Organized Crime Research (kvl-homepage)

The Use of Models in the Study of Organized Crime

Paper presented at the 2003 conference of the European Consortium for Political Research (ECPR)
Marburg, Germany, 19 September 2003
by Klaus von Lampe


I would like to talk about the use of models in the study of organized crime. This touches the more fundamental question of how the social sciences should go about exploring such a highly complex and elusive issue as organized crime.
At the same time I try to avoid the intricacies of the philosophical debate on the nature, purpose and value of scientific models.
Models, this much is clear, are representations of reality, though on a lower level of complexity. But there is some controversy on what exactly models should do and how meaningful they can really be. In a way I side-step this debate by, first, briefly reviewing three key works that represent the prevailing convention in the organized crime literature on what a "model of organized crime" is supposed to be. Then I will turn to a general classification of models in the social sciences which distinguishes between causal and analytical models and which I find quite helpful, and I will argue for adopting an analytical model as the most appropriate conceptual framework for further research on organized crime.

"Models of Organized Crime"

Jay Albanese

The notion of "models of organized crime" has in the past been most closely linked to a threefold classification proposed by Jay Albanese (1989: 91-102; 1994), who distinguishes a "hierarchical model", a "patron-client model" and an "enterprise model" of organized crime. All three "models" originally referred to the American Cosa Nostra. The "hierarchical model" pertains to the official view of the Cosa Nostra as a nationwide bureaucratic organizational entity (Cressey 1969). The "patron-client model", represented by the works of Joe Albini (1971) and Francis and Liz Ianni (1972), re-conceptualizes the Cosa Nostra as a web of asymmetric ties embedded in local or ethnic networks, whereas the "enterprise model", proposed by Dwight C. Smith (1980), centers around economic activities and the primacy of market forces over group structures. As Jay Albanese himself has emphasized, these "models" really represent different ways of looking at organized crime, different "paradigms" in Albanese's wording, which he believes can fruitfully be combined to get a more complete picture (Albanese 1994).

Boronia Halstead

This understanding of "models of organized crime" has been further elaborated by Boronia Halstead in a paper published in a 1998 issue of the journal "Transnational Organized Crime" (Halstead 1998), and more recently in a paper by Phil Williams and Roy Godson that appeared in a 2002 issue of the journal "Crime, Law and Social Change" (Williams & Godson 2002).

Halstead (1998) distinguishes different "models" not only by the underlying conception of the nature of organized crime but also by specific social conditions that are assumed to be responsible for the emergence of one or the other manifestation of organized crime. She distinguishes two broad categories, "group-focused models" and "economic models", and within these categories differentiates various "models" that emphasize particular aspects, for example, the structure, activities and social embeddedness of criminal groups. Halstead highlights the explanatory power of these models with regard to factors that lead to or facilitate the emergence or shaping of organized crime phenomena on the micro or macro level.
On the micro level Halstead discusses, for example, how illegal enterprises can be perceived as organizations influenced by internal and external stakeholders. Halstead (1998: 8), drawing on the multiple-constituency approach in organization theory (see Bedeian & Zammuto 1991: 68), explains:

"Applying this model to organized crime, a particular illicit enterprise might be analyzed by identifying the various factions or stakeholders with an interest in the enterprise, examining the nature of the interest and assessing how the range of interests interact and what the power relationships between the interests might be. For example, in the market for an illegal commodity such as cannabis, these interests would include cannabis users, cannabis wholesalers, cannabis retailers, law enforcement policy makers, law enforcers, health policy makers, corrupt public officials, and other less obvious groups, such as the media. The interaction between these constituencies and the relative power relationships between them will determine the nature of the illicit enterprise. The multiple constituencies approach draws into focus the fact that agents that might have an impact on the structure and operation of illicit enterprise are not just those who gain directly from it."

Other models identified by Halstead relate to macro level phenomena such as illegal markets, which are discussed with regard to external regulation and internal business culture (Halstead 1998: 16).

Williams and Godson

Williams and Godson (2002) take the discussion yet another step further by linking certain social conditions with certain manifestations of organized crime and these, in turn, with certain social consequences or impacts.

In their discussion of a methodology for anticipating "the further evolution of organized crime", Williams and Godson distinguish several potentially predictive "models" that emphasize causal relations between certain environmental conditions, certain manifestations of organized crime and certain outcomes. "Political models", they argue, can explain the increase in particular types of crime and the emergence of criminal structures as the result of a weak state, an authoritarian form of government, and a low degree of the institutionalization of the rule of law (Williams & Godson 2002: 315-323). "Economic models", in their typology, include those approaches which attempt to predict organized criminal behavior with a view to the dynamics of supply and demand and the levels of control of illegal goods and services (Williams & Godson 2002: 322-328). "Social models", the third type of models defined by Williams and Godson (2002: 328), emphasize the cultural basis for organized crime, the idea of criminal networks as a social system, and the importance of trust and bonding mechanisms as the basis for criminal organization. The "strategic or risk management model", in turn, conceptualizes the activities of criminal enterprises, for example the corruption of public officials or the exploitation of safe havens, as means to minimize risks emanating from operating in a hostile environment (Williams & Godson, 2002: 335-339). Finally, Williams and Godson's typology includes "hybrid or composite models" which variously combine political, economic, social, and strategy factors to predict, for example, that in certain states characterized by weak government, economic dislocation, and social upheaval, transnational criminal organizations will take control of much of the domestic economy to use it as a basis for operating in host states where lucrative markets and supporting ethnic networks exist (Williams & Godson, 2002: 340-347).

Organized Crime as an Object of Study

What the "models" identified by Albanese, Halstead and Williams and Godson have in common is a strong orientation to concrete events and settings. The "models" are largely constructed with specific manifestations of organized crime in mind that have emerged under specific historical and cultural conditions.
The quoted authors try to overcome the resulting limits in applicability by combining different approaches. However, these composite models, although they potentially touch a wide range of issues, still fall short of an overall framework that is designed to consistently analyze and compare phenomena across historical and cultural variations, because these composite models arrange and link phenomena more or less as if the only possible constellations are those defined by specific historical cases.

In contrast, when we speak of organized crime research as a process of creating a cumulative body of knowledge (von Lampe 2002), we need a conceptual framework that allows for the empirical existence of any conceivable constellation of the phenomena that fall under the umbrella concept of organized crime, regardless of whether or not they resemble commonly known events or stereotypical imagery.

At this point, I think, I should briefly clarify my understanding of organized crime as the object of study.

My approach is based on the assumption that "organized crime" is first and foremost a construct, a 'notion vulgaire' in the Durkheimian sense (Durkheim 1973: 22), which reflects social reality as much as emotions, prejudices and ideologies of those involved in the construction process.
From a sociological perspective, such constructs cannot be accepted at face value. Rather, it is the duty of the social sciences to define and categorize the underlying phenomena and to explore through empirical observation what intricate links exist that would justify placing all these diverse phenomena in one theoretical context.

This very basic idea about the meaning of sociological inquiry, I believe, constitutes an appropriate measuring rod for determining the suitability of a model for the study of organized crime.

Causal Models and Analytical Models

In the following discussion I will use the term "model" in a narrow sense as proposed by sociologist Jonathan Turner (1991). He defines a model as "a diagrammatic representation of social events. The diagrammatic elements of any model include: (1) concepts that denote and highlight certain features of the universe; (2) the arrangement of these concepts in visual space so as to reflect the ordering of events in the universe; and (3) symbols that mark the connections among concepts, such as lines, arrows, vectors, and so on" (Turner 1991: 15-6).

Turner distinguishes two essential types of models: causal models and analytical models (Turner 1991: 17). Causal models (Fig. 1) aim at explaining a dependent variable by one or more independent variables and present a simple lineal view of causality. Analytical models (Fig. 2), in contrast, portray a complex set of connections among a set of variables (Turner 1991: 18).

Figure 1: Causal Model

Figure 2: Analytical Model

Causal Models of Organized Crime

Causal models are implicit in many works on organized crime. They tend to conceptualize organized crime, or one particular aspect of it, as a one-dimensional phenomenon varying along a spectrum from bad to worse. The "models" described by Halstead and Williams and Godson largely fall into this category.

A classical example (Fig. 3) is provided by Donald Cressey, who in his treatise "Theft of the Nation" on the American Cosa Nostra (Cressey 1969) discussed how the demands for illegal goods and services nurtured the emergence of ever more centralized and ever more powerful crime syndicates who eventually succeed in undermining public morals, neutralizing law enforcement through corruption and infiltrating the legal economy unless appropriate countermeasures are taken. This theoretical proposition can be depicted in a model comprising four elements: government, society, illegal markets and organized crime (von Lampe 1999: 308). While interrelations are acknowledged in both directions between the model elements, in the last instance the purpose is to explain variations in the power and reach of organized crime in the sense of an ultimately unified organizational entity.

Figure 3: "Cressey Model"

The major shortcoming of causal models is their narrow focus as a consequence of the conceptualization of organized crime, or an essential aspect of it, as a one-dimensional dependent variable. This makes it difficult, if not impossible, to account for variations in the manifestation or for different paths of development of organized crime. Cressey's model, for example, fails to consider the possibility that despite flourishing illegal markets, criminal groups may not necessarily become ever more powerful, but instead remain in a position of inferiority vis-à-vis corrupt political and business elites who maintain and possibly extend their control over state and economic resources (von Lampe 1999: 313-4), as has vividly been described by Bill Chambliss in his analysis of crime and politics in the city of Seattle (Chambliss 1978).

Analytical Models of Organized Crime

In comparison with causal models, analytical modeling schemes have two advantages for the study of organized crime: (1) they correspond much more to the complexity and multi-dimensionality of the structures, events and processes that are lumped together under the term "organized crime", and (2) they better fulfill the present needs of organized crime research, which is really still in its infant stage, by helping to tentatively order data and to formulate research questions. Given the fragmentary nature of the current knowledge on organized crime, analytical models can be employed as heuristic devices that display aspects of interest and map connections between them either based on existing empirical findings or based on plausible assumptions. Understood in this way, analytical models may form the starting point and a comprehensive framework for more systematic and better coordinated future research.

An Analytical Model of Organized Crime

In the remaining section of this paper I will outline one such analytical model of organized crime. The model is intended to capture the situation in a given geographical area that shows some level of political coherence, for example a municipality or a nation state.

Selecting Model Elements

The first step in the construction of an analytical model of organized crime is to determine what aspects of the social universe to include. This is not so much a definitional question than a matter of tentatively marking out a field of study. In the absence of a generally accepted, authoritative definition of the term "organized crime", the only fairly inclusive approach seems to be one which outlines the field of study by the scope of the public and scientific debate. Organized crime, then, is what people so label. This includes just about any kind of cooperation for the rational, i.e. non-impulsive, commission of illegal acts, regardless of the social status or the motives of the perpetrators (von Lampe 2002: 195). Accordingly, the elements of the model should be selected with a view to those factors that either represent or significantly influence the emergence and continued existence of patters of criminal cooperation for the rational commission of criminal acts.

I would argue that any meaningful model of organized crime has to include six basic elements, three representing what has variously been labeled organized crime, and three elements representing environmental factors.
The three core elements are:
- the actors who cooperate in rational, non-impulsive criminal activities,
- the structures that connect these actors,
- the criminal activities these actors are involved in,
The three environmental elements are:
- society
- government
- and the realm of public discourse, i.e. the media.

Even on this high level of abstraction, a variety of interesting connections can be mapped. To begin with, there is no organized crime without organized criminals, and these organized criminals are, at least in part, a product of their social environment (Landesco 1929), characterized, for example, by social and cultural cleavages. In turn, the types of crimes these criminal actors are engaged in may depend on personal skills, just as certain cooperative structures may in part depend on individual properties such as reliability and sociability. At the same time, criminal structures are influenced by the type of activity they serve. An extortion gang, for example, requires structures different from an illegal casino (Block 1983). Indirectly, then, social factors can influence the shape of criminal structures, for example, through the demand for particular illicit goods and services that require one or the other type of group structure. The same applies, of course, for the government. It has been repeatedly argued that law enforcement pressure impedes the emergence of complex criminal structures. Government plays a role in other ways too, for example by adopting crime prevention policies or by creating opportunities for criminal activities such as through raising taxes for goods like alcohol or cigarettes, not to mention the power to define certain acts as illegal. The Media, finally, can be a crucial factor by raising or shifting attention to certain phenomena, for example certain areas of crime, certain criminal groups or certain ethnic groups.

Determining the Level of Abstraction

The second step in the construction of an analytical model of organized crime is to determine the level of abstraction.

There are manifold ways to further differentiate the model and to dissolve the elements into ever smaller units of analysis. How far one should go depends largely on the respective research question that the model is supposed to help formulate or answer. There are, however, three differentiations pertaining to the nature of criminal structures that I think need to be made in order to avoid comparing apples and oranges.

The first differentiation that I propose is between criminal networks in the sense of webs of criminally exploitable ties as latent structures, on the one hand, and patterns of criminal cooperation as manifest structures, on the other hand. The underlying assumption is that a great deal of what is labeled organized crime involves the flexible use of personal ties for the commission of criminal acts (Van Duyne 1996; Fijnaut et al. 1998; Potter 1994). This means that there are webs of personal ties connecting criminal actors that could be activated for criminal cooperation but only a certain share of these ties are actually used at any given point in time.

The second differentiation accounts for the fact that criminal structures can serve different functions. I have argued elsewhere (von Lampe 2001a; 2003) that all criminal structures serve essentially one or more of three purposes: economic, social and quasi-governmental. The crucial distinction here is between economic and non-economic functions because it allows to analytically separate illegal enterprises in a broad sense, including market-oriented groups as well as predatory criminal groups, from fraternal associations which only indirectly, through their individual members, are involved in crime for material gain. It should be obvious that there are substantial differences in the conditions for their emergence and their impact on the immediate and broader environment, notwithstanding the fact that economic and non-economic functions may empirically overlap.

The third differentiation pertains to micro and macro structures within the sphere of illegality. There are not only criminal collectives that form out of economic or socio-cultural interests, but there are also, potentially, overarching structures that concentrate power in a given illegal market or geographical area. Following the distinction between economic and non-economic structures, these entities can be divided into two categories: illegal market monopolies and quasi-governmental 'power syndicates' (Block 1983).
On this level of abstraction, the model can address two key questions that have been raised in the organized crime debate: (1) how patterns of criminal cooperation emerge and are transformed, and (2) how within those criminal structures positions of power develop that are relevant for the criminal structures themselves and for society at large (see von Lampe 2001b).

Connecting the Model Elements

The third step in the construction of an analytical model is to determine by what kind of links the model elements are connected. In the way modeling is understood here, this is an ongoing process which accompanies the entire research process. From the onset it is important to note that elements are not necessarily connected simply by unidirectional causal links. Rather, it can be expected to find direct and indirect connections working in different directions.

Figure 4: An Analytical Model of Organized Crime

In the shown diagram (Fig. 4), connections between the elements are mapped based on a review of academic literature, mostly pertaining to the United States (von Lampe 1999: 322). It seems clear that the limits of diagrammatic representation are quickly reached. However, the model, I would argue, still serves to capture some important complexities.

To illustrate this point, let me as an example briefly discuss some of the factors potentially influencing the emergence of an illegal monopoly of violence over a given territory in the hands of a crime syndicate. The ideal typical picture is that of illegal actors operating in this geographical area being subordinated to a criminal group which controls the use of violence, sets and enforces rules of conduct and levies a tax on criminal activities. Such a constellation can in some respects be regarded as the natural outcome of a process that takes effect under conditions of illegality. Assuming a state of anarchy as the starting point of the development, criminal actors have to decide how to allocate their resources between productive and unproductive activities, or, in other words, between generating an output or influencing its distribution. The assumption is that one group of actors will emerge with an ever increasing potential for violence while all other actors will find it increasingly useless to invest resources in the potential for violence themselves (Skaperdas & Syropoulos 1995). In fact, the monopolization of violence can work for the benefit of all, because it will tend to reduce the overall level of violence (Hellman 1980; Luksetich & White 1982).
The emergence of a 'power syndicate', however, appears to be dependent on the presence or absence, respectively, of a number of conducive and countervailing factors. First of all, 'power syndicates' need to have personnel capable of using violence effectively. Secondly, the existence of a 'power syndicate' implies the ability to monitor the criminal activities of other criminal actors. Given the clandestine nature of most areas of crime, this will tend to be possible only where criminal actors are integrated in dense networks that guarantee a free flow of information, or where illegal activities are characterized by high visibility, continuity and fixed places of business, such as in the cases of illegal casino gambling and the street sale of illicit goods (Schelling 1971). This, in turn, presupposes not only a demand for these goods and services but also a certain level of tolerance on the part of the society at large and a lack of motivation, possibly brought about through corruption, or a lack of resources on the part of law enforcement to stop such illegal activities.
Furthermore, in order to maintain their position in a cost efficient way and to fend off free-riders, 'power syndicates' will try to establish an identifiable reputation for the effective use of violence (Gambetta 1993; Reuter 1983). This requires a sufficient level of recognizability which, at the same time, increases the visibility for law enforcement.
Finally, the media potentially play an ambiguous role in this context. On the one hand, the media can help establish a reputation for particular criminal groups. On the other hand, concentrated media coverage will put pressure on law enforcement to focus on these same groups. These considerations imply a wave-like path of development characterized by the rise and subsequent downfall of a 'power syndicate' as the most likely scenario.
This example, of course, would have to be elaborated further in a thorough analysis. There are more ramifications to it than can be displayed in such a diagram. Additional propositions would surely show an even higher degree of interconnectedness and a wider circle of influencing factors. Despite these limits, I would argue that the model does help to put all the issues in perspective that have been raised in the organized crime debate.


To summarize my argument, let me stress again that I consider the importance of models in the study of organized crime to be not so much in presenting final conclusions. Rather, I see models as heuristic devices that guide and systematize research.
The analytical model I have only sketched is designed, one the one hand, to account for all dimensions of the problem that need to be considered, be they sociological, psychological, cultural, economic or political. On the other hand, the model is designed to apply to all conceivable historical and geographical settings. This does not mean that all manifestations of organized crime are more or less identical and that the model elements will always have the same significance. On the contrary, it is the diversity of manifestations of organized crime under varying circumstances from which we can expect to gain the most valuable insights. The model provides a conceptual scheme for the analysis of each and every case in its own right, but from a comparative perspective with the use of the same terminology, within the same broad conceptual framework and with the same research questions in mind.


Albanese, Jay, Organized Crime in America, 2nd ed., Cincinnati, OH: Anderson, 1989.
Albanese, Jay, Models of Organized Crime, in: Robert J. Kelly, Ko-lin Chin, & Rufus Schatzberg (eds.), Handbook of Organized Crime in the United States, Westport, CT: Greenwood, 1994, 77-90.
Albini, Joseph L., The American Mafia: Genesis of a Legend, New York: Appleton-Century-Crofts, 1971.
Bedeian, Arthur G., & Zammuto, Raymond F., Organizations: Theory and Design, Chicago: Dryden, 1991.
Block, Alan A., East Side, West Side: Organizing Crime in New York 1930-1950, New Brunswick: Transaction Publishers, 1983.
Chambliss, William J., On the Take: From Petty Crooks to Presidents, Bloomington: Indiana University Press, 1978.
Cressey, Donald R., Theft of the Nation: The Structure and Operations of Organized Crime in America, New York: Harper & Row, 1969.
Durkheim, Emile, Les Règles de la Méthode Sociologique, Paris: Presses Universitaires de France, 1973.
van Duyne, Petrus, The phantom and threat of organized crime, Crime, Law and Social Change, 24(4), 1996, 341-377.
Fijnaut, Cyrille, Bovenkerk, Frank, Bruinsma, Gerben, & Van de Bunt, Henk, Organized Crime in the Netherlands, The Hague, Netherlands: Kluwer Law International,1998.
Gambetta, Diego, The Sicilian Mafia: The Business of Private Protection, Cambridge, MA: Harvard University Press, 1993.
Halstead, Boronia, The Use of Models in the Analysis of Organized Crime and Development of Policy, Transnational Organized Crime, 4(1), 1998, 1-24.
Hellman, Daryl A., The Economics of Crime, New York: St. Martin's Press, 1980.
Ianni, Francis & Reuss-Ianni, Elizabeth, Family Business: Kinship and Social Control in Organized Crime, New York: Russell Sage, 1972.
von Lampe, Klaus, Organisierte Kriminalität unter der Lupe: Netzwerke kriminell nutzbarer Kontakte als konzeptueller Zugang zur OK-Problematik, Kriminalistik 55(7), 2001(a), 465-471.
von Lampe, Klaus, Not a Process of Enlightenment: The Conceptual History of Organized Crime in Germany and the United States of America, Forum on Crime and Society, 1(2), 2001(b), 99-116.
von Lampe, Klaus, Organized Crime: Begriff und Theorie organisierter Kriminalität in den USA, Frankfurt am Main: Peter Lang, 1999.
von Lampe, Klaus, Afterword: Organized Crime Research in Perspective, in: Petrus C. van Duyne, Klaus von Lampe, & Nikos Passas (eds.), Upperworld and Underworld in Cross-Border Crime, Nijmegen: Wolf Legal Publishers, 2002, 189-198.
von Lampe, Klaus, Organising the nicotine racket: Patterns of criminal cooperation in the cigarette black market in Germany, in: P. C. van Duyne, K. von Lampe, & J. Newell (eds.), Criminal finances and organising crime in Europe, Nijmegen, Netherlands: Wolf Legal Publishers, 2003, 41-65.
Landesco, John, Organized Crime in Chicago, in: The Illinois Crime Survey, Chicago: The Illinois Association for Criminal Justice, 1929, 823-1087.
Luksetich, William A., & White, Michael D., Crime and Public Policy: An Economic Approach, Boston: Little, Brown and Company, 1982.
Potter, Gary W., Criminal Organizations: Vice, Racketeering, and Politics in an American City, Prospect Heights, IL: Waveland, 1994.
Reuter, Peter, Disorganized Crime: The Economics of the Visible Hand, Cambridge, MA: The MIT Press, 1983.
Schelling, Thomas C., What is the Business of Organized Crime?, Journal of Public Law, 20(1), 1971, 71-84.
Skaperdas, Stergios, & Syropoulos, Constantinos, Gangs as primitive states, in: Gianluca Fiorentini & Sam Peltzman (eds.), The economics of organised crime, Cambridge, UK: Cambridge University Press, 1995, 61-82.
Smith, Dwight C., Paragons, Pariahs, and Pirates: A Spectrum-Based Theory of Enterprise, Crime & Delinquency, 26(3), 1980, 358-386.
Turner, Jonathan H., The Structure of Sociological Theory, 5th ed., Belmont, CA: Wadsworth, 1991.
Williams, Phil, & Godson, Roy, Anticipating organized and transnational crime, Crime, Law and Social Change, 37(4), 2002, 311-355.