Guest post by Graham Denyer Willis, Centre of Development Studies and Centre of Latin American Studies, University of Cambridge. Graham has a PhD from MIT and has held positions or undertaken research in Cape Verde, Uganda, Zambia, and Brazil, both as an academic and as a policy and NGO worker. He is the author of The Killing Consensus. Graham is on Twitter @gdenyerwillis. This post is the fourth installment of Border Criminologies’ themed week on the Anthropology of Police organised by Paul Mutsaers.  

‘We were so poor,’ a civil police officer in São Paulo once told me, ‘that when someone in the neighbourhood died, my grandmother was the first on their doorstep to ask for their clothes.’  

Scholars have written and researched abandonment in Latin American cities for many decades. Often at the forefront of these discussions are the built environments of abandonment: vast urban territories that seem to exist outside the realm of the state. These ‘slums’―comunas, favelas, villas, periferias, subúrbios, and the like―are spaces where the centralized redistribution of resources, goods, and services has often never meaningfully existed, or has existed with marked difference. This is acutely true of the ways that many urban communities experience policing and the idealized redistribution of security as a supposed public good.

This abandonment is by design. It’s not the case that states and political authority lack capacity to provide for citizens on the edge. Such a position is an illusion. Abandonment is a form of governance: millions in Latin American cities are systemically devalued, left to die. Such a form of governance is profoundly elucidated by ethnography, ideally suited to detail the conflicted contours of police and policing.

Consequentially, some such communities have become consumed by ‘their own’ violence. As cities like Tegucigalpa, Medellin, and Caracas implode in violence (especially) at the margins, let them kill each other is a refrain that reverberates across morality, social relations, policy, and practices of governance. Countless people continue to die unremarkably in great number, through violence, ill-health, and a dulled suffering that remains politically and socially unimportant. To enumerate the deaths of the systemically unvalued is a conflict in terms and practice.

In all of this, where should we locate those who become the region’s vast corpus of low paid police people? These are the empirical terrains of everyday policing―and of individual police―in Latin America. Not only did the civil police officer’s grandmother ask for the clothes of the dead, she got them. Routinely. This police officer explained that he not only grew up in a context of violence, he literally grew up in the clothes of dead people. As one size became too small, the next size up was inevitable.

© Graham Denyer Willis
Often lost from the discussion and research on police in the global south is a sophisticated portrayal of the difficult reality that police themselves are often―if not typically―systemically abandoned individuals. After three years of studying homicide and other detectives of the civil police in the city of São Paulo Brazil, two things stand out, contextually. First, from well before São Paulo’s police became police, violent death was often a mundane part of their everyday existence. I detail at length in my recent ethnography, The Killing Consensus, how such death was part and parcel of the formative years of many police. Some could only become police because they escaped many open-air shootouts, the premature expiration of close friends, and the cloudy disappearances of others, including family members. In the urban periphery of cities like São Paulo, the happenstance of dead bodies lying in the streets―sometimes for days―especially through the 1990s, marked the consciousness of those thousands of disposable, underpaid (yet pensioned!) armed urban bureaucrats that today fill the streets. Who survived, and why, continues to deeply inform the logic and street-level justifications of contemporary policing.

Second, the guardians of the valued are themselves unvalued. In places like Brazil, just as police assassinate without exception, they are assassinated without exception by the hundreds each year. It’s impossible to separate the routine shooting of primarily young black men by police in São Paulo―at a rate of roughly 2.2 per day―from the targeted killings of police officers―106 in 2012, the last year of my field research. In 2014, to contextualize, 114 police were assassinated in the world’s next Olympic city, Rio de Janeiro. If killings of police were tallied like murders, we would see that they die at a rate of 265 per 100,000.

There’s a common response and sentiment about police assassinations in such a context: Who. Cares. Police funerals reflect as much, being confined to distant public cemeteries, rarely with media present and only attended by small numbers of colleagues and family members. As it goes, these are violent people who are probably ‘dirty.’ But any insinuation that dead police officers must have done something to deserve it only serves to further legitimate violence against not just police, but, by extension, anyone who lives in a historically abandoned community.

© Graham Denyer Willis
The point here isn’t to paint a sympathetic picture of policing or to minimize the scale of everyday police violence in São Paulo or elsewhere. It is, rather, to nod at a larger point about governance: the conditions of abandonment and the devalued state in which police and countless others live allows―if not justifies―a continued need for violence in a supposed attempt to order society. But a system cannot leave vast populations to die and then respond meekly to spiralling let them kill each other violence by using similarly abandoned lives to combat deep historical and political inequities. To do so politically and morally legitimates extermination―the current status quo.

The issue is twofold and cyclical: The more police are made to feel vulnerable and under-protected by their own political handlers, the more likely they are to take violence into their own hands. For the homicide detectives I’ve studied who are responsible for investigating and ‘cracking down’ on police lethality, looking the other way when others do take violence into their own hands is the status quo. There is little reason to believe detectives will disagree, except where such violence takes a valued life. This is exceptional.

There are other systemic consequences: Why put faith in a system that consumes itself? If policing in much of Latin America amounts to unvalued lives fighting unvalued lives, with both being consumed unremarkably, one of the only solutions for a person in the midst of it all is to make your own individual solution. One of the most common things I heard from police in São Paulo was that, when the going gets rough―when there is a rash of police killings or a spike in urban violence―the best thing you can do is go home, lock the door, and protect your family. Lest your clothes be worn by someone else.

Any comments about this post? Get in touch with us! Send us an email, or post a comment here or on Facebook. You can also tweet us.


How to cite this blog post (Harvard style):

Willis, G.D. (2016) Policing, Ethnography, and Abandonment in Latin America. Available at: (Accessed [date]).