, international coalitions to combat terrorists and criminals will perform better. However, before examining these questions and our three cases, which illustrate the structure, operation, and performance of dark networks, we wish to discuss the current state of network research.

PRESENT STATE OF NETWORK RESEARCH

The term network has been one of the most widely used notions in the social sciences for the last two decades. In a very broad definition it can be understood as a set of social entities linked directly or indirectly by various ties. Besides a general metaphorical use, at least three concepts are attached to the term network. (1) The first concept is applied as an empirical tool to describe social structure. Second, network is used as a label for a specific type of social structure. Third, it is a concept to describe and analyze forms of governance, that is, forms of coordinating social activity. It is this usage that those in public management and policy most often use when they discuss networks (Milward and Provan 2000):
  Networks are structures of interdependence involving multiple
   organizations or parts thereof,
   where one unit is not merely the formal subordinate of the other in
   some larger hierarchical
   arrangement. Networks exhibit some structural stability but extend
   beyond formal established
   linkages and policy legitimated ties. The notion of networks excludes
   mere formal
   hierarchies and perfect markets, but it includes a wide range of
   structures in between. The
   institutional glue congealing networked ties may include authority
   bonds, exchange relations
   and coalitions based on common interest, all within a single
   multiunit structure. (O'Toole
   1997, 45)


The networks in all three of these concepts, while having empirical referents, are to a certain extent the construction of the researcher.

As both a paradigm and an empirical tool, network analysis has been one of the major innovations in the social sciences in the last thirty years and has recently been applied more and more in policy analysis as well (see, e.g., Kenis 1991; Laumann and Knoke 1987; Pappi, Konig, and Knoke 1995; Schneider 1988; Schneider and Werle 1991). There are reasons for this success. First, network concepts were based on relations between actors rather than attributes like race, class, or gender. Concentrating attention on the ties between social entities rather than on the qualities possessed by them forces social scientists to think in terms of constraints and options that are inherent in the way social relations are organized (Emirbayer and Goodwin 1994, 1414; Poucke 1979, 181).

Network in the meaning of social structure is seen as an actor system with very specific features. In policy making it is regarded as an arrangement characterized by a predominance of informal communicative relations, a horizontal as opposed to a hierarchical pattern of relations, and a decentralized pattern of actors' positions (Kenis and Schneider 1991, 32). (2) These relations are based on the exchange of information, money, political support, or credible commitments for cooperative behavior among state, societal, and private actors (Schneider 2000, 259). In both the management and economic sociology literatures, networks are often seen as horizontally structured systems of production and exchange with a predominance of informal ties and loosely coupled relations in contrast to tightly coupled, rigid, hierarchical organizations (Jones, Hesterly, and Borgatti 1997; Powell and SmithDoerr 1994).

Starting with Williamson's Markets and Hierarchies (1975), in which the author elaborates Coase's (1937) basic ideas about the determinants for the organization of economic activities, a rich body of literature has developed on different forms of governance over the last two decades. For example, transaction cost, a concept that originated in economics, is now viewed as the major factor explaining what form an institution takes--markets, hierarchies, or networks. In economics, organizational sociology, and political science a common literature and nomenclature have emerged to discuss the conditions under which different institutions materialize. Central to this discussion have been questions about which factors lead to hierarchical, network, or market arrangements and the conditions under which the different forms have comparative advantages (Hollingsworth 1997; Hollingsworth and Lindberg 1985; Hollingsworth, Schmitter, and Streek 1994). A major issue in this discussion is whether networks are simply a combination of elements of market and hierarchy--and may therefore be placed on a continuum somewhere between market and hierarchy--or whether they are better understood as a unique form of governance (Powell 1990).

One of the basic features of research on networks, which repeatedly appears in almost all studies, is the statement that networks are often the only governance form that is able to deal with today's complex problems, which span the established formal (hierarchic) structure of state bureaucracies and territorial boundaries. (3) The actors in a network share common interests with regard to a policy and exchange resources and information to pursue these shared interests, acknowledging that cooperation is the best way to achieve common goals (Borzel 1998, 254).

In management and organizational theory, networks are advocated as a way to disseminate information quickly, foster innovation, make large hierarchical organizations more flexible, and enhance competitiveness. Networks are seen as an ideal arrangement to preserve organizational independence and flexibility while offering multiple organizations the possibility of reaching common goals in limited areas such as research and development or enhancing delivery of certain goods and services. Joint ventures, strategic alliances, consortia, or more loosely coupled informal systems of cooperation are seen as vital to overcome resource shortages and to combine very specific knowledge in different areas, especially in the newly developing "high-tech" industries like information technology and biotechnology (Porter Liebeskind et al. 1995; Powell 1990: Powell and Brantley 1992: Powell, Koputt, and Smith-Doerr 1996).

In public management, networks are seen as a way for the effective delivery of public goods and services, especially in fragmented systems and in systems in which public, private, and nonprofit organizations have to operate together (Agranoff 1986; Chisholm 1989; O'Toole 1997). However, the few studies that have looked at the effectiveness of networks have found mixed results and demonstrate that networks must have specific structural characteristics besides sufficient resources and a clear and transparent control system in order to be effective (Provan and Milward 1995). A major concern in this regard is how public officials can be efficient and effective managers in these settings (Kickert, Klijn, and Koppenjan 1997; Klijn, Koppenjan, and Termeer 1995; Lynn, Heinrich, and Hill 2000; Meier and O'Toole 2001).

This is a very brief summary of what we know about bright networks, that is, a legal and overt governance form that is supposed to create benefits for the participating actors and to advance the common good and does not--at least intentionally--harm people. Recently, however, especially after the terror attacks of 11 September 2001 on Washington, D.C., and New York City, the destructive capabilities of networks have come to the attention of a wider audience. Although there have been niches in criminology and political science that have used the network concept at least metaphorically for quite some time, the research communities working on networks and governance on the one hand and on corruption and organized crime on the other have hardly talked to each other. Exceptions to this general trend are a few studies published in the journal Transnational Organized Crime (discussed below) and work published in a special issue of Political Studies in 1997 on corruption (e.g., Cartier-Bresson 1997).

Ronfeldt and Arquila, two researchers at the Rand Corporation, published an edited volume called Networks and Netwars in 2001 in which different contributions explore the possibilities of using network forms or network organizations to wage war or induce harm. Although it was largely put together before the terror attacks of 11 September, it is not surprising that the book received enormous attention, especially in newspapers and magazines. After the 11 September attacks the "network community" reacted with some work on terrorist and criminal networks that was presented at the International Social Network Association's Annual Conference in February 2002, and some of this work was published in Connections. (4) We draw on this work below. There was also considerable work presented in this area at the American Political Science Annual Meeting in summer 2002 (see, e.g., Kahler 2002; Kenney 2002; Keohane 2002).

Although the "turn to the dark side" has long been overdue in the discussion on network forms of governance, it also adds more conceptual difficulties to it. Is network simply a metaphor to circumscribe more loosely coupled systems that are clearly not markets or formal hierarchies? Is it simply a method we use to look at social structures, whether they produce good or evil outcomes? Or is it a form of governance that can be used to coordinate activities either to benefit all human beings or to inflict great harm on them? What are the benefits of using the network concept in this respect, and do the special circumstances of illegal and covert activity pose additional challenges for data collection and analysis?

Before we are able to turn to these questions, we will present three cases (heroin trafficking, the Al Qaeda terrorist group, and the smuggling of "blood diamonds" in exchange for illegal weapons in the failed states of West Africa) in which there is some initial evidence that the groups concerned might be organized as networks. It is obvious that all three violate legal norms in the Western world and that they have inflicted great harm on many innocent people. Therefore, they are dark networks--both illegal and covert. Proceeding inductively, we try to determine what the similarities and differences are with respect to the organizational form and in what sense we are able to speak of networks in each of these instances.

SELECTED CASES

Drug-Trafficking Networks

Organized crime and drug-trafficking operations have been described in network analytic terms (see, e.g., Jackson, Herbrink, and Jansen 1996; Klerks 2001), or scholars have at least used the network as a metaphor to describe illegal social structures that are not simply vertically integrated formal organizations (see, e.g., Halstead 1998; McIllwain 1998; Schmid 1996). (5) Because it is very tempting to use network as a metaphor to describe these structures between hierarchies and markets (Powell 1990), it is worth taking a closer look at their structural features. As we will see in the following description, the heroin trade and the logistics required to produce the drug and bring it from Southeast Asia or South America to the main consumer countries in Europe and North America have been increasingly specialized and decentralized. The basic feature, however, is a chain of bilateral contracts; a group or a person will only know his or her immediate predecessor and successor in the supply chain. Moreover, these distribution chains lack exclusivity; individuals at one level contract with different individuals at other levels (Layne et al. 2001, appendix C). Although this is not a dense, horizontal structure, it nonetheless resembles the supply chains that are regularly termed networks in the business and management literature. Production of designer footwear is organized in a very similar way, and it was only the consumer protest and threatened boycott (because of sweatshop labor conditions along this chain) that forced companies like Nike to take responsibility for what went on in the overseas factories they contracted with.

The time of large, powerful, and vertically integrated drug cartels of the 1980s and 1990s is over (Brzezinski 2002, 46). Like John D. Rockefeller's Standard Oil Company in the early twentieth century, the cartels tried to hang onto their product through its entire chain of manufacturing, distribution, and sale to consumers. But they ultimately paid the price for their hierarchical control; a single broken link could bring down the whole organization (Brzezinski 2002, 46). Dark networks that do not respond strategically to changes in the tactics of their legal adversaries will fail.

Systems for the provision of goods and services that are decentralized and organized through a series of subcontracts can also be found in areas like health care and human services (Provan and Milward 1995; Provan, Milward, and Isett 2002). This discussion of bilateral contract versus hierarchical control of the whole production and distribution process points to a question that has not yet been fully explored in the discussion on network governance--how well do networks adapt to changes and outside pressures, and what are the characteristics and contingencies that let some networks adapt more effectively than others?

The principal difference between legal economic activity and all illegal trade and commerce is that the latter must operate in the shadow of the law. This has at least two major consequences for the organization of the activity. First, almost all exchanges, especially over borders, become dangerous and difficult. While legal businesses can use the established services of global financial institutions and international sea and air transport firms, illegal organizations have to devote a majority of their attention and resources to transportation and money laundering. On the other hand, the greater the risks, the greater the profits. (6) Brzezinski (2002) describes in a fascinating way how the price of heroin increases more than a thousand times on its journey from the producer to the consumer. A kilo of raw opium gum sells for about $135. After refinement in Southeast Asia, it sells for about $2,500 in the country of origin. The wholesale price in New York City will then be $200,000. One of the greatest markups (almost one hundred times the preceding price) happens when the product crosses the American border and arrives in the first major city. Therefore, the networks must be covert, and consolidation and vertical integration are extremely risky because one broken link can destroy a tightly coupled network as police roll up the network by moving from one link to the next:
  As a result, the narcotics industry has adapted what might be called
   the Osama bin Laden
   approach to management: base your operation in remote safe havens,
   the more war-torn and
   chaotic the better; stay small and shifty, use specialized
   subcontractors or freelancers on a
   need-to-know basis; vary your routes and routines whenever possible;
   and most important,
   always insulate yourself with plenty of expendable intermediaries in
   case someone gets
   caught and talks. (Brzezinski 2002, 26)


Examining the drug smuggling from the Caribbean, Griffith (1997) comes to a similar conclusion. Although drug traffickers use sophisticated planning, technology, and communications, they are not well organized vertically and appear to exhibit few of the characteristics of formal organizations. Griffith characterizes traffickers as adaptive individuals who pay close attention to countermeasures taken by governments to interdict and deter drug shipments. This is even more true after 11 September, when the pressures through border security have considerably increased especially in the United States. Nonetheless, supplies are at an all-time high.

Second, disputes between two parties in the chain cannot be resolved either by hierarchy or by the legal system of a state. This increases transaction costs considerably. It is not surprising, therefore, that while there is some cooperation between different groups, the majority of trafficking between the producer, transit, and consumer countries is done within ethnically homogeneous groups where kinship ties still play an important role: "In America, as in Europe, immigrants are the major importers and distributors of heroin. This is simply because the smuggling networks prefer to deal with their own and don't trust outsiders. Nigerians will sell to fellow West Africans, Chinese to immigrants from back home and so on" (Brzezinski 2002, 46).

The major trends in the late 1990s, however, included further decentralization of larger networks of drug-smuggling organizations into smaller units and the increasing presence of short networks, small self-contained groups linked only by function and need. Not only were the organizations getting smaller and more insular; they were linked only by function, information, and immediate need, with little permanent structure and durability of functional integration (Layne et al. 2001, appendix C).

With its enormous profit margins and small size, heroin is ideally suited to be shipped tens of thousands of miles and smuggled across multiple borders. Far smaller networks are required to smuggle heroin than to smuggle cocaine and marijuana, because the profit margins per kilo are much higher in the first case; that is, smugglers need to transport only a fraction of the amount of product to make the same profits, which can be done using a much smaller infrastructure. Spreading loads over many individual couriers minimizes risk. Once the drug is in the country, the most promising and profitable strategy is for couriers to hook up with established distribution centers (Brzezinski 2002, 46)--not unlike what international corporations do with legal imports.

The Al Qaeda Terrorist Network

Unlike the drug-trafficking networks, the membership of Al Qaeda seems to be quite diverse in terms of ethnic, national, and social background. (7) Operatives are from a variety of countries in Europe, the Middle East, and Asia; are born Muslims and converts; and range from highly educated people from middle- or even upper-class families to petty criminals with only limited education (see New York Times 2001 b, 2001 c). The key to understanding the attraction seems to be the sense of identity Al Qaeda gives its members, especially young Muslim men in Europe who are alienated from their home countries but at the same time not at all integrated in the life of their host countries. Their defining commonalities are their training in the camps in Afghanistan or combat in places like Bosnia or Chechnya and zealous religious beliefs (see New York Times 2001 b). Often they were educated in the West or have been living in Western societies, so they have acquired the knowledge and ability to blend in and provide Al Qaeda with necessary camouflage.

First, the most striking and surprising feature of Al Qaeda that makes it different and much more dangerous for Western societies is its seeming ability to draw on people who have blended into their respective societies and have not been in contact with the core organization for years. Second, Al Qaeda seems to be able to plan and execute complex operations over a very long time and keep up the commitment and motivation of the operatives even for suicide missions in a hostile environment. (8) The hijackers of 11 September appear to have had a high normative consensus overlaid with religious fervor.

It is not possible to gain a complete picture of the organization at this time, especially concerning its connections to other illegal and legal activities, groups, and governments. Moreover, since the successful ousting of the Taliban regime, the expulsion of large parts of Al Qaeda and its leadership from Afghanistan, and the arrest or death of major operatives, Al Qaeda has presumably changed its organizational form and strategy (see New York Times" 2002d; Washington Post 2002a). Analytically, however, the following three levels can be distinguished. (9)

Political Level

It seems evident that numerous ties exist to radical Islamic groups and Islamic welfare organizations worldwide (see Washington Post 2001 a). (10) A variety of ethnic and national groups exist whose belief presumably binds them to the cause and not necessarily to a given organization or leader (Rothenberg 2001, 37). It is also likely that this large network consists of a multitude of legal political, charitable, economic, and social organizations whose members and functionaries do not necessarily know of the terrorists' goals or even share them. (11) They are nonetheless being used as a political and social infrastructure (for the United States, see Emerson 2002, 37ff.).

Organizational Level

The evidence we have so far suggests that Al Qaeda functions as an umbrella organization that initiates attacks. It seems similar in structure to the hollow corporation model (Business Week 1986; Milward and Proven 2000, 362), where central functions are finance, planning, strategy, and marketing but where the operating units are relatively autonomous and operate under contractual relationship with the corporation. Nike is the exemplar for this type of organizational structure. Attacks are apparently organized and carried out in a remarkably decentralized fashion, rather independently, like the actions of project teams. The command and control structure of Al Qaeda itself, however, appears to resemble a more traditional hierarchy or corporate structure with a clear division of labor--a leader (bin Laden), a council of about a dozen advisers, and committees or persons responsible for "military operations, religious affairs, finances, and the production of false travel and identity documents" (New York Times 2001a, B4).

It is assumed, however, that the original tightly knit core of Al Qaeda exists based on organizational as well as family ties. The tightly knit leadership is made up primarily of Saudi and Egyptian dissidents who developed their ideas of jihad during the CIA-funded insurrection against the Soviet occupation of Afghanistan between 1979 and 1989 (tied together by common experiences, a common ideology, common goals, and now family or marital ties [Washington Post 2001b]). On the other hand, presumably hundreds if not thousands of Muslim men, scattered over the world, have a more or less loose association with the organization; some are organized in cells, and most have gone through the military training camps in Afghanistan. It seems, however, that there is only limited over the cells in the different countries and that they are organized and controlled locally. Having a secure base with training camps and logistics in Afghanistan, the group was, as a covert secret organization, in an advantageous position. The picture emerging from government documents, court transcripts, and interviews is of an underground army so scattered and self-sustaining that even the elimination of bin Laden and his closest deputies might not eradicate the threat they have created. (12)

Operational Level (of the 11 September Attack-Project Team)

In terms of organizational analysis the Al Qaeda organization demonstrates some very remarkable features. It was possible to assemble a large group of intelligent and educated men for a suicide mission and to secure their commitment over a very long time in foreign territory. Although it seems that only the pilots knew the final goal of the mission, it is still surprising that there was presumably no defection (New York Times 2001a). The operational level of the network seems very hard to detect. It probably forms for attacks and then dissolves again, unlike the terrorist groups in Europe in the last thirty years. (13)

Every secret organization has to solve a fundamental dilemma: how to stay secret and at the same time insure the necessary coordination and control of its members (see Baker and Faulkner 1993). Every link increases the risk of detection and of the destruction of the organization if a member is discovered and detained. How this dilemma was solved in the case of the 11 September attacks is quite remarkable (for the data and analysis, see Krebs 2001). The organizational structure seems to have been based on prior trusted contacts between the members (New York Times 2001a, 2001 b). However, these ties were very limited, and as a consequence a very sparse network evolved in which team members of the same flight were sometimes more than two steps away from each other. There were apparently hardly any prior contacts between team members of different flights except for the alleged pilots of the three flights, who were members of the "Hamburg Cell." (14) Given bin Laden's comment that the groups did not know each other and that those who were trained to fly did not know the others, Krebs (2001, 46) concludes that the network traded effectiveness for secrecy. But, as we mentioned above, covert networks exist for a reason, and to accomplish their task, members must coordinate their activities. The hijackers of 11 September achieved that through the "judicious use of transitory shortcuts" (Krebs 2001,46), that is, rare meetings that connected distant parts of the network. After the coordination is accomplished, the cross ties go dormant until the need for their activity arises again.

Since the victory over the Taliban in Afghanistan, Al Qaeda is reorganizing its activities to accommodate the loss of its territorial base. Clearly the U.S.-led military action and vigorous law enforcement in the United States, Europe, and Pakistan have weakened the organization. Western security officials "believe that Al Qaeda is breaking down into small widely dispersed cells of up to five people. These are believed to be connected to a wellorganized central command whose location has yet to be established and whose means of communication is unclear" (Humes-Schulz 2002, 6). It also appears that Al Qaeda, as a reaction to the immense pressure, has changed its strategy from trying to attack highly symbolic targets in Western countries to attacking soft and more remote targets, as in the attack on the French tanker or the bombing on Bali, at least for the moment.

This seems to indicate that organizational choices by covert networks are as much a reaction to attempts by governments to destroy them as they are strategic choices. In this view, organizational structure of a covert network is a matter of the second- or third-best option.

Arms-Trafficking Networks in West Africa

A truism of the network approach is that, at some level, everything is connected to everything else (Watts 2003). This is no less true of dark networks. There is increasing evidence of a close connection between Al Qaeda and the failed states of Liberia, Sierra Leone, and Burkina Faso in West Africa. The connection appears based on Al Qaeda's need to exchange cash for diamonds. This is fueled by the pressure from the United States and Western Europe to clamp down on Al Qaeda's use of legitimate banks for international monetary transactions. Diamonds provide a ready currency for Al Qaeda, and the failed states of the region have perhaps provided a safe haven for Al Qaeda operatives in the wake of 11 September in exchange for arms and money for the warlords of the region (Farah 2002).

The trade in diamonds in exchange for guns and military hardware exists at the intersection of the failed nation-states of Africa with tribal warlords who control both natural resources and sometimes the organs of state power, such as national banks, commerce, and foreign relations. Unlike gold or silver, which have fixed values, diamonds are a compact currency that can be used to back international loans, pay debts, pay bribes, and buy arms (Cockburn 2002, 20). In addition, diamonds can be carried through airport security without alerting anyone; they are easy to hide, cannot be sniffed by dogs, and do not set off alarms. Their singular virtue is that they are easily exchanged for cash (Farah 2001a).

Sierra Leone came to the world's attention in the late 1990s as a rebel group called the Revolutionary United Front (RUF) began chopping off the arms and legs of innocent men, women, and children as a way of spreading terror and maintaining their control over that country's diamond fields. Led by Foday Sankoh, a former corporal in Sierra Leone's army, the RUF was able to smuggle out tens of millions of dollars worth of stones a year through Liberia and other neighboring countries (Cockburn 2002, 28).

In Liberia a former warlord-turned-elected president, Charles Taylor, proved eager and adept at helping the RUF launder its diamonds on the world market. Since he was a warlord in the late 1980s, one of several trying to overthrow the regime of Samuel K. Doe, Taylor proved skilled at "turning a stalemated war into a lucrative business enterprise. He became the prototypical gangland impresario thriving in a lawless market" (Berkeley 2001,24).

After Taylor and other warlords overthrew the Doe regime and Taylor was elected president (he threatened to continue the war if not elected), Monrovia, the capitol, became a mecca for all types of dubious individuals seeking to extract Liberia's natural resources, sell guns and military hardware, and launder money for diamonds. Liberia under Taylor, like Afghanistan under the Taliban, was like a gang become a state--where the state itself is a gang, the law does not exist, and the Mafia controls not just turf but also the state itself, including the police, army, secret police, courts, central bank, government departments, legislature, and media (Berkeley 2001, 15).

In the period after Taylor became president, the Republic of Liberia became a nexus for many dark networks. There are linkages among various dark networks; some are more central than others, and some are only loosely linked with the others. Networks are based on trust, reciprocity, and shared experience. In the case of Taylor these links go back to the 1980s when he trained with Foday Sankoh at Libyan terrorist training camps under the tutelage of Colonel Moammar Gaddafi. During the 1970s and 1980s Libya was the terrorist training center for the world. During this period, Libya ran a nondenominational version of the Al Qaeda camps in Afghanistan, where the Irish Republican Army, Basque separatists, and many others received military and organizational training. Gaddafi was also interested in training a cadre of African revolutionaries to spread his version of revolution and cement Libya's position in the region. Relationships formed in the camps help to explain much of West Africa's misery over the past decade (Farah 2001b).

Charles Taylor, a fugitive from a bank robbery in the United States who was wanted for embezzlement in Liberia, came to Libya in the late 1980s based on a recommendation from an army officer in Burkina Faso, Blaise Compaore, who is now Burkina Faso's president. In Libya, Ibrahim Bah, a Senegalese rebel, trained Taylor and Sankoh. Bah is a key figure in terrorism circles in both West Africa and the Middle East. Bah fought with the mujabedeen against the Soviet-backed regime in Afghanistan and with Hezbollah in southern Lebanon; he returned to Libya at the end of the 1980s to train African revolutionaries (Farah 2001b).

Taylor, Sankoh, Compaore, and Bah are at the center of a network that uses war for means other than power. Their power is based on terror and the use of the state for their own purposes. The means for this are guns, diamonds, and increasingly timber, all under the mantle of whatever legitimacy a failed state can provide. (15) After Taylor and Sankoh left Libya, they began their rebellions against the governments of Liberia and Sierra Leone, respectively. Taylor and Compaore are now presidents; Sankoh is in jail, and his rebel movement, the RUF, has disarmed; but for much of the 1990s these three men, aided by Bah, used the diamonds that Sankoh's forces controlled to buy weapons and supplies for the RUF and to grow rich on the proceeds of these deals.

By the late 1990s Liberia had become the center of a great deal of diamond-related criminal activity. Taylor oversaw the trade in weapons for diamonds with the RUF and did the same for other diamond-producing countries. This trade facilitated war and provided cover for all types of organized crime (Smillie, Gberie, and Hazleton 2000, 2-3). Over the last three years, Al Qaeda bought diamonds from the Sierra Leone diamond fields, which are very close to the Liberian border, at below-market rates and resold them in Europe for millions of dollars in profit, according to U.S. and European intelligence sources. Al Qaeda's purchases of diamonds increased greatly, and the price it was willing to pay went up during summer 2001 (Farah 2001a). This put much of Al Qaeda's assets beyond the reach of Western governments. European banking officials believe that Al Qaeda was able to shift its assets from cash to gold and diamonds before 11 September (Frantz 2002). Ibrahim Bah is the RUF's principal diamond dealer. According to these same sources, Bah acted as the middleman in the trade between top RUF officers and buyers from both Al Qaeda and Hezbollah. (16) The United Nations estimates that the blood diamonds that left Liberia in 1999 were worth about $75 million on the open market. Taylor received a commission on each transaction, and Bah and other brokers shared the rest (Farah 2001a): "That Taylor has ties to Al Qaeda shouldn't be terribly surprising. Since becoming president in 1997, Taylor has run Liberia like a giant criminal enterprise, attracting South African mercenaries, Latin American drug lords, and Ukrainian mobsters to Monrovia. Middle Eastern terrorists were bound to find their way there eventually" (Lizza 2001, 21). A recent European investigation found evidence that both Liberia and Guinea harbored senior Al Qaeda operatives for at least two months after 11 September in exchange for $1 million transferred to Charles Taylor (Farah 2002).

The trade in diamonds is probably smaller now than it has been in many years. Foday Sankoh was captured in a raid in Freetown, Sierra Leone, that netted documents that clearly link Taylor to the RUF diamond sales. He is in jail, and the new leadership of the RUF has signed a peace agreement with the Sierra Leone government and claims that it has disarmed and wants to become a legitimate political party.

With the RUF in eclipse and the imposition of U.N. economic sanctions against Taylor's government, which includes a travel ban on Taylor, Bah, and many other top Liberian officials for their role in the diamonds-for-weapons trade, Taylor had to look for other sources of revenue not covered by U.N. sanctions to support his regime and to buy weapons to put down a growing insurgency. Taylor turned to the virgin timber that covers much of Liberia's forests as a source of funds. The Oriental Timber Company is one of several operations clear-cutting Liberia's forests at an unsustainable rate. It has its own private militia, which it rents out to Taylor to help fight rebel groups springing up in opposition to Taylor, some of which are simply envious of his success in appropriating the country's resources for himself. Weapons are brought in by ships, which are off-loaded under heavy guard and then loaded with lumber. The Oriental Timber Company roads through the forest provide the means to move troops to wherever the front is and to move weapons to neighboring countries that wish to buy them. (17)

Weapons that continue to fuel the wars of West Africa have to come from somewhere, and nowhere is there a bigger armory than in the republics of the former Soviet Union. In addition to weapons, the means of transport are there too in the aging transport aircraft of the former Soviet Air Force. When governments and U.N. officials discuss arms dealers, the name that comes up time and time again is Victor Bout. Bout is a former Soviet pilot who has become the leading entrepreneur of the post-Soviet arms trade. Based out of the emirate of Sharjah, he flies a fleet of sixty battered Antonovs and Ilyushins to conflict zones throughout Africa and Central Asia. Bout flew weapons, poison, chemicals, operatives, and cash to both the Taliban and Al Qaeda in Afghanistan in addition to his African ventures. In U.N. reports Bout has been accused of violating weapons embargoes in Angola, Sierra Leone, and Liberia. His partner in these ventures in the emirates is Sheik Abdullah bin Zayed, the former emirate ambassador to the United States (Pasternak and Braun 2002). Most of the aircraft used by these two men were registered in Liberia.

Recently, a Kenyan named Sanjivan Ruprah was arrested in Belgium. Ruprah was a former air transport firm and diamond mine owner who was authorized to represent the Liberian government to expand its airline registry. Just as the Liberian ship registry provides a flag of convenience for companies wishing to avoid taxes, scrutiny, and regulation, so does the Liberian aircraft registry provide the same service for air transport companies. These registries are another source of income for Taylor.

The reason that Ruprah, and until recently Bout (who moved from Belgium to Sharjah when he became known to the world as an arms dealer), required access to Belgium was for proximity to Antwerp. The arms dealers were paid in diamonds for the weapons they delivered, and Antwerp is one of the diamond centers of the world where blood diamonds can be laundered (Braun, Pasternak, and Rotella 2002).

Bout sells weapons that are hard to hide, like helicopter gunships and heavy artillery, as well as all manner of small arms and explosives, but he creates layers of companies and many links between himself and his arms trafficking. Brokers and middlemen can work out of a hotel room in Brussels using the telephone and a fax machine to arrange an arms deal between Liberia and the Ukraine. (18) The arms dealer is breaking no laws because no arms ever go through Belgium (Frontline 2002).

A chance cocaine bust in Milan of a hotel room filled with prostitutes led to the arrest of a notorious arms dealer named Leonid Minin. In addition to the cocaine and prostitutes, the police found "more than $35,000 in cash, a half million dollars in diamonds, and more than 1,500 documents detailing a tangled web of business dealings in oil, diamonds, timber and gun shipments to Africa" (Frontline 2002). To underscore the global nature of the arms-for-commodities networks, Minin, an Israeli citizen with ties to the Odessa Mafia, is being prosecuted in Italy for selling Ukrainian arms to Liberia and Sierra Leone through a broker in Russia and banks in Turkish Cyprus, Switzerland, and the United States (Frontline 2002). (19)

What conclusions can we draw from an examination of these global networks that center on one failed state? One conclusion is that the nature of war has changed:
  The point of war may not be actually to win it, but to engage in
   profitable crime under the
   cover of warfare.... Over the years, the informal diamond mining
   sector, long dominated
   by what might be called "disorganized crime," became increasingly
   influenced by organized
   crime and by the transcontinental smuggling not just of diamonds,
   but of guns and drugs,
   and by vast sums of money in search of a laundry. Violence became
   central to the advancement
   of those with vested interests. (Smillie, Gberie, and Hazleton 2000)


Covert networks have come together with warlords controlling access to resources to create commodity wars. These wars are fought over control of diamonds, petroleum concessions, and coca leaves and poppies, which yield narcotics, not for any real ideological or political reason. War is the only business open at the entry level to large numbers of displaced or simply greedy people in weak or failed states (Hoagland 2002).

How can the United Nations and strong states control these covert networks that are laying waste to so much of the Third World? These networks flourish when strong states do not really care or have greater issues to deal with. In the case of Liberia, U.N. sanctions played a positive role. Nongovernmental organizations like Global Witness alerted the world to what the arms-for-diamonds trade was doing to West Africa. They touched the world's conscience by connecting children with severed limbs with diamonds. They coined the phrase blood diamonds. When the United Nations imposed sanctions and the threat of a boycott of diamonds from the region, the conflict in Sierra Leone began to wind down.

Another lesson is that failed states can be extremely useful to covert networks. Not only are there commodities to extract for arms or supplies, but members of these networks can be provided with diplomatic passports, and in this case the Liberian ship and air registries give covert networks the ability to give their transport arms some degree of legitimacy. (20) Perhaps the United Nations and the world community need some means of declaring states like Liberia and Sierra Leone bankrupt and taking away the privileges of statehood by turning them into U.N. protectorates like Bosnia?

LIGHT ON THE SIDE: QUESTIONS, CHARACTERISTICS, AND PROPOSITIONS ON COVERT AND ILLEGAL NETWORKS

This section consists of a comparison of the cases presented above and attempts to answer the four central questions raised at the beginning of this article:

1. Are there any commonalties among the structures of dark networks? Do they adopt similar structures to cope with the need for secrecy necessitated by the illegal nature of their operations? Have these dark networks invented new organizational forms?

2. Is there a point where dark networks come together? As depicted in figure 2, are there actual linkages between and among them--weak ties, if you will--that allow them to collaborate across different fields of illegal activity? Do they require unfettered access to territory to be effective? Is it possible to maintain effectiveness without a safe haven like Taliban-controlled Afghanistan?

3. What are the differences and similarities between dark (illegal and covert) and bright (legal and overt) networks?

4. According to intelligence and police officials in many nations, international criminal and terror activity seems to be organized in network form; what can we learn from studying these networks that might help legitimate states combat them? (21)

We realize that our cases were selected in an opportunistic and not systematic fashion. The cases vary on some dimensions and not on others. We also acknowledge that our information, which comes from many different public sources, is incomplete and may suffer from intentional or unintentional distortion. These are problems that apply to any illegal and covert network. Given the importance of these networks, we believe that it is legitimate in an exploratory study of this kind to take three of the most prominent cases, analyze them, and inductively develop propositions that can be used to guide future research on dark networks. Although we have used the phrase dark networks ourselves, we have to admit that the term is inherently problematic. It contains an evaluation of goals that is by its nature normative (see also Erickson 1981, 209); terrorists to some people are freedom fighters to others. A neutral concept that can be operationalized according to social science standards is the dichotomy between overt versus covert networks. The first are public and out in the open; they are legal, and people participating in their activities are not facing risks such as imprisonment, injury, or death. Covert networks, on the other hand, are mostly illegal; that is, their activity is contrary to the law that is enacted in the geographic area where the activity takes place. The members of covert networks have to hide their activities and face constant risks if discovered. They face a constant dilemma: The more active they are, the more likely it is that they will be detected because the more activity they generate, the more visible they become to government authorities.

Question 1. Are There Any Commonalities among the Structures of Dark Networks? Do They Adopt Similar Structures to Cope with the Need for Secrecy Necessitated by the Illegal Nature of Their Operations? Have These Dark Networks invented New Organizational Forms?

As we have seen from the description of the case studies, covert networks, like overt networks, come in all forms and shapes, from complex overlapping and interlinked structures of the diamond and arms trade to the chain of bilateral contract relations of the heroin trade that facilitates the distribution of the drug downstream. Moreover, trying to determine whether all covert networks are structured in similar fashion is as fruitless as trying to find one pattern among overt networks. Given our strategic and contingent perspective, the answer is that it very much depends on the specific opponents and competitors. Some networks are dense at the top and loose at the bottom, like Al Qaeda before 11 September. Others, like the drug-smuggling networks, can go from a vertically integrated network to a series of bilateral contracts based on the threats to their existence. Networks can be constructed on a need-to-know basis where the costs of discovery are high or on larger and more durable structures where risk of detection is low. The common feature of dark networks that survive is their ability to stay flexible and adapt quickly to changing pressures and circumstances.

Actors in covert networks must constantly evaluate the risk and react accordingly. The higher the risk, the more resources and attention will go into measures to avoid detection. Here, trust relations are of crucial importance. As we have seen in the examples, every illegal activity that needs continuing coordination is based on ties of trust that were often formed long before the illegal activity started. This occurs because recruitment of new members follows the path of established trust relations. The relations between major players in arms and diamonds smuggling were built in the training camps in Libya in the 1970s and 1980s. In a similar fashion, the members of Al Qaeda underwent a comparable personal development during the fight against the Soviets in Afghanistan or their training in the camps of Osama bin Laden, where they formed a similar worldview, built common interests, and very often established personal relationships. These trust relations are of vital importance not only to minimize the risk of treason and detection but also to minimize or resolve disputes. Because covert networks cannot rely on formal institutions or the legal systems of nation-states, disputes have to be resolved in other ways or the actors risk mutual destruction if they revert to violence, leading to Mafia-like vendettas. Often relationships are secured with virtual or literal mutual hostage taking, which is a very old technique to secure an agreement. A breach of trust is connected to mutual destruction or at least the death of loved ones. Bin Laden seems to have married several of his children to close allies. Because risk is such a big factor, professional and personal lives are intermeshed and almost indistinguishable.

We do not think that there is a major new form of organization visible in the cases that have been presented. On the contrary, it is rather interesting to see that in many instances they resemble successful organizational models that we know from legal activity. The organizational structure of heroin production and distribution, for example, looks very much like the production and distribution chain of multinational companies that produce consumer goods. Al Qaeda before 11 September seems to have had a functionally differentiated corporate core that financed and supported terrorist cells that came up with ideas and plans for actions very much like we would see in a hollow corporation.

Proposition 1 Covert networks like overt networks come in all forms and shapes. Dark networks must have a very flexible structure that enables them to react quickly to changing pressures from nation-states and other opponents in order to survive.

Question 2. Is There a Point Where Dark Networks Come Together? Do They Require Unfettered Access to Territory to Be Effective? Is It Possible to Maintain Effectiveness without a Safe Haven like Liberia or Afghanistan?

We believe that we have shown that dark networks from different areas are connected through actors who function as brokers between these different networks. Very often these brokers are dictators, rebel chiefs, warlords, or presidents of failed states with private armies. They facilitate the operation of all types of covert networks and in that process achieve enormous material gains. In our West African case, we have documented the connections among failed states (with warlords and rebel armies), terrorism, diamond smuggling, and arms trafficking. Covert networks of all kinds do not operate in a vacuum, despite the fact that they now frequently use modern communication technology for their activities. They need a territorial base of some sort either for the production or brokering of their illegal goods or as refuge for planning, training, and recovering. If Al Qaeda does not control a territorial base, then, even in a failed state, terrorist training camps are impossible. Likewise, if Al Qaeda decides to refrain from all but the most critical electronic communication and uses trusted messengers for face-to-face communication instead, then communications interceptions become very difficult for Western intelligence agencies. At the same time, while Al Qaeda has decreased the probability of detection, it has increased the difficulty of communication among its members all over the world to mount complex operations like the hijackings of 11 September. The strategy may change from hitting high-profile targets in world centers, like New York, London, Paris, and Berlin, to a second- or third-best strategy of bombing soft targets in remote locations, like a nightclub in Bali or a synagogue in Tunisia.

West Africa, and especially the subregion of Liberia, Sierra Leone, and Burkina Faso, seems to be a major geographical center where many of the dark networks we have described come together. This is because this is a region where no nation-state with a legitimate monopoly of coercive power exists and where diamonds, a highly valuable and exchangeable currency for dark networks, are mined and sold. In the case of Al Qaeda and the illegal weapons and diamonds trade, President Charles Taylor of Liberia and President Blaise Compaore of Burkina Faso are presumably the brokers.

Covert networks have to face a clear dilemma between security and effectiveness also with regard to potential territorial bases. If they withdraw to their comparably safe territorial bases, then they lose effectiveness because these bases are usually in rather remote and underdeveloped regions with little public infrastructure. Therefore, covert networks will try to establish a foothold in places where communication, travel, and arranging deals are far easier than in failed states. At the same time, in a strategic contingent sense, this is much more difficult when the Western states are "at war" with covert networks.

Proposition 2 Dark networks need a territorial base to be effective. These bases are usually in regions tom apart by civil war and dominated by current or former warlords and where no state with a legitimate monopoly of coercive power exists. At the moment, several countries in West Africa constitute the geographical center where a number of dark networks come together to exchange resources that each finds valuable for its own illegal purposes.

Question 3. What Are the Differences and Similarities between Dark and Bright Networks?

The central question for all networks is, what holds a network made up of different, more or less autonomous, but interdependent actors together? What are the factors that provide the glue that holds networks together? What common elements allow them to function effectively? Both overt and covert networks are differentiated by function or by geography. They sometimes consist of hundreds of people or organizations and have to find a way to coordinate and control the activities of their nodes. Here, exchange of information is vital for both overt and covert networks. However, while persuasion, exchange, and negotiation are the central mechanisms for management and conflict resolution in overt networks, coercion and physical force are the distinctive characteristic of covert networks. Because the activities of actors in covert networks are illegal, they cannot revert to the external legal system as a last resort to resolve conflicts. As a consequence, transaction costs in covert networks are higher than those in overt networks. The dominant orientation of actors in overt networks is to manage a common problem or achieve a common goal, whereas in covert networks actors try to fulfill complementary interests and avoid detection. This can especially be seen in the cases of arms and diamonds smuggling and heroin trafficking, wherein either personal material interest or political power in its most naked form is the dominant motivation for the actors.

This also has consequences for the dominant mode of integration. Overt networks are integrated through the orientation of the actors toward a common issue, problem, or goal and through the exchange or pooling of resources and information. In addition, institutionalized linkages that are based on formal agreements or even laws play a vital role. Of course, this mechanism of integration is not available for covert networks. Here integration is primarily based on trust relations between individual persons and their complementary interests. In the case of networks like Al Qaeda with overarching political goals (e.g., the removal of Crusaders and Jews from the Middle East), shared beliefs and ideology play a major role in integrating sometimes very diverse people. With overt and covert networks, as in all organized activity, a reasonable trade-off between differentiation and integration must be achieved if goal achievement is even a possibility.

If the basic prerequisites of overt and covert networks are the same for them to function effectively, on what dimensions are they different? Erickson (1981) argues in this respect that the major difference is the personal risk that members (and their families) take with their activities. In his view, the major sources of the structure of what he calls secret societies are found in preexisting social structures rather than in psychological factors. Risk enforces recruitment along lines of trust and, thus, through preexisting networks of relationships, which set the limits of the secret society's structure (Erickson 1981, 188). This also limits the talent available to the covert network; thus the social structures of overt and covert networks tend to move in the opposite direction. Whereas in the case of overt networks the structure will tend toward density, the structure of covert networks will tend to be as sparse as possible to achieve the goals of the participating actors.

Proposition 3 Overt and covert networks face similar challenges in terms of differentiation and integration. However, covert networks have to find different answers to these challenges because on the one hand their activity is under constant risk of detection and institutionalization is rather weak but on the other hand they tend to revert to physical force and coercion to compel adherence to their goals.

Question 4. According to intelligence and Police Officials in Many Nations, International Criminal and Terror Activity Seems to Be Organized in Network Form; What Can We Learn from Studying These Networks That Might Help Legitimate States Combat Them?

The more covert networks have to invest in staying secret, the less attention and resources will be available for their illegal activities. The more actors must invest in creating and maintaining trust relationships and staying undercover, the less capacity they have to act toward their nefarious goals. Therefore, keeping up the pressure in as many countries as possible through concerted international collaboration is vital. When put under intense pressure from legitimate states, dark networks have to constantly worry about the security of their organizational structure, modes of communication, and so forth, so that fewer resources and less attention remain for their specific activity.

There is a clear analogy between terrorism and drugs. In both cases, countries like the United States have declared war on them. Heroin traffickers today tend to operate more like highly compartmentalized terrorist cells than the sprawling Colombian cartels of the Pablo Escobar era. When there is a war on, it is very hard to reap the advantages of scope and scale because of the risks of discovery: "This type of activity does not allow concentration of power like legitimate business.... If smugglers get too big, they develop security and personnel problems and get targeted by law enforcement" (Brzezinski 2002, 26). This can be generalized to almost all current illegal activities that have roused the ire of powerful countries. "Flying under the radar" to try to avoid detection may be the only viable strategy for covert networks that have too much visibility for their own good. Western nations only targeted pirates when these states' losses increased and they could no longer use the pirates for their own purposes, for example, attacking the shipping of other Western countries. There is a historical analogy between England using pirates as surrogates to attack French and Spanish shipping and the CIA's use of mujahedeen as proxies against the Soviet Union in Afghanistan. Carrying this analogy a bit further, if a covert network becomes too successful, as Al Qaeda clearly did after 11 September, it provokes a furious response that increases the difficulty of achieving other goals like forcing U.S. troops from Saudi Arabia, where the Islamic holy places of Mecca and Medina are located.

Moreover, this should also be a warning to nation-states that try to use such groups for short-term political gains. In today's globalized world it is almost impossible for any nation, even for a hegemonic power like the United States, to contain the threat it once helped to create. Therefore, helping countries to build stable and reasonably representative state structures with a legitimate monopoly of power should be--in the long run--one of the major strategies of the West to fight criminal organizations and networks.

Proposition 4 The strategy of Western countries to fight criminal organizations and networks should be twofold. First, the West should increase the pressure on illegal networks in as many countries as possible through internationally coordinated action and should sanction states that voluntarily provide a base for these activities. Second, Western nations should help countries avoid becoming failed states by encouraging and assisting them to build stable and reasonably representative state structures with a legitimate monopoly of power in order to deter illegal, covert networks from using them as bases.

CONCLUSION AND PERSPECTIVES FOR FURTHER RESEARCH

This article has attempted to move our discussion of networks toward a strategic and contingent perspective. The "network as problem" perspective we have introduced helps us understand the relationship between oppositional networks. It is as if we are back in the "spy versus spy" days of the Cold War when the CIA mirrored every move by the KGB and vice versa. The actions of one are directly related to the actions of the other. In a world where one side acts to close off options of the other, second- or third-best options may be all that are possible. This is especially true for covert networks that have aroused the ire of the world community or large parts of it.

Given that this is exploratory research, we freely admit that many of our conclusions are speculation. In this article we have tried to bring together as much information as possible on three different covert networks to discover their similarities and differences. At the same time, we have tried to frame this research in the broader context of the ongoing work on networks that is proceeding across a number of disciplines. We have discussed the policy implications of research on covert networks and where legal and overt networks involving international collaboration might have the best chance of countering them.

In our three examples of dark networks (heroin trafficking, Al Qaeda, and arms and diamond smuggling) we have shown that while the networks can take on unique structural features, they also share important characteristics among themselves and with overt and legal networks. Moreover, it has been demonstrated that organizational characteristics are used that have already been successful in the overt world. The defining difference between overt and covert networks is the need for secrecy and the use of physical force. In addition, secrecy and information-processing capacity are two competing and contradictory needs in illegal networks. Concealment usually is regarded more highly, and therefore illegal networks have a tendency toward a decentralized structure unless the task necessitates a centralized one (Baker and Faulkner 1993): If, for example, Al Qaeda were on the verge of taking power in Saudi Arabia, it might reorganize along more conventional lines, like the Vietcong and North Vietnamese fighters did as they began their push south after the American withdrawal.

Although using the network concept to analyze the social organization of illegal activity adds more conceptual difficulties to the approach, we believe that the benefits outweigh the problems. The network concept is useful as an analytical tool because it calls for a systematic collection of information about the relations among the social units, be they individuals, groups, parts of organizations, or whole organizations, it is further useful as a description for the social organization because, as we showed in the cases presented, there are many instances in both the covert networks and the overt networks combating them in which neither a clear hierarchical order nor a market structure of autonomous actors exists. In our view, elements of network as a form of governance can be found in overt and covert networks alike. The goals that are necessarily normatively set by the participating actors or by outside observers do not have a major influence on the way the activity is organized. Organizational structure is, rather, a function of the need for security. In this respect, networks can be used to coordinate activities to benefit all people and to inflict great harm, which is also true for formal hierarchical organizations. The circumstances of illegal and covert activity clearly pose additional challenges for data collection and analysis. Although it is often very difficult for scholars to get sufficient relational data in the case of overt networks, it might, in many instances, be impossible in the ease of covert networks. Therefore, a viable research strategy for the future might be to look at closed cases for which ample material is publicly accessible in the form of news reports, court transcripts, interrogation protocols, and so on.

While we have at least partly answered some of the important questions, others were barely touched, for example: How shall we classify the many different ways that an organization can be something other than hierarchical in form? Empirically, how much variation does there tend to be in covert networks, and how can this variation best be conceptualized? Where are the fault lines of dark networks? What is their point of greatest vulnerability? While these are research questions, what about the policy implications of this research?

One of the major policy implications is, of course, how best to combat them. Very often it is assumed that the most central actors are the most powerful and that if they are removed, then the network will collapse. While there is a strong logic to this assumption, criminal gangs frequently remain active even after their leaders are behind bars: "This suggests that the most important actors in any group may not necessarily be ... the most central" (Jackson, Herbrink, and Jansen 1996, 86). It may also mean that in all organized activities there is a contest for authority and that with leaders removed, others rebuild the structure under different leadership (for experimental results, see Carley, Lee, and Krackhardt 2001). If covert networks, like overt networks, are also characterized by a need to integrate functionally or geographically differentiated elements, then the most likely way to disrupt them is to attempt to find the link that joins as many of these differentiated elements as possible. Geographically, West Africa seems to be the point where currently many dark networks come together. From a policy standpoint, this argues for the Western states to pay more attention to the situation in West Africa, especially in the fight against blood diamonds and illegal arms trafficking as well as international terrorism.

Whatever the locale, dark networks will remain a problem as long as the basic causes that brought them into existence and sustain them are not eradicated: drug trafficking will persist as long as there is a high demand for drugs in the Western countries with huge profit margins. Terrorism will remain an essential problem as long as people feel suppressed, alienated, and impoverished and are susceptible to violent solutions. Diamond and arms smuggling will continue as long as there are (civil) wars. The activity will move elsewhere once one failed state has ceased to be a convenient operative base, although it is possible that it will be difficult to find another and the new base may be less advantageous than the last one. Therefore, until we are able to eradicate these long-term causes, containment of dark networks is the best choice we may have.

A first version of this article was presented at the International Conference on the Empirical Study of Governance, Management, and Performance, Barcelona, 4-5 October 2002. We would like to thank the participants of the conference, especially Larry O'Toole, for their helpful comments on the revision of this article.

(1) We owe this idea to Patrick Kenis (1991). The basic idea can also be found in Pappi 1993.

(2) Kenis and Schneider actually talk about "the predominance of informal, decentralized and horizontal relations" (1991, 32). Using the analytical distinction between network as social structure and network as governance from, one notices that they are actually talking of structure and that it should be formulated as above because it is hard to imagine a "decentralized relation."

(3) It is interesting that the administrative response taken by the Bush administration to "homeland security" after the terror attacks in the United States is more hierarchy than network. Rather than emphasizing horizontal coordination, numerous existing departments and whole agencies (among them the Coast Guard, the Immigration and Naturalization Service, and the Federal Emergency Management Agency) were reorganized in a new cabinet-level department (Department of Homeland Security Reorganization Plan, www.whitehouse.gov/news/releases/2002/11/ reorganization_plan.pdf, 25 November 2002).

(4) Connections is to members of the International Social Network Association what Political Science is to members of the American Political Science Association; it is a journal that chronicles the profession rather than a journal that publishes only refereed articles.

(5) The information about the following case has been taken from an article by Matthew Brzezinski (2002). In the article he gives a unique and very interesting account of the ways heroin travels from the production sites in Southeast Asia or South America to the United States. In a personal journey, he investigated the heroin channels downstream from the source to the consumer at a street corner in Philadelphia, and he analyzes the changes that have occurred in the heroin trade in the last decade with a management framework. For a broad and detailed overview of the organizational characteristics of drug smuggling, see Layne et al. 2001.

(6) Reviewing prior research, it is concluded in the report by Layne et al. (2001, appendix C) that merely seizing more drugs in the absence of arrest was unlikely to yield results in deterrence, as smugglers regarded such losses as part of the cost of doing business, whereas a high probability of personal apprehension had a deterrent effect. This is only true, however, as long as dealers can roll over these costs onto the consumer.

(7) Although there is evidence that in the past at least some of the clandestine organizations that tried to achieve political goals through violent actions were organized in network-like structures (see, e.g., the FLN liberation movement in Algeria, Irish Republican Army, Basque separatists), because of the terrorist attacks of 11 September we concentrate exclusively on Al Qaeda. This seems justified because some initial evidence suggests that at least some aspects of Al Qaeda are particularly "networky" and differ significantly from prior terrorist organizations. Although it is possible that some of the information is actually misinformation planted by Western intelligence services and it is clear that the current publicly available information is only a fraction of the whole story, it is quite surprising how large the body of confirmed and consolidated public information already is.

(8) In a report for PBS's Frontline, journalist Hedrick Smith asserted that the biggest mistake of Western intelligence agencies was a failure of imagination that such an attack could be launched on U.S. soil. He quotes intelligence people: "We had no idea that you could motivate and maintain the motivation over a long period of time and over a long distance" (www.pbs.org/wgbh/pages/frontline/shows/network).

(9) Note that "the investigation in this sense has not turned up evidence that the same groups were responsible for all those plots but rather that there is a kind of interlocking terrorist directorate, with one group taking the baton from another, and one group's goals becoming those of the next group. The form of terrorism that struck on Sept. 11 involves a still shadowy and fluid network of people and groups, and it clearly shows that since the mid-1990's, many parts of that network have gravitated toward Al Qaeda. Sept. 11, with its three separate groups of young men from scattered places coming together to shatter America's calm, was the culmination of that process" (New York Times 2002c).

(10) The attacks in Kenya, Tanzania, Yemen, the United States, Tunisia, Bali, and now presumably again Kenya clearly show that Al Qaeda can operate effectively on a global basis.

(11) It was recently reported that despite joint international efforts it is still very hard to seriously disrupt this support structure (New York Times 2002a).

(12) Emerson (2002, 60), however, states that the motives of the terrorists are not simply religious fanaticism. There seem to be mixed motives as well. Some seem to be participating for the mere sake of the intrigue, others for money; others are unbalanced and susceptible to charismatic leadership, and still others are genuine fanatics.

(13) Even after the video that was broadcast by the Qatar-based TV station Al Jazzeera, in which the planning of the 11 September attacks is described, it is not entirely clear how much the hijackers knew in advance, although there is evidence that they knew they were on a "martyrdom mission" (Makiya and Mneimneh 2002), Recent evidence gathered by German criminal investigators confirms that several key figures in the attacks were in Afghanistan at the same time in 1999-2000 and that the final decision to launch the attack was made there, although it is not known yet whether they were together. Further coordination presumably took place in a meeting held in Kuala Lumpur in January 2000 that connected the "Hamburg Cell" with other hijackers (see New York Times 2002b).

(14) New evidence suggests that alleged ringleader Mohammed Atta and other members of the cell were recruited by a middleman, Mohammed Haydar Zammar, a Syrian-born German citizen who served as a liaison to the senior Al Qaeda leadership and usually made the travel arrangements for recruits from Germany to Afghanistan (see The Observer 2002: Washington Post 2002b).

(15) After over two decades of fighting, Liberia as a state is little more than a forested version of Afghanistan. There are virtually no public services or functioning institutions. The state exists to enrich Charles Taylor.

(16) Bah is believed to have introduced Abdullah Ahmed Abdullah to Taylor. Abdullah is on the FBI's most-wanted list of Al Qaeda terrorists as a top bin Laden adviser who helped plan a number of Al Qaeda attacks.

(17) In 2001, 52 percent of Liberia's timber production was managed by Oriental Timber Company, which exported 61 percent of the timber products of the country (Goujon 2002).

(18) Bout is believed to employ at least three hundred people in his operations, including pilots and brokers.

(19) Italy's Court of Cassation ordered the release of Minin because it believed that Italy did not have jurisdiction to prosecute him for arms trafficking in other countries (BBC Monitoring 2002).

(20) The ship registry has 1,700 ships that fly the Liberian flag of convenience. This contributes $18 million per year to the government of Charles Taylor (Arieff 2002).

(21) Besides Al Qaeda, see, for example, recent findings about the Basque terrorist group ETA (Washington Post 2003).

REFERENCES

Agranoff, Robert. 1986. Intergovernmental management: Human service problem solving in six metropolitan areas. Albany: State University of New York Press.

Arieff, Irwin. 2002. UN pushes Liberia to audit shipping, timber funds. Reuters English News Service. 4 September.

Baker, Wayne E., and Robert R. Faulkner. 1993. The social organization of conspiracy: Illegal networks in the heavy electrical equipment industry. American Sociological Review 58:837-60.

BBC Monitoring. 2002. Italian Court of Cassation orders arms trafficker's release from jail. British Broadcasting Corporation, 19 September.

Berkeley, Bill. 2001. The graves are not yet full: Race, tribe, and power in the heart of Africa. New York: Basic Books.

Borzel, Tanja A. 1998. Organizing Babylon--On the different conceptions of policy networks. Public Administration 76:253-74.

Braun, Stephen, Judy Pasternak, and Sebastian Rotella. 2002. Al Qaeda linked to Russian arms broker. Los Angeles Times, 16 February.

Bryson, John M., and Barbara C. Crosby. 1992. Leadership for the common good. San Francisco: Jossey-Bass.

Brzezinski, Matthew. 2002. Re-engineering the drug business. New York Times Magazine, 23 June:24-29, 46, 54-55. Business Week. 1986. Special report: The hollow corporation. Business Week, March:56-85.

Carley, Kathleen M., Ju-Sung Lee, and David Krackhardt. 2001. Destabilizing networks. Connections 24, no. 3:79-92.

Cartier-Bresson, Jean. 1997. Corruption networks, transaction security and illegal social exchange. Political Studies 45:463-76.

Chisholm, Donald. 1989. Coordination without hierarchy. Berkeley: University of California Press. Coase, Ronald H. 1937. The nature of the firm. Economica 4:386-405.

Cockburn Andrew. 2002. Diamonds: The real story. National Geographic, March:6-34.

Emerson, Steven. 2002. American jihad: The terrorists living among us. New York: Free Press.

Emirbayer, Mustafa, and Jeff Goodwin. 1994. Network analysis, culture, and the problem of agency. American Journal of Sociology 99:1411-54.

Erickson, Bonnie H. 1981. Secret societies and social structure. Social

Forces 60:188-210.

Farah, Douglas. 2001 a. Al Qaeda cash tied to diamond trade. Washington Post, 2 November:A1.--. 2001b. An "axis" connected to Gaddafi. Washington Post, 2 November:A22.--. 2002. Report says African's harbored Al Qaeda: Terror assets hidden in gem-buying spree. Washington Post, 29 December:A1.

Frantz, Douglas. 2002. Front companies said to keep financing terrorists. New York Times, 19 September:A17.

Frontline. 2002. Gunrunners. Available at www.pbs.org/frontlineworld/ stories/sierraleone/bout.html.

Goujon, Emmanuel. 2002. Sanction-hit Liberia uses wood to stoke the flames of war. Agence France-Presse, 10 August.

Griffith, Ivelaw Lloyd. 1997. Drugs and security in the Caribbean: Sovereignty under siege. College Park: Pennsylvania State University Press.

Halstead, Boronia. 1998. The use of models in the analysis of organized crime and development of policy. Transnational Organized Crime 4:1-24.

Hjern, Benny. 1992. Illegitimate democracy: A case for multiorganizational policy analysis. Policy Currents 2, no. 1:1-5.

Hoagland, Jim. 2002. Diamond-backed warriors. Washington Post, 28 February:A23.

Hollingsworth, Joseph R. 1997. Coordination of economic actors and social systems of production. In Contemporary capitalism--The embeddedness of institutions, edited by Joseph R. Hollingsworth and R. Boyer, 1-55. Cambridge: Cambridge University Press.

Hollingsworth, Joseph R., and Leon N. Lindberg. 1985. The governance of the American economy: The role of markets, clans, hierarchies, and associative behaviour. In Private interest government: Beyond market and state, edited by Wolfgang Streeck and Philippe C. Schmitter, 221-54. London: Sage.

Hollingsworth, Joseph R., Philippe C. Schmitter, and Wolfgang Streeck. 1994. Governing capitalist economies: Performance and control of economic sectors. Oxford: Oxford University Press.

Humes-Schulz, Stacy. 2002. Dispersal of al-Qaeda dictates shift in anti-terror strategy. Financial Times, 11 September:6.

Jackson, Janet L., Janet C. M. Herbrink, and Robert W. J. Jansen. 1996. Examining criminal organizations: Possible methodologies. Transnational Organized Crime 2:83-105.

Jones, Candace, William S. Hesterly, and Stephen P. Borgatti. 1997. A general theory of network governance: Exchange conditions and social mechanisms. Academy of Management Review 22, no. 4:911-45.

Judt, Tony. 2002. Its own worst enemy. New York Review of Books, 15 August: 1-9. Available at www.nybooks.com/articles/15632.

Kahler, Miles. 2002. Networks and failed states: September 11 and the long twentieth century. Paper presented at the Annual Meeting of the American Political Science Association, Boston, 29 August-1 September.

Kenis, Patrick. 1991. The preconditions for policy networks: Some findings from a three-country study on industrial restructuring. In Policy networks: Empirical evidence and theoretical considerations, edited by B. Marin and R. Mayntz, 297-330. Boulder/Frankfurt: Westview Press/Campus Verlag.

Kenis, Patrick, and Volker Schneider. 1991. Policy networks and policy analysis: Scrutinizing a new analytical toolbox. In Policy networks: Empirical evidence and theoretical considerations, edited by B. Marin and R. Mayntz, 25-62. Boulder/Frankfurt: Westview Press/ Campus Verlag.

Kenney, Michael. 2002. The challenge of eradicating transnational criminal networks: Lessons from the war on drugs. Paper presented at the Annual Meeting of the American Political Science Association, Boston, 29 August-I September.

Keohane, Robert O. 2002. The globalization of informal violence, theories of world politics, and "the liberalism of fear." Paper presented at the Annual Meeting of the American Political Science Association, Boston, 29 August-1 September.

Kickert, Walter J. M., Erik-Hans Klijn, and Joop F. M. Koppenjan, eds. 1997. Managing complex networks: Strategies for the public sector. London: Sage.

Klerks, Peter. 2001. The network paradigm applied to criminal organisations: Theoretical nitpicking or a relevant doctrine for investigators? Recent developments in the Netherlands. Connections 24:53-65.

Klijn, Erik-Hans, Joop F. M. Koppenjan, and Katrien Termeer. 1995. Managing networks in the public sector: A theoretical study of management strategies in policy networks. Public Administration 73:437-54.

Krebs, Valdis. 2001. Mapping networks of terrorist cells. Connections 24:36-42.

Laumann, Edward O., and David Knoke. 1987. The organizational state: Social choice in national policy domains. Madison: University of Wisconsin Press.

Layne, Mary, et al. 2001. Measuring the deterrent effect of enforcement operations on drug smuggling, 1991-1999. Report for the Office of National Drug Control Policy Office of Programs, Budget. Research and Evaluations. Available at Abt Associates, Inc., www. whitehousedrugpolicy.gov/ publications/enforce/measure_effect_2001.

Lizza, Ryan. 2001. Can Charles Taylor's apologists explain his ties to Al Qaeda? New Republic, 19 November:21

Lynn, Laurence E., Caroline Heinrich, and Caroline Hill. 2000. Studying governance and public management: Challenges and prospects. Journal of Public Administration Research and Theory, 10. no. 2:233-61.

Makiya, Kanan, and Hassan Mneimneh. 2002. Manual for a "raid." New York Review of Books 49, no. 1 (17 January):18-21.

Mcillwain, Jeffrey Scott. 1998. An equal opportunity employer: Opium smuggling networks in and around San Diego during the early twentieth century. Transnational Organized Crime 4:31-54.

Meier, Kenneth J., and Laurence J. O'Toole. 2001. Managerial strategies and behavior in networks: A model with evidence from US public education. Journal of Public Administration Research and Theory 11, no. 3:271-93.

Milward, H. Brinton, and Keith G. Provan. 2000. Governing the hollow state, Journal of Public Administration Research and Theory 10, no. 2:359-79.

New York Times. 2001a. Al Qaeda: Sprawling, hard-to-spot web of terrorists-in-waiting. New York Times, 30 September:B4.--. 2001b. A European dragnet captures new clues to bin Laden's network. New York Times, 12 October:B 10.--. 2001c. The inner workings of plot to blow up the U.S. Embassy in Paris. New York Times, 28 October:B1.--. 2002a. Front companies said to keep financing terrorists. New York Times, 9 November. Available at www.nytimes.com/2002/09/19/international/asia/19QAED.html?pagewanted= print&position=top.--. 2002b. Germans lay out early Qaeda ties to 9/11 hijackers. New York Times, 24 August.--. 2002c. On the path to the U.S. skies plot leader met bin Laden. New York Times, 9 October.--. 2002d. Terror attacks on "soft" targets complicate security. New York Times, 11 November.

Nye, Joseph S., Jr. 2002. The paradox of American power: Why the world's only superpower can't go it alone. Oxford: Oxford University Press.

The Observer. 2002. Just as you thought the world might be safer.... The Observer, 16 June.

O'Toole, Laurence J. 1997. Treating networks seriously: Practical and research-based agendas in public administration. Public Administration Review 57:45 52.

Pappi, Franz U. 1993. Policy-Netze: Erscheinungsformen moderner Politiksteuerung oder methodischer Ansatz. In Policy-Analyse. Kritik und Neurorientierung (PVS Sonderheft), edited by Adrienne Heritier, 84-96 Opladen: Westdeutscher Verlag.

Pappi, Franz Urban, Thomas Konig, and David Knoke. 1995. Entscheidungsprozesse in der Arbeits- und Sozialpolitik: Der Zugang der lnteressengruppen zum Regierungssystem uber Politikfeldnetze; ein deutsch-amerikanischer Vergleich. Frankfurt: Campus Verlag.

Pasternak, Judy, and Stephen Braun. 2002. Emirates looked the other way while al-Qaida funds flowed. Los Angeles Times, 21 January.

Peterson, John, and Margaret Sharp. 1998. Technology policy in the European Union. Basingstoke, Hampshire: Macmillan.

Porter Liebeskind, Julia, et al. 1995. Social networks, learning, and flexibility: Sourcing scientific knowledge in new biotechnology firms. Washington, D.C.: National Bureau of Economic Research.

Poucke, Willy van. 1979. Network constraints on social action: Preliminaries for a network theory. Social Networks 2:181-90.

Powell, Walter W. 1990. Neither market nor hierarchy: Network forms of organization. Research in Organizational Behavior 12:295-336.

Powell, Walter W., and Peter Brantley. 1992. Competitive cooperation in biotechnology: Learning through networks. In Networks and organizations, edited by N. Nohria and R. G. Eccles, 366-94. Cambridge: Harvard Business School Press.

Powell, Walter W., Kenneth W. Koput, and Laurel Smith-Doerr. 1996. Interorganizational collaboration and the locus of innovation: Networks of learning in biotechnology. Administrative Science Quarterly 41 : 116-45.

Powell, Walter W., and Laurel Smith-Doerr. 1994. Networks and economic life. In The handbook of economic sociology, edited by Neil J. Smelser and Richard Swedberg, 368-402. Princeton: Princeton University Press.

Provan, Keith G., and H. Brinton Milward. 1995. A preliminary theory of interorganizational network effectiveness: A comparative study of four community mental health systems. Administrative Science Quarterly 40:1-33.

Provan, Keith G., H. Brinton Milward, and Kimberly Roussin Isett. 2002. Collaboration and integration of community-based health and human services in a nonprofit managed care system. Health Care Management Review 27, no. 1:21 32.

Rittel, Horst W. J., and Melvin M. Webber. 1973. Dilemmas in a general theory, of planning. Policy Sciences 4:155-69.

Ronfeldt and Arquila, eds. 2001. Networks and Netwars The Future of Terror, Crime. and Militancy. Santa Monica, Calif.: The Rand Corporation.

Rothenberg, Richard. 2001. From whole cloth: Making up the terrorist network. Connections 24:36-42.

Schmid, Alex P. 1996. The link between transnational organized crime and terrorist crimes. Transnational Organized Crime 2:40-82.

Schneider, Volker. 1988. Politiknetzwerke in der Chemikalienkontrolle--Eine Analyse transnationaler Politikentwieklung Berlin: de Gruyter.--. 2000. Organisationsstaat und Verhandlungsdemokratie. In Gesellschaftliche Komplexitat und kollektive Handlungsfahigkeit, edited by Raymund Werle and U. Schimank, 243-69. Frankfurt/Main: Campus Verlag.

Schneider, Volker, and Raymund Werle. 1991. Policy networks in the German telecommunications domain. In Policy networks: Empirical evidence and theoretical considerations, edited by B. Marin and R. Mayntz, 97-137. Boulder/Frankfurt: Westview Press/Campus Verlag.

Simon, Herbert. 1973. The structure of ill-structured problems. Artificial Intelligence 4:181 201.

Smillie, lan, Lansana Gberie, and Ralph Hazleton. 2000. The heart of the matter: Sierra Leone, diamonds and human security. Partnership Africa Canada (Ottawa), January.

Washington Post. 2001a. Disconnect the dots: Maybe we can't cut off terror's head but we can take out its nodes. Washington Post, 17 September:C1. --. 200lb. A few loyal men direct bin Laden's sprawling network. Washington Post, 27 September:A1.--. 2002a. Al Qaeda's new leaders. Washington Post, 29 October:A1.--. 2002b. Key figure in Sept. 11 plot held in secret detention in Syria. Washington Post, 6 June.--. 2003. Washington Post, 19 January:A23. Available at www.washingtonpost. com/wp-dyn/ articles/A11802-2003Jan18.html.

Watts, Duncan. 2003. Six degrees: The science of a connected age. New York: W. W. Norton.

Williamson, Oliver. 1975. Markets and hierarchies: Analysis and antitrust implications. New York:

Free Press.

Jorg Raab

University of Konstanz

H. Brinton Milward

University of Arizona
COPYRIGHT 2003 Oxford University Press
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2003 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Raab, Jorg; Milward, H. Brinton
Publication:Journal of Public Administration Research and Theory
Date:Oct 1, 2003
Words:15005
Previous Article:Bounded rationality and political science: lessons from public administration and public policy.
Next Article:Transaction costs and institutional explanations for government service production decisions.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters