Gangs, Slums, Megacities and the Utility of Population-Centric COIN

Editor’s Note: The following interview was originally published by

Small Wars Journal


Small Wars Journal: As we are very close (at time of the interview) to the 20th anniversary of the Battle of Mogadishu, don’t you feel that the kind of threat that we faced back then (non-state warlord gangs that operate across domains in the crime world, insurgency and illicit economy) is proliferating around the world? Looking at what is happening in the slums of megacities – such as Cite Soleil in Haiti, Tivoli Gardens in Kingston, Jamaica – or with the Al Qaeda affiliates, do we have a pattern here?

Vanda Felbab-Brown: I am not sure that I would say that the threat has necessarily proliferated, because it was already there in the 1990s, including at the time of the Mogadishu battle. But perhaps we were not paying so much attention to it. Many of the areas that are experiencing problems today and are ruled by non-state actors have had chronic problems for a long time. Certainly, if we look at the favelas in Rio de Janeiro exactly in the 1992-1994 timeframe when the Somalia crisis exploded, the favelas were fully run by gangs and there were a lot of violent confrontations between them and the state. We see the same pattern also in other parts of Africa or Karachi or Colombia, with gangs often linked to politicians. So it is not necessarily a proliferation, but rather what we witness now is our paying more attention to these unstable areas.

SWJ: Are these trends here to stay with us? Are we heading towards a world that provides more opportunity for these dark networks to proliferate, to incubate? On one side, we see concentrated urbanization in coastal, hyper-connected areas that will need to accommodate more and more waves of rural immigrants. Presumably this will put pressure on an already overstretched city infrastructure. On the other side, we see what Moisés Naím is talking about in his latest book on the decline/decay of power, the decoupling of power from size, and the decoupling of the capacity to use power effectively from the control of a large Weberian bureaucracy.

Felbab-Brown: They will definitely stay and perhaps will intensify, partially for the reason that urbanization is taking place in a way and magnitude with which many governments struggle to cope. Much of Mexico City, a megacity of 20 million people, for example, is really disconnected from the central government and the central business areas. Another example is Karachi: yes, the blood bath that we see there is partially instigated by state actors, but it is also a phenomenon of a very tenuous and limited control of the state in many areas of the city. I am intrigued by the Moisés Naím suggestion of the decoupling of the effective exercise of power from Weberian bureaucracy. I would say that the Weberian construct of power has never been as widespread as we imagine. Many areas in the world – Africa being the prime example – have had a different notion of the state, and one much closer to the medieval conceptualization, where the purpose of the state and the purpose of power competition for controlling the state apparatus is to make money for oneself, as opposed to the public service state whose main raison d’être is to deliver public goods to citizens in exchange for legitimacy and sustainability. In the end the social contract emerged because the self-interested elites understood that they needed to offer something to the population. The notion that the state is weakening and collapsing might not really be appropriate for many parts of the world where the state has always been defined as a mafia bazaar – where one takes over the state to issue exceptions from law enforcement to friends, so that he or she and friends can make money.

In other parts of the world, the state might not have also been quite conceptualized in terms of social contract and Weber-like notions. For example, in parts of Latin America, the state lies somewhere in between the European/Western model and some of the Asian or African states. In Latin America, the state is often captured by and serves a very narrow elite. Many countries in Latin America are still highly exclusionary. Even in places like Colombia, with all of its progress over the past decade, there is still a fundamentally unequal society. Yes, the middle class is increasingly able to participate in state making, along with the political and economic elites, and being able to demand accountability. But still, vast segments of population are really not experiencing the state in any positive manner or even experiencing the state at all. Much of Central America faces massive challenges: tax collection might be as low as 10%, and the political and economic power is enormously concentrated in the hands of narrow elites.

All that said, the future – if troubling trends are not mitigated now – might bring some fundamental challenges. Imagine if in the long term there is substantial global warming, major sea level rise, and large parts of Bangladesh, for example, are submerged. This will generate huge waves of immigrants, but also radically empower non-state actors who are able to provide some sort of protection and public goods to the population.

SWJ: In one of your most recent articles, you talked about the emergence of crime-war battlefields. What causes crime and war/warfare, or crime-terror, or crime-insurgency to overlap? What made traditional criminal group, narco-gangs and cartels get into the business of insurgency, becoming criminal insurgents? What explains the shifting of traditional crime towards insurgency?

Felbab-Brown: First of all, I fundamentally do not believe that overall there is a merger of or a confluence between insurgents, militants, or terrorists and criminals. Yes, many insurgencies exploit illicit economies and appropriate crime for their purposes. And yes, in some parts of the world, they often operate in the same space, interact, and occasionally may force an alliance or even merge. Second, I would also make the argument that even the traditional criminal organizations often take on the trappings of governance, behaving in some ways like insurgents sometimes do. An old example of this phenomenon, the paradigm of the traditional crime world, if you’d like, is the Italian Mafia. They didn’t declare a war against the state, didn’t want to topple the state, but for all practical purposes they provided security, norms, and enforced contracts. They extorted money, which is a form of tax collection; they could control politicians and government officials. In many ways, they really behaved like the dream outcome for an insurgency. The difference is that they didn’t put up a flag and declare an Independent Republic of Mafia, but behaved in many ways like a very successful insurgency. Many of today’s warlords in Afghanistan or Somalia would love to be as successful as the Mafia has been. Of course, various criminal and various militant groups have different capacities and talents for being effective governors. The Taliban in Afghanistan, the Mafia in Sicily, and the gangs in the slums and shantytowns are all in the business of providing socio-economic goods, regardless if they are politically-motivated actors or just criminal enterprises. The more they do so – the more they provide order, security, and economic goods – the more they behave like de facto proto-state governing entities.

SWJ: Clausewitz wrote about a fundamental trinity between the government, the army, and the people. This image used to give us the flavor of a Westphalian world, a traditional state monopoly. Are the criminal insurgents, fundamentally non-state actors, becoming proficient in mastering this trinity?

Felbab-Brown: I’d say perhaps. But just as the states are not uniform in their capacity to govern, so the non-state actors are not uniform in their capacity to govern. Some are just abusive thugs with little political capital, and little legitimacy, who rule merely through brutal force. And of course, there’s a complex relationship between their capacity to govern and the governance of the state. Take for example, Jamaica’s Tivoli Gardens. The growth of the gangs in the garrison towns has historically been very much enabled and exploited by leading politicians and political parties. The politicians used the gangs to deliver money, such as donations, and votes; intimidate rivals; and rule via proxies in areas where the state chose not to or was not able to govern properly. In many ways, the gangs emerged and were nurtured, if not outright created by the state and the political parties. But at some point, the state lost control over the gangs – at least, the level of control it used to have. A similar evolution also took place in Pakistan, with many of the militants groups as well as the urban gangs in Karachi. In Mexico, the narco-groups were not created by the state, but they were coopted and used by the state for decades.

SWJ: How can we explain the presence of ‘the people’ in this non-state Clausewitzian picture? What makes people support criminal insurgents? What creates and preserves the bond between people and criminal insurgents?

Felbab-Brown: It comes down to the state leg of the trinity, or more precisely, its absence or only deficient presence. In areas with minimal state presence, in conditions of social, economic, and political marginalization, people come to be dependent on illicit economies. For many of them, participation in illicit economies may be the only way to satisfy to some extent their human security needs. So the more the state is absent or deficient in providing some of the basic public goods (access to justice, public safety, suppression of street crime), the more likely it is that local communities will become dependent on and supporters of the non-state actors who sponsor these illegal economies. Even in criminal domains, illicit economies, or informal economies there is no such thing as ungoverned space. Even criminals need order.

Chaotic criminal markets are difficult to operate in even for criminals, just as constantly contested control and unpredictable violence is often the most difficult environment for people to bear and adapt to. One of the worst criminal markets to operate in today is Mexico. The criminal underworld is way too violent, way too unpredictable, way too contested. Transaction costs are very high. The hardship that unpredictable, widespread violence and constant contestation over territory imposes on people also explains why the Afghan population in the 1990s found the Taliban, brutal as the group was, preferable to the warlords. The Taliban provided predictability, order, and stable, if brutal, rules.

Similarly, to function as an order-and-rule provider brings criminal entities important support from the community. That support might be thin, but it is often crucial. How thin and how important such political capital is for the criminal group depends to a large extent on who the alternative competitor for the allegiance or the implicit support of the people is, and whether that entity, say the state or another warlord, can better provide a public good, such as security? Ideally, the state out-competes the violent non-state actors and bonds between the state and the people are strengthened. In other words, in many parts of the world, we should think of crime not merely as an aberrant social activity to suppress, but rather as a competition in state-making. In weak and fragile states, where there is socio-political marginalization and poverty, non-state actors are often able to out-compete the state and secure the allegiance of large segments of society. Ultimately, it is about competitive state-building between the government and the non-state actors.

SWJ: At its core a criminal insurgency is still a business-oriented entity. Is it profitable to also be in the business of providing some form of welfare governance, of providing governmental services?

Felbab-Brown: I would say very much so, but this depends on the notion of profits, and different criminal groups make different decisions. It is profitable in the sense of being valuable not just with respect to short-term monetary gains. There might be some immediate monetary costs to groups from providing public goods and services, but doing so brings long-term sustainability and the critical benefit of having a sufficiently good relationship with the population.

Often militant groups or criminal groups are tempted to rule simply through the coercive force, through repression. Mexico’s Zetas or Somalia’s Al-Shabab are prime examples. Al-Shabab became very brutal and predominantly ruled through repression. This is not to say that they provided zero governance, no public goods; they did. Especially in the group’s early phase, Shabab sometimes changed the behavior of clan elites in ways that the population elites quite liked. Nonetheless, compared with other insurgencies, they are very brutal and provide little, if anything, in terms of public services and goods. And they have become progressively more brutal.

The criminal and militant groups which survived for decades, and some of which are still alive and robust today, are hanging on so long or thriving precisely because they have been able to generate enough support among the population.

SWJ: How useful do you see population-centric counterinsurgency (COIN), tailored to each context, in dealing with these challenges? How valuable do you find this intellectual construct in fighting against the gangs in the slums? Winning the trust of the local community and building the legitimacy of the state is at the core of population-centric COIN and is something that is missing in the slums.

Felbab-Brown: I absolutely think so. I am not in favor of arguments that say COIN is a lost proposition and we should get out of doing COIN. Instead, we want to do COIN smartly and selectively and recognize our limits as to when and where we get involved.

I am not an intervention advocate and in fact I want to see U.S. foreign policy become far more selective about getting militarily involved abroad. But population-centric COIN is an approach, a methodology to deal with non-state actors. If we do get involved in situations where we face non-state actors, or for many governments, in their domestic anti-crime settings, what is the alternative? Go and shoot the militants or the criminals up, kill-and-capture-oriented policies? Often, governments will not be able to kill their way out, drone their way out, of dealing with violent non-state actors.

That does not mean that military forces should be sent to deal with crime problems, but it does mean that for addressing many violent slum situations, such as in Karachi or Brazil, simply kill-or-capture policies will not be adequate. Even police forces might need to adopt what de facto approximates population-centric COIN. For example, for decades, the response of the Brazilian state to violence emanating from the favelas was kill and capture. The military police would roll into the favelas, shoot up people for a day, and get out. Other than producing suffering and more violence, such a policy didn’t accomplish anything. Rio de Janeiro’s latest program, of Pacifying Police Units (Unidade de Policía Pacíficadora or UPP), on the other hand, is a very counterinsurgency-like approach without it being called a counterinsurgency. But the logic and basic scheme are there: permanent presence of the state, including its law enforcement apparatus as well as linking the people to the state via the extension of public goods, services and a legal economy – the essence of what population-centric counterinsurgency is about. But there is no denying that implementing population-centric insurgency or similar anti-crime programs is hard, there are few success stories, and often a lot of deeply problematic policies have been mixed into supposed population-centric COIN or anti-organized-crime approaches. With the disappointments in Iraq and Afghanistan, there is a tendency in the United States right now to throw out the COIN baby with the bath water. But to say that the United States would never do COIN again is highly problematic because what is the alternative? That would leave only kill and capture policies or never getting involved in any setting where one has to confront violent non-state actors.

SWJ: What would the “build” phase mean or imply in retaking or pacifying a “lost neighborhood”?

Felbab-Brown: It implies some physical “building,” such as building infrastructure and connecting the slums physically with other parts of the city. Then it implies building, bringing in legal economies, a task that is often very difficult. It might be no less difficult in urban slums than in Afghanistan, for example.

Often the slum area doesn’t have electricity, clean water supply, or regular Internet access. Yet bringing in such physical infrastructure are still the easiest fixes. Other issues that need to be addressed to jump start a legal economy and foster more positive social dynamics in the area include addressing very low human capital, such as by investing in education and health care.

It is also necessary to address the absence of legal dispute resolution mechanisms and access to formal justice and contract enforcement. Historically, such regulation and predictability would be provided by the slumlord.

Frequently, the police would be deeply deficient in providing citizens’ safety; rather, they would frequently be highly abusive and very corrupt. Permanent and benevolent police presence in a slum – truly community-oriented policing – are needed to change the perception of the population toward the state and its role in contract enforcement and delivering public safety.

Frequently by far the most complex task is creating legal job opportunities. Building roads and bridges and extending electricity are far more straightforward propositions than generating legal jobs. Indeed, job creation is not something that frequently the state can do easily. The state can employ some people, such as in cash-for-work programs, but ultimately private businesses need to come to the slum areas and find it cost effective to employ its residents. Such a scenario then goes back to the quality of human capital in area, such as whether the population is sufficiently skilled to offer a competitive labor force. It often takes a lot of time before businesses come in with sustainable enterprises and create jobs. Indeed, the building-up-the-legal-economy and creating new legal jobs is a component with which Rio’s UPP policy, for example, is very much struggling. These efforts take a lot of time and sustained effort.

Implementing such complex socio-economic and security policies to address militancy or entrenched crime is complex, difficult, and poor design and implementation oftentimes kill the efforts.

The overall effort in the end is about extending a multifaceted, benevolent presence of the state, the components of which include law enforcement efforts, economic policies, and fostering access to formal justice and dispute resolution mechanisms.

Such a multifaceted state-building effort involves varied actors across government, really necessitating a whole-of-government approach. At its core should be the goal to strengthen the bonds between the state and the marginalized communities vulnerable to participation in illicit economies for reasons related to economic survival and physical security. This approach requires governments to address all of the causes and structural drivers of the popular support for the illegal activities or militants.

SWJ: How would you assess Rio’s Pacification Policy (UPP) program?

Felbab-Brown: The verdict on the UPP is still out. The policy is clearly much better than previous violent raids. Its emphasis on community policing and permanent police presence and its socio-economic components make for a very good overall policy design. In its conception, the UPP is an exciting program. However, the UPP has struggled in implementation and some aspects of the program cannot be evaluated for years to come. The handover from the raid takeover forces to the community UPP forces has lagged behind in some of the favelas. Reports of police abuse and retaliatory violent targeting of policemen by gangs persist. UPP’s socio-economic programs have often been cast too narrowly. Job creation takes years to implement.

Bringing formal justice mechanisms or formal courts to the favelas has also dragged behind. A big question is whether the will to persist with the efforts will remain after the Olympics Games and the football World Cup – on both the part of the policymakers and Brazil’s taxpayers. But the UPP is a tremendous opportunity: yes, it needs to be tweaked and adjusted in some aspects, but to waste it would be tragic and would sentence Rio’s favelas to crime, poverty, and marginalization for years to come.

SWJ: Should the community-policing models implemented in London or New York be considered or used in the lost neighborhoods, slums or shantytowns of the global megacities?

Felbab-Brown: There is a tremendous variation among what is called community policing. As the concept became popular, a government and/or a local police force would re-label “community policing” whatever they were doing before. New York’s program has not been about community policing at all; it’s been about zero-tolerance approaches, based on theories such as “broken windows” – i.e., targeting aggressively even minor infractions with the hope of deterring more serious crime. NYPD’s aggressive stop-and-frisk component has been highly controversial among minorities and indeed, increasingly, New York residents overall. Such zero-tolerance approaches often tend to be inappropriate models for slum neighborhoods or countries with small, corrupt, inadequate police forces. Indeed, Rio’s UPP program, which prioritizes violence reduction over tackling other criminality, for example, looks nothing like New York’s zero-tolerance approaches that aggressively target visible drug use and distribution. Similarly, exporting New York’s zero-tolerance approaches to Karachi, for example, would accomplish little; likely, the policy would be perverted and counterproductive. For countries and areas with small, overwhelmed police forces and high criminality, other policies, such as community policing, selective targeting, and focused deterrence, tend to be far more effective.