About a year ago, I wrote my dissertation about — well, I suppose it changes depends on who I’m talking to, because it was about a lot of things — but mostly about transnational organized crime. I’d attempted to bring in a lot of things we learned about fighting networks — and building our own networks — from counter-terrorism into countering transnational organized crime. In the course of my research, I explored the British Suppression of the Atlantic Slave Trade and the Coast Guard’s Rum War during Prohibition, and in both cases, I found that questions of governance were inextricably intertwined with questions of network suppression.
In the case of the Rum War, a draconian law was imposed through process politics without actually gaining the sort of public support required to make it enforceable. In the case of the Slave Trade, after decades of ineffectual ‘slaver-plinking’ — interdicting and seizing individual slaving vessels — the British captains pushed ashore (generally against the wishes of the Foreign Office.) As dramatized in the final scene of Amistad, this allowed them to destroy the slaver’s infrastructure ashore, but the challenge was not finding the forts — the cruisers knew exactly where they were and would picket them — but getting the legal authorization from the West African potentates to conduct attacks into sovereign territory. Demand for these treaties soon outstripped supply as the slave trade retrograded into its last remaining favorable climes, and this led to the British picking sides in succession struggles, which entwined them in imperial questions which outlasted the campaign to destroy the trade.
As in COIN, questions of governance are inextricably linked to these sorts of campaigns. Which means that we need to have some sort of working bedrock theory of ‘what is governance’ if we are going to seriously engage in COIN, transnational persistent counterterrorism (CT), or counter-transnational criminal organization (TCO) campaigns. Fortunately, the National Defense University has done some excellent work on convergence, which explores links between these sorts of campaigns. In this essay, excerpted from an appendix of my dissertation, I attempt to build a rudimentary working bedrock theory of governance, which we might apply to COIN, CT and counter-TCO efforts.
Foundations: Understanding the Convergence Literature. As articulated by Admiral (Ret.) James Stavridis, convergence is the “dark side of globalization.” This is where low-level illicit actors of all sorts make use of common global information and transportation architectures, and hence begin to emulate and even partner with each other. There are two key elements of this line of thought — 1) common infrastructure and the erosion of governance, and 2) the collision of illicit economics and radical ideology. The latter will occupy the bulk of our time, and will lead us to a rudimentary general working theory of contested governance.
Material Elements: Criminal-Insurgency-Terror Infrastructure Overlap.To the first point, there are three levels where illicit markets and violent extremist organizations (VEO) can share infrastructure. The deepest level is explicit cooperation, which is the most involved and straight-forward, but not necessarily the most dangerous. The intermediate layer, shared knowledge and practices, may qualify for that distinction: knowledge of border holes, access to smuggling and safe house networks, etc. facilitates unscripted nefarious activity. It is the fluidity and unpredictability of this layer that makes it dangerous. Finally, on the least deliberate level, criminality erodes governance, which lowers the tide for all sorts of badness as well. We explore these three presently:
Explicit Cooperation. The simplest level is explicit cooperation — where terror groups are selling illicit drugs, or mafia are contracted by violent extremists to smuggle weapons, or groups are doing both political violence and illicit economics all at once. This level of cooperation is deceptively complex — the temptation is to assume that all bad things go together, the interaction between illicit actors depends on specific needs of both organizations and the governance context. These relationships are rarely as simple as they seem. For instance, while a smuggling ring could plausibly serve both terror infiltrators and human trafficking, a number of incompatible constraints would generally prohibit a ring that does one from doing the other. Human trafficking requires an invasive commodification and exploitation process, which involves a number of security vulnerabilities for the ring that might deter a terrorist. Similarly, a human trafficking ring most likely could not afford to weather the kind of heat that a terror association would bring.
These problems come to a head over terror-narcotics connections. Insurgent groups like the Fuerzas Armadas Revolucionarias de Colombia (FARC) and the Taliban engage in illicit narcotic production. A simple model would attempt to eradicate these illicit crops in order to dry up the funding base for these groups, but such a policy often backfires. As Vanda Felbab-Brown argues, a state needs strong governance in an area if it is going to successfully conduct an eradication campaign, and the whole point of an insurgency is to weaken (and thereby eventually replace) state governance. Therefore, eradication campaigns in poorly governed areas are likely to aid the insurgents by alienating the population, which exacerbates the problem. The relationship between terror, crime and governance must therefore be more complex.
Shared Knowledge. The second layer of convergence looks at shared practices and knowledge. To continue our trafficking-terror case, while one ring is unlikely to engage in both due to incompatible security concerns, knowledge of low-risk cross-border roots would be equally useful to both. Every time any group penetrates a border, some knowledge is generated about how to do so again in the future. Similarly, if a drug trafficker learns how to use ‘bit-coin’ as a virtual hawala to throw law enforcement money-chasers off his scent, that tactic is now a part of illicit corporate knowledge and plausibly accessible by terrorists. Sharing tactics, techniques and procedures (TTP) does not require nearly as much commitment and does not impose the same degree of vulnerabilities, especially when shared across globalized information infrastructures. Do-it-yourself improvised explosive device (IED) websites speak to this point. This ‘shared practices’ layer broadly parallels our previous insights about structural efficiency and ‘market-like’ forms from network theory.
Assertion: Shared practices and knowledge are key centers of gravity for illicit market suppression. Tactical learning and dissemination are functions of structural configurations, with ‘market-like’ structures performing best. The suppression regime has the lever of interdiction to de-optimize these structures.
Corrosive Anti-Governance. The third layer of convergence considers the corrosive effect of violent political and criminal actors on governance. Crime will tend to corrupt state officials, terror disrupts confidence, and insurgencies disrupt infrastructure; all of these are bad for the efficient provision of government functions. We will explore this point at length later, but it seems uncontroversial that illicit actors of all sorts reinforce each other in this third and most abstract layer. This broadly parallels our intuitions about the necessity of social support in illicit market suppression.
Assertion: Governance is a key center of gravity for illicit market suppression. Governance allows the suppressor to bring force to bear, and supports demand restructuring. Poor governance allows the illicit market to thrive by filling their coffers through illicit demand. The suppression regime must extend some form of governance into the spaces occupied by the illicit market; the illicit market will attempt to erode governance and thereby expand.
Altogether, these three points of logistical convergence illustrate that there are some natural equities shared between myriad criminal groups and violent political actors. This is common-sensical, but we can therefore discuss these overlaps while avoiding cheap rhetorical moves that place all the badguys in the same ‘league of evil.’ This is rarely true — just as in traditional states compete, cooperate and clash in conflict, illicit actors too have no permanent allies or enemies, but permanent interests. So human trafficking and terrorism might go together in one circumstance — for instance, ISIS engages in aspects of both, but is able to because of territorial control — but not in others, where the time and interaction required to exploit trafficked people presents more of a liability than an asset to a group.
In general, following Shapiro’s Terrorist Dilemma, we should expect to see low-exposure criminality (clandestine laundering) for groups with high security concerns, and high-profit and high-exposure crimes (narcotics production, human trafficking) for groups with low security concerns. The most common way to reduce security concerns is by gaining some degree of territorial control, which means that we need to consider governance. In this, we find the greatest contribution of the convergence literature — crime is effectively illicit economics, which is low politics, and the violence deployed toward some ideological end, as in terrorism or insurgent shadow governance, is high politics. Together, these paint a mirror image of the state and civil society — in order to understand convergence, we must understand governance.
Legitimacy Elements: A Rudimentary Theory of Contested Governance.To explore this facet of convergence, we’ll use arguments from mafia scholar Diego Gambetta and Mao Zedong’s theories of revolutionary warfare. Mao argues that guerrilla war takes three phases. First, the revolutionaries organize and prepare; second, they conduct terror attacks and transition to an insurgency; finally, they carry out a conventional war to consolidate control. Each phase increases the degree of the governance enjoyed by the revolutionary group. However, governance requires answering legions of practical, banal questions about service provision and the like. Accordingly, each phase adds more low-politics distributional concerns to the high-politics ideology-centric revolutionary group.
Max Weber’s work on legitimacy explains this centralizing, stabilizing tendency. In a simplified version of his model, there are two forms of power — coercive and legitimate. Coercive power is the raw exercise of compulsion, which enables you to get exactly what you want, but at a high price. Legitimacy makes power go much farther by embedded it in a web of nested bargains. However, with each of these bargains, you promise to abide within certain restraints, which stabilizes things for everyone but limits your freedom of maneuver. Therefore, as the revolutionary group becomes more successful, it enters into these sorts of civil society bargains in order to efficiently extend its power. Accordingly, they lose some of their revolutionary fervor as they have to put it into practice. This results in a mix ideological high politics with distributional low politics that resembles statecraft.
The opposite process, low-politics group moving toward high-politics, is described in Gambetta’s analysis of the Sicilian mafia. Describing the Mafia as a provider of private protection, he presents a model of organized crime that looks remarkably like statecraft. Channeling Charles Tilly’s “state as a protection racket,” the various Italian organized crime groups perform protection and contract enforcement, which are core tasks of the state. Since illicit groups do not directly have access to these state resources, larger illicit groups can trade on their reputation to fill these functions.
Starting as a pure low politics group — solely about making money — these groups venture into higher politics in order to better make money. Contract enforcement and protection are cheaper mechanisms of conflict resolution than combat, but in order to provide these things, one must venture into ideology to explain why one group is authorized to perform these functions. Climbing even further into high politics, corruption provides an excellent means to hijack the state. This allows a criminal group to extract rents from the system, perform banking functions, and use legal protection strategies. At some point, the criminal network becomes indifferentiable from the state. As Mao’s model moves from pure high politics into a mix of high and low politics, Gambetta’s model moves from pure low politics toward a similar mix of politics.
This model provides two key assumptions that undergird a model of illicit market suppression. First, illicit market suppression is fundamentally about governance. To a very broad reading of regime theory, any place where people conduct business has some degree of governance. The problem with the West African coastal seas during late-phase Atlantic slave trade suppression was not so much a lack of governance, but governance of the wrong kind. Slave traders had strong expectations about how business was conducted in this space, and the British continuously struggled to supplant that governance with one more conducive toward their ends.
To this point, one key strategy debate in the British suppression centered on the relationship between sovereignty and governance. The British approach took a tack of increasing governance over the commons through shared sovereignty. They did so through international courts and treaty networks. The Americans, allergic to the idea of participating in any sort of shared sovereignty, preferred an approach that attacked slaver governance by removing from them the protections of sovereign flags. This piracy-based approach treated slavers as hostis humanis generis (‘enemies of all humanity’) and therefore protected under no flag.
Assertion: The core struggle in both COIN and Counter-TCO is about governance. Both sides need to build governance conducive to their ends, and their ends are mutually exclusive.
This model parallels an intuition from our building blocks: the natural advantage of the suppressor is the ability to marshal forces, but they are constrained in how they use them. The legitimacy bargains that underwrite the state allow it to achieve tremendous economies of force, but it comes at a cost of the limitations imposed by those bargains. Accountability, oversight and political constraints all dramatically reduce the suppressor’s ability to innovate and adapt, but they legitimacy it provides grants deep pockets so long as public support for the campaign remains. Conversely, the natural advantage of the illicit market is its adaptability, but it cannot win a pitched battle against the state because of the state’s ability to draw on incomparable reinforcements. Continuing the resource exchange ratios analogy, the state’s large-scale legitimacy advantage gives it an excellent ‘check-writing ratio,’ and the adaptability of the illicit market gives it excellent ‘check-cashing ratios’ on small scales. (These erode as the scale increases, which result in stalemates and nested bargains.)
Assertion: The suppression regime enjoys significant advantages in its ability to marshal forces (‘check-writing ratio’), but disadvantages in small-scale innovation (‘check-cashing ratio.’) The converse is true for the illicit market.
Theory: Governance Arc Model. To further develop the idea that governance is at the heart of the struggle against illicit markets, we’ll formalize this intuition. If governance is the ability to implement a set of rules, then there should some trade-off between the volume of rules and compliance with those rules for a given amount of governance. This results in a ‘Pareto frontier’ of all possible combinations of rules (polity depth) and rule-followers (polity breadth.) Building on this assumption, setting a policy on a given issue results in a combination of rules and rule-followers along this frontier. These rule followers, the ‘constituents,’ and these rules, the ‘constitution,’ together make up a polity.
Changes to rules result in public choice problems. If increasing the depth of rules decreases the number of rule-followers, adding rules results in a split between the ‘reformed polity’ (those who follow the new rules) and an ‘excluded polity’ (those who do not.) The resulting conflict pits these two against each other, with the reformed polity attempting to enforce the rule, and the excluded polity doing their best to impede the process. The excluded polity’s resistance (clandestinely carrying out proscribed practices, not turning in others who do, hampering enforcement efforts, and/or public campaigning) increases the cost of governance in the hopes of making the new rule unenforceable or at least uneconomical.
Therefore, in order to successfully conclude the reform attempt, one must expand governance. These expanded enforcement powers re-incorporate the excluded polity into an expanded polity through coercion, persuasion or co-optation. This broadly aligns with what we know of historical and contemporary suppression attempts. During the British suppression of the Atlantic Slave Trade, the Royal Navy massively expanded operations and infrastructure in West Africa to enforce international law. Eventually, the British slave-holding elites agreed to a compensated emancipation plan and were co-opted into the official abolitionist policies. The expansion of law enforcement during Prohibition was no less dramatic, which parallels the cost and infrastructure of the contemporary anti-narcotics campaign. The anti-human-trafficking movement is struggling to move forward from rule reform into expanded governance.
Relating this process back to our previous insights about the relationship between governance and illicit market suppression, we can imagine the magnitude of this governance vector as ‘efficiency.’ The more efficient the means of governance are, the more ‘bang for the buck’ we get out of one unit of governance. Similarly, governance support for the suppression campaign can be imagined as the angle of the governance vector, as it signifies the will to maintain the higher level of rules. The suppressor must maintain the will to continue the campaign, and increase efficiency to expand governance in order to enforce the new sets of rules.
This model implies that suppression efforts are subject to exogenous governance shocks. A positive shock enables new rules, while a negative shock forces returns to lower levels of rules. The Prohibition case supports this assertion: the creation of the income tax (in lieu of alcohol excise taxes) provided resources to begin the campaign; the stock market crash dried up these resources and made a reinstated alcohol excise tax, and hence reinstated alcohol, far more attractive.
Assertion: Governance can be described as a combination of support and effectiveness. Support is the political will to continue a campaign (‘check-writing ratio,) and effectiveness is the how efficiently that will translates into effects in relevant domains (‘check-cashing ratio.)
Testable Implication: Exogenous shocks to the available resources for governance should alter the set of possible outcomes available to the suppressor.
Extending this model to include competing sub-groups reveals a key insight about illicit market suppression: groups that are strongly connected internally but poorly connected to the whole stand to gain the most by more restrictive rule changes. This ‘best marginalized group’ stands to gain the most from an illicit market, due to its internal organization, and stands to lose the least from the sanctions of the larger society. For instance, the various ethnic organized crime groups during prohibition, or the pirates who inherited the late-phase illegal Atlantic slave trade, both were strongly internally organized but loosely connected to the larger society.
We can picture this process on the model by depicting a number of competing models of governance on the governance arc graph. The dominant polity enjoys a ‘Peter principle,’ where they can incorporate more groups, which in turn increases their governance, which then allows them to incorporate even more groups.
There are three possible relationships between one of these sub-groups and the larger dominant form of governance. First (green arc) a sub-group can be fully incorporated into the larger governance construct (blue area.) These are groups that under normal sets of policies will remain under the aegis of the primary polity unless extreme policies are pursued. An example of this is the Green Party in the United States on the issue of environmentalism. Normal political bargaining is the best strategy for the main governing body for these groups. Second (red arc), a subgroup may have views so wildly divergent from the main group that no reconciliation is possible. This might be Al-Qaeda’s relationship with the American polity on any of a number of issues. Such a group can only be managed as a threat. Between these two, and most interestingly (orange arc), a subgroup may have some equities in the main group’s arc of possible polities and some outside. With such a group, a truce is possible where members of the subgroup comply with some lessened set of rules in exchange for stability (purple area). Consider, for instance, implicit truces and stalemates with areas of inner city crime.
When we introduce rule changes into this model, we find that the structure of truces and hostile groups changes. Increasing the rule-set causes groups with whom truces had been established to become fully excluded and therefore hostile, and groups that had previously been incorporated into the polity become marginalized. One interesting implication here is that the best-organized liminal group inherits the space previously occupied by licit pursuits (contested polity in the top graph below.) However, by increasing governance, these contested groups can be re-incorporated into the polity and with truce possible with their remnants (bottom graph below.)
Incorporating this insight back into the larger convergence model reveals the awkward possibility that one of the extenders of state governance is a network of implicit bargains associated with stalemates against illicit groups (graph below.) Fully hostile groups may survive below a sense-making frontier, provided they do not grow large enough. However, groups that share some degree of overlap with the core governance can reach implicit truces. This is due to a power dynamic familiar to counter-insurgency — if the population does not support you, then it becomes very difficult to gain the low-level intelligence necessary to prosecute minor crimes.
Accordingly, to maximize the impact of the limited governance resources available, law enforcement may elect to forgo full enforcement of these sorts of crimes in areas where there is a lack of population support. Such a strategy likely involves probing enforcement, where these stalemates are revisited occasionally with periodic enforcement efforts, or skirmishing enforcement, where overt illicit behavior is prosecuted in order to force the illicit organization to take basic security countermeasures. Should the area or group in question actively defy sovereignty or break rules the society is fully committed to enforcing, such as murder, then the state will marshal resources and the stalemate no longer holds. This results in partial compliance with laws, and hence implicit truces.
For example, a veteran law enforcement officer tells a story from a ride-along with the Las Vegas Police Department. Investigating an unrelated event, two police officers hear shots fired in the distance. Within a few minutes, a call to investigate a gunshot wound comes across the radio. Upon arriving on the scene in a high-crime area, they encounter an individual who was stable but clearly shot. Presented with standard queries, this individual responded that he had fallen down and neither needed nor desired any assistance. The officers asked whether or not he was sure, and he responded that he was. They proceeded to drive away and continue their patrol.
This whole scenario demonstrated a cost-imposition strategy on the part of a hostile population — the individual clearly would not assist an investigation, which would make finding the perpetrator and achieving a conviction nearly impossible. Moreover, since he was the aggrieved party yet not interested in a police-mediated outcome, clearing the case would result in little benefit to the department. Accordingly, the officers chose to abandon a high-cost/low-benefit potential case in order to focus resources elsewhere. Law enforcement also can pursue a cost-imposition strategy. If a crime wave or a politically salient murder catches the attention of a mayor, law enforcement can respond by surging enforcement to the affected area. Doing so greatly increases the cost of doing business for previously tolerated practices, resulting in selection effects for criminals who avoid triggering these sorts of responses. These two countervailing forces produce an equilibrating tendency.
The insight here is that mutual cost-imposition strategies on the part of law enforcement and criminals result in stalemates, which lead to implicit bargains. In poorly governed areas, criminal networks can make routine enforcement so costly as to become prohibitive. Law enforcement can similarly make it difficult for criminals to do business through random patrols and occasional large-scale busts. These two strategies give both sides leverage on the other, which results in high-priority laws being enforced at the expense of low-priority laws. On a larger level, these sorts of bargains bring stability to an overall enforcement strategy. However, changes to the rule-set lead to changing balances of power. This results in a struggle to control the newly opened illicit market space, which alters this network of implicit bargains. In turn, this yields conflict, which is the primary negative externality of suppression campaigns.
Assertion: The ‘best marginalized group,’ or the most organized liminal group, stands to inherit any proscribed previously licit markets. They have the most to gain from provisioning the now-illicit good or practice, and the least to lose due to their lack of integration in the larger society.
Assertion: In a stable state, law enforcement and criminals can stabilize stalemates resulting from mutual cost imposition strategies by way of implicit bargains. However, a change in the rule-set disrupts these bargains, yielding conflict until a new equilibrium can be achieved between law enforcement and all criminal players.
There is an upside of this gritty realization — since some set of bargains is going to exist regardless, we can exert some leverage in deciding with whom we bargain. Paralleling COIN strategy, both law enforcement and illicit networks must navigate within the larger population. For both sides, civil society can either amplify or inhibit their available power. If illicit networks are ‘wrong kind of civil society,’ then we can also bargain with the right kinds of civil society. Neighborhood watch and community policing initiatives are legitimate bargains with civil society, where law enforcement proves itself helpful and trustworthy, and the population cooperate by providing intelligence and cooperation in prosecutions. In this way, law enforcement can shape the environment by choosing formal bargains with neighborhoods rather than informal bargains with illicit networks. Ideally, this allows positive civil society to supplant uncivil society and improves long-term governance.
We can then formulate James Q. Wilson’s ‘Broken Windows’ theory as a population-centric la enforcement strategy. If a hostile population dramatically increases the cost of investigating even major crimes, then eradicating minor crimes is a relatively inexpensive way of changing the operating environment. Becker’s focus on major crimes, much like kinetic-centric COIN strategy, may be more expensive in the long run if it cedes the population to criminal networks.
Therefore, civil society partnerships are an essential conductor of governance between the state and the street level. This cashes out in two different ways — first, if you lose the population on the suppression issue, then you lose your support and with it the whole campaign. Second, if you network well with the population, you gain suppression efficiency through intelligence and general support in the legal system.
Assertion: Civil-society formal bargains, such as in neighborhood watches and community policing, serve as force extenders. These replace the need for implicit bargains with illicit networks, and reinforce suppression regime efficiency and support.
Application: Sensemaking Below the Line. Bringing this idea back to James C. Scott’s model from Seeing Like a State, the state must constantly grapple with the limitations its ‘sense-making frontier’ when enacting any rule changes (‘Detect/Counter Line’ on graphic below.) Illicit groups fare far better than the state at low-level innovation, but they do so for the same reasons that positive civil society and economic actors prosper in local areas. Since the state has a difficult time making sense below a certain level of organization, it must find a way to deal with these low-level threats.
The traditional approach to solving this ‘sense-making below the line’ problem is to ‘drop the line.’ This is done by improving sense-making through surveillance and enforcement technologies. Classic suppression (graphic below) uses more resources to achieve higher fidelity on a given problem set, thereby dropping the sense-making frontier further. However, this still runs into the problem of geometrically worsening resource exchange ratios on the part of the state as we examine more localized problems.
A more draconian approach to this problem denies low-level threats access to civil society space by disrupting that space altogether (graphic below.) This ‘scorched earth’ solves the ‘sensemaking below the line’ problem by making the space below the line uninhabitable. For obvious reasons, this approach is unacceptable for liberal democracies. It has the additional unpleasant side effect of radicalizing anyone who manages to endure the use of these tactics. Still, the logic of this approach should be understood, as its use is not unprecedented — as depicted in the classic film The Battle of Algiers.
The third, and preferred, approach to the ‘sensemaking below the line’ problem is ‘partnering below the line.’ While state actors have a difficult time making sense of the world below a certain level of specificity, civil society actors do not (graphic below.) Through public-private partnerships, these actors can partner with the state in order to allow the state to bring its force to bear in a locally relevant and contextualized way. Community Policing strategies follow the broad contours of this logic. The problem with this approach is that it low-level law enforcement intelligence becomes subject to the preferences of civil society actors, though from a civil liberties perspective, that limitation may be acceptable.
If the ‘partnering below the line’ approach is indeed the preferred method of dealing with this problem, then we return to the primacy of networks and partnerships. As we discussed with the ‘check-cashing ratio,’ if civil society is on board with a suppression campaign, they will allow the resources of that campaign to be used far more effectively provided they are effectively integrated into that campaign. There are two key pieces here that echo a refrain from the rest of the model: this depends on the support of civil society to provide this information, and it depends on the efficiency of the suppression regime to receive this information. Therefore, we return to efficiency and support.
Assertion: Public-private partnerships are crucial for success in a suppression campaign. Civil society actors can make sense of low-level contextualized problems better than the state, but the state retains more power; in partnership both gain the resource advantages of the state and the contextualization advantages of the low level actor. This partnership is contingent upon public support and upon the efficiency of the suppression regime to absorb that support.
This public-private partnership model will provide foundational assumptions to the applied policy recommendations in this dissertation, as it is the optimal option for dealing with the crucial sense-making problem. Altogether, the convergence literature has given us a foundation for solving this problem, as well as for considering broader issues of governance amidst a spectrum of terror, insurgent, and criminal threats.
Dave Blair is a U.S. Air Force officer and is a graduate of the United States Air Force Academy. He holds a PhD and a Masters degree from Georgetown, and a Masters in Public Policy from the Harvard Kennedy School. The views expressed in this article are the author’s alone, and do not reflect those of the U.S. Air Force, the Department of Defense, or the U.S. Government.
Have a response or an idea for your own article? Follow the logo below, and you too can contribute to The Bridge:
Enjoy what you just read? Please help spread the word to new readers by sharing it on social media.
 Jacqueline Brewer, Michael Miklaucic, and James G. Stavridis,Convergence: Illicit Networks and National Security in the Age of Globalization(National Defense University Press, 2013).
 Vanda Felbab-Brown, Shooting Up: Counterinsurgency and the War on Drugs (Brookings Institution Press, 2009).
 Rick Atkinson, Left of Boom: The Struggle to Defeat Roadside Bombs(Washington Post, 2007).
 Gambetta, The Sicilian Mafia. Mao Tse-Tung and Zedong Mao, On Guerrilla Warfare (University of Illinois Press, 1961).
 Max Weber, “Legitimacy, Politics and the State,” Legitimacy and the State, 1984, 32–62.
 Gambetta, The Sicilian Mafia.
 Charles Tilly et al., War Making and State Making as Organized Crime(Cambridge University Press, 1985),http://static.ow.ly/docs/0%20Tilly%2085_5Xr.pdf.
 Stephen D. Krasner, “Sharing Sovereignty: New Institutions for Collapsed and Failing States,” International Security 29, no. 2 (2004): 85–120.
 Martinez, The Slave Trade and the Origins of International Human Rights Law; Krasner, “Sharing Sovereignty.”
 Martinez, The Slave Trade and the Origins of International Human Rights Law.
 Like any modeling assumption, this fails to represent the world at its extremes. While an increase in regulatory burden is likely to result in diminished compliance, if there is an expectation of no rules, it is unlikely that people will comply with minor rules; if there are too many rules, as in the Soviet Union, liberalization efforts can result in less compliance rather than more. Also, increasing rules can remove groups from the polity, which reduces resources for governance — we address this in the dynamic model. But, in mid-ranges of policy on most issues in a consolidated democracy, this tradeoff holds — even on an issue as sensitive as gun rights or abortion, the vast majority of people will continue contributing to governance by paying taxes and the like even if they do not achieve their preferred policy outcome.
 We assume that the production of governance is not directly affected by this conflict, but in extreme circumstances it can be. If the situation gets so bad that people cease paying taxes, declare independence, and so on, then this can result in a ‘death spiral’ where enforcement attempts reduce governance, which further complicates enforcement.
 Hochschild, Bury the Chains; Drescher, Econocide.
 Okrent, Last Call.
 This is not necessarily a wise strategy — it may be that a lack of enforcement in these areas is exactly what leads to the lack of population support. Local conditions matter greatly in population-centric strategies (Kilcullen 2010.)
[18 Personal Interview, Sgt. Bruce Blair, Montgomery County Police Department (Ret.), US Dept of Justice, National Institute of Justice, National Law Enforcement and Corrections Technology Center.
 Gillo Pontecorvo, “The Battle of Algiers,” 138477, 1965, http://dome-dev.mit.edu/handle/1721.3/45240.
 R. C. Trojanowicz and B. Bucqueroux, Community Policing: A Contemporary Perspective (Anderson Publishing Company Cincinnati, OH, 1990), https://www.ncjrs.gov/App/Publications/abstract.aspx?ID=174124.