Insurance technopolitics: Car theft,recovery,and tracking systems in São Paulo

Abstract

In response to high levels of car theft, insurance companies in São Paulo have developed new systems and technologies for tracking and recovering stolen vehicles. These interventions are driven by an insurance rationality that seeks to manage risk and ensure these companies’ profitability. However, this article draws on the notion of technopolitics to argue that the tracking devices and other technologies mobilized in this way also exercise their own agency. They help to mediate and reorganize the power dynamics and relations between diverse actors who operate within São Paulo’s stolen car market and vehicle recovery processes, presenting both challenges and opportunities for each as they pursue their respective aims. The notion of ‘insurance technopolitics’ emphasizes this conjunction between risk governance and the contingent, technologically mediated relationships and conflicts to which it may give rise.

Keywords

Brazil car theft insurance security technology

Introduction: Car theft and insurance in Brazil

Around half a million cars are stolen in Brazil every year, approximately one every minute. An outsized proportion of these thefts occur in the country’s largest metropolis, São Paulo, which is also home to its largest market in auto parts. There are three main destinations for stolen vehicles. First, they may be dismantled and their parts sold to ‘chop shops’ (car dismantling shops).¹ Second, they may be used to commit other crimes or for joyriding and later abandoned. Third, they may be exchanged for drugs or weapons at Brazil’s borders with Paraguay or Bolivia (Feltran, 2018; Raldes, 2013).² Insurance companies consulted during this research estimate that about 60% of the cars stolen in São Paulo end up supplying the city’s chop shops.

Against this background, representatives of the insurance sector have invested resources and lobbied for the creation of legislation to control such activities. Referred to as the Dismantling Law (Lei do Desmonte) and inspired by a law introduced in Argentina,³ this legislation seeks to regulate activities related to the dismantling of cars, as well as creating a system for tracking car parts. The latter is based on tags that indicate to consumers and the authorities the vehicle from which a part originated and its legal status. The legislation aims to reduce car robberies and, with that, cheapen the costs of car insurance (Fromm, 2019). It was first implemented in the state of São Paulo, and, later, in 2014, a similar version was created at the federal level. This piece of legislation, among other examples, demonstrates the capacity of insurance companies to mobilize and exercise influence on all levels of government. They are important actors in the policymaking process within which public security interventions are designed.

As will be outlined in greater detail below, insurance companies have also developed new technologies and procedures to assist them in the recovery of stolen vehicles. In developing such systems, these actors pursue an insurance rationality that seeks to capitalize risk, constituting what key theorists of insurance, drawing on Foucault, understand as a ‘technology of risk’ (Ewald, 1991; Lobo-Guerrero, 2011). The approaches put forward by these theorists offer powerful insights into the ways in which this entrepreneurial form of governance operates and the ways it interacts, often in symbiosis but also in tension, with the sovereign functions of the state (Lobo-Guerrero, 2011). However, they tend to pay little attention to technological devices themselves, largely seeing these as subordinated to, and merely serving, an insurance rationality. By contrast, this article draws on the introduction to this special issue (Müller and Richmond, this issue) and theories of technopolitics (Hecht, 2011) to argue that such devices themselves exercise agency. That is to say, in the context of the present study, that they mediate and reorganize power dynamics and relations between the diverse actors who operate within São Paulo’s stolen car market and vehicle recovery, potentially presenting both challenges and opportunities for each as they pursue their objectives. The notion of ‘insurance technopolitics’ emphasizes this conjunction between risk governance and the contingent, technologically mediated relationships and conflicts that may arise from it.

I address these themes through the presentation of original ethnographic research in which I followed the everyday routines of a tracking team from a large insurance company in São Paulo. This research shows how relationships and conflicts between private companies, public security forces, and criminal groups are entangled with technologies. The different human actors who appear in these accounts either deploy technological solutions to pursue their objectives or otherwise adapt their practices in the response to the use of technologies by others. In both cases, they frequently subvert those technologies’ intended purposes in the process. These dynamics reflect Hecht’s (2011) understanding of technology as a mode, rather than merely a tool, of politics, and support Amicelle et al.’s (2015) claim that the production and appropriation of security devices are integral to governance of (in)security. Applying a technopolitics lens to the operations of insurance companies in the recovery of stolen cars therefore makes it possible to understand the power and impacts of these companies on public security and to map their complex relations with other security actors.

The remainder of this article is divided into five parts. In the next section, I discuss how the rationality of insurance companies operates, its relation to the state, and how the notion of technopolitics complements and extends these Foucauldian analyses. The article’s third section details the technical features of the tracking devices adopted by insurance companies in São Paulo and how they support the profitability of those companies. The fourth section briefly outlines the methodology adopted in the research, drawing on the ‘ethnography of things’ approach. The fifth section then presents ethnographic findings. Here, the discussion is divided into two subsections. The first outlines the processes of and actors involved in car theft and recovery, while the second examines how tracking technologies blur and reshape the relationships between public and private security actors. The conclusion summarizes the key points and offers some final considerations.

From technologies of risk to insurance technopolitics

Drawing on Foucault’s theory of governmentality, influential approaches to insurance conceive it as a ‘technology of risk’ (Defert, 1991; Ericson et al. 2003; Ewald, 1991; Lobo-Guerrero, 2011). François Ewald (1991) draws attention to the two meanings that are simultaneously implicated in the term ‘insurance’. On the one hand, it refers to insurance institutions and companies. On the other, it refers to an abstract technology: ‘as a technology of risk, insurance is first and foremost a schema of rationality, a way of breaking down, rearranging, ordering certain elements of reality’ (Ewald, 1991: 199). Technologies in this Foucauldian sense ‘refer not to tools, machines, or the application of science to industrial production, but refer to methods and procedures for governing human beings’ (Behrent, 2013: 56). As will be discussed below, this article proposes that technologies in the material sense are also crucial for understanding the relationship between insurance companies and the various actors who operate in the stolen car market and car recovery in São Paulo.

Insurance is the practice of compensation and reparation that is conceptualized within a rationality that is based on the calculation of probabilities. It is a justice model that aims not to punish a culprit for the damage they caused, as the judicial system does (Ewald, 1991); rather, the insurance company buys risk and sells a guarantee of protection in the case of an unfortunate event. Insurance displaces the question of punishment to one of economic reparation. As all of this implies, risk is an essential constitutive element of the rationale of insurance. Indeed, Ewald (1991: 199) goes as far as to argue that ‘risk is a neologism of insurance’. At the core of this set of techniques, risk refers to ‘a specific mode of treatment of certain events capable of happening to a group of individuals – or, more exactly, to values or capitals possessed or represented by a collectivity of individuals: that is to say, a population’ (Ewald, 1991: 199). Ewald (1991: 201) also identifies three fundamental characteristics of risk: ‘it is calculable, it is collective, and it is a capital’.

Addressing recent transformations to the management of risk, O’Malley (2009) argues that ‘risk in and of itself’ is not growing exponentially, but that the propensity and the disposition to treat processes according to the logic of risk is. Paying particular attention to this process, part of the literature addresses the issue of policies for risk management and the technization of urban security (Graham, 2017; Grassiani and Müller, 2019; Wilson and Weber, 2008). Urban territories have become fertile ground for the growing market for services and technology in the security realm, part of which is made up of the insurance market, including car insurance companies. This is particularly so in contexts of ‘fragmented sovereignty’ (Davis, 2010), where non-state groups challenge the state’s monopoly of coercive power.

Luis Lobo-Guerrero (2011) offers valuable insights for conceiving the relationship between the ‘entrepreneurial’ form of power exercised by insurance companies and the ‘sovereign function’ of states. He notes that insurance provides an instrument that allows private companies to operate – and individuals to circulate – in environments where sovereign forms of security are not consolidated:

Conducting petroleum-prospecting operations in Colombia might not require ensuring the capacity of the national government to exercise the sovereignty of the state in the totality of the country’s territory. It involves, instead, a programmed risk management strategy that adapts to the shifting circumstances of the region where the company seeks to operate and managing the risks as they arise. (Lobo-Guerrero, 2011: 105)

By extension, although insurance companies may often act in close collaboration with state agencies and actors, they do not seek to extend the reach of the state’s sovereign power. Instead, they pursue their own entrepreneurial objectives based on risk governance. Where necessary, they may deploy their own security personnel, technologies, and procedures to this end, thus duplicating some (though not all) of the state’s sovereignty functions. In this way, insurance companies represent an important example of how, under contemporary liberal regimes of risk, security governance expands beyond the state, contributing to conditions of fragmented sovereignty (Ericson et al., 2003).

This analysis may also be extended to relationships with criminal actors. Insurance rationality does not seek to reduce crime, but simply to reduce the material damage it causes to the insured. As one participant of this research and founder of an insurance company’s vehicle tracking center aptly put it to me during a personal conversation, ‘the car must be stolen (first), then we all start to work’. Indeed, insurance rationality leaves ample room for crime to occur, and may even encourage acquiescence, whether direct or indirect, with criminal activity. In this respect, it is not different from the rationalities of other private businesses that may be able to benefit from cultivating relationships with criminal actors that can furnish valuable information or influence. Garmany and Galdeano (2017: 1), for example, argue that São Paulo’s ever-growing number of private security agencies operate in ‘complicated networks between the state, private capital and organized crime’. The ‘security’ they offer often constitutes a form of brokerage that may be associated with increasing, rather than reduced, levels of crime and corruption in the city. In this sense, to understand security governance in a city like São Paulo, a less state-centric approach is needed, which considers the presence of actors – including the legal insurance market, private security providers, and organized crime – that coexist with the normative regime of the state (Feltran, 2020).

If the notion of insurance rationality helps to analyze the behavior of insurance companies and their interactions with both state and criminal actors, the concept of technopolitics (Müller and Richmond, this issue) offers further insights. Insurance rationality not only drives insurance companies to develop risk calculations, risk management strategies, and relationships with diverse security actors, but also leads them to develop and deploy material technologies. As will be discussed below, central among these are tracking devices that assist in the recovery of stolen cars. These represent a good example of technopolitics as described by Hecht (2011: 3): ‘a concept that captures the hybrid forms of power embedded in technological artifacts, systems, and practices’. However, the technical properties of these technologies not only further the interests of insurance companies, but also mediate their relationships with other actors involved in the stolen car market and vehicle recovery, presenting these actors with challenges but also opportunities. In this way, it makes sense to expand the notion of insurance rationality to one of ‘insurance technopolitics’. While a ‘technology of risk’ underpins the actions of insurance companies, a technopolitical perspective offers a more materialized, decentralized, and contingent (Müller and Richmond, this issue) understanding of how they interact with other security actors than is implied by dominant Foucauldian approaches.

Tracking vehicles and people in São Paulo

In Brazil, and especially São Paulo, car theft is a major concern for insurers. They have invested significant resources and developed new technologies for tracking and monitoring their clients and their clients’ property. When a car is stolen, the insurance company becomes responsible either for paying indemnity or for recuperating and returning the vehicle to the insured, if it is still in perfect condition. The insurance company only offers financial compensation: what is insured has no relation to any injury suffered by the client; insurance is merely a guarantee for capital loss. In the case of insurance for car theft, any violence suffered during theft, or possible setbacks experienced as a result of it, are not compensated by the insurance company. There is also no guarantee that the vehicle itself will be found and returned to the insured. The conditions prior to the incident will never be re-established. However, the insurance company guarantees compensation for financial loss that results from these events.

It is estimated that, in Brazil, eight out every ten stolen cars have insurance coverage.⁴ One of the main insurers in the country alone has an average of 100 car thefts or robberies per day. More than 12,000 cars from the three insurance companies that make up this insurance group totally disappear annually. The loss of these assets represents a great economic loss for insurers, which is why they seek out these stolen cars. The average rate of recovery for these cars is 50%, but may increase to 90% if the cars have a tracking device installed by the company. Claiming police inefficiency in addressing these criminal cases, insurers have invested in their own autonomous structures for investigating and retrieving vehicles – both for cars with tracking devices and for those without. In this way, the companies try to establish independence from the police, becoming autonomous and specialized in conducting operations that that the police would previously have led. At the same time, they hire service providers that have specific know-how and connections to the police, typically small companies run by former police or military officers, off-duty police officers, or private providers with personal contacts in the police force. Numerous companies have become specialized in retrieving cars, though this phenomenon remains underexplored in the literature. This is a critical object of study for understanding the new political and economic arrangements of the security market in Brazil.

A number of technologies are implemented by insurers seeking greater control and better calculation of risk. Vehicle-tracking technologies are a good example. There are three types of technology available in Brazil for tracking vehicles: through satellite (GPS), radiofrequency, and GPRS/GSM. The insurance company’s team that I accompanied opted to use radiofrequency to retrieve its insured cars. The radiofrequency system functions by sending out radio signals and requires a functioning phone chip from a telephone company within the tracking device (in this case, the telephone company and the insurance company were parts of the same group of companies). As they explained to me, the signals from this system are more effective than others for tracking vehicles in tunnels or underground. In addition, unlike satellite tracking, radio signals do not suffer interference from signal inhibitors. According to the team, it is a system with less signal oscillation. Cars are monitored by the central team using Google Street View.

In addition to the significant sums of money accrued on a daily basis from the retrieval of vehicles by the COMTV (Central Office for the Monitoring and Tracking of Vehicles) team, which covers some of the insurance’s company’s expenses with indemnities, the tracking device enables even greater profits. The information collected regarding insured clients’ daily movements is used in various other sectors of the company, for purposes beyond automobile insurance itself. This information, when combined with advertising and publicity techniques, for example, makes it possible to develop campaigns and products directed at clients with specific profiles, selected on the basis of categories such as income, class, gender, age, the neighborhood in which they live or work, and so on. On the basis of the most common movements of a specific group of clients, it is possible, for example, to choose the best location in which a company should place billboards advertising a given product targeted at a particular client profile. Information from the monitoring system can also be used to develop the concept and design of products themselves. These products are increasingly targeted at, and efficient for, a specific and predetermined group of people. Meyers and Van Hoyweghen (2018, 2020) have shown a growth in the ‘behavior-based personalization’ of products and prices in the insurance market.⁵ Furthermore, by learning about their clients in increasingly detailed ways, insurance companies can also gather more information about the risks involved in a client’s habits and behaviors and they can optimize gains and select profiles that most interest them: in the company where I conducted fieldwork, they focused on clients who paid more and posed a lower risk (class A).

The tracking device has also become an important technology for combatting insurance frauds. COMTV has employees who are specialized in investigating frauds. At the moment that a claim is filed, if there is any suspicion of fraud, particularly in cases involving high-value vehicles, an investigation process begins within the insurance company. The client’s capivara – all of their personal data – is pulled up. In other words, the company can check if the person is in debt, any companies or real estate they may own, and whether they are subject to any judicial processes. Part of the investigation process involves tracing the vehicle’s movements over the last few months. During such an investigation, it is common to discover that the client went by various chop shops (car dismantling shops) on the days or weeks preceding the reporting of the claim. The staff already know about a large portion of these establishments (clandestine chop shops) and can see their storefronts using Google Street View. In these cases, a report is written up with all the suspicions and evidence collected. In most cases, this information cannot be used judicially against a client. It is most common for the company to pay the indemnity but to place the client’s name on a ‘blacklist’. Then, the next time the person tries to buy insurance from the same insurance company, the policy amount will be exorbitant or insurance will be denied.

Tracking devices installed in vehicles make it possible for the insurer to monitor the vehicles and behavior of their clients, as well as recovering the insurance companies’ assets. As will become clear in the next section, while they assist the insurers to govern urban security for their own purposes, at the street level these technologies escape these original purposes and produce new subjects and challenges to state governance.

Methodology and context

The empirical data presented in this article are drawn from fieldwork conducted between 2017 and 2020. I conducted participant observation using a multi-sited approach (Marcus, 1995; Hannerz, 2003), during different events associated with car recovery and dismantling, and in different organizations and companies operating within the Brazilian insurance market. This included visits made between April and June 2018 to the COMTV of a major insurance company in São Paulo. On the basis of the methodological strategies proposed by the ‘ethnography of things’ approach (Appadurai, 1988; Knowles, 2014), I sought to follow the journey of a car from the moment it was stolen until it was recovered by the insurance company. The journey of a stolen car was developed fictitiously on the basis of data and typical cases documented during research in the field.⁶

I do not describe a single individual Hyundai HB20, but rather outline a feasible combination of real elements and subjects to reconstruct a typical trajectory.⁷ It was only through the use of this methodological strategy, structured around the need ‘to follow a car’ (Feltran, 2022; Knowles, 2014; Tsing, 2015), that it was possible to connect the different points of the car-theft market chain. Also, this approach is well suited to a theoretical framework focused on the relationship between material objects and technologies and human beings. The methodological choice is ideally suited to the conceptual framework adopted here and the notion of ‘insurance technopolitics’. The ‘ethnography of things’ directs our focus to the material objects bound up in the insurance market and the ways in which they can be disassembled, reassembled, and repurposed by different actors. The approach adopted here includes tracking technologies and shows how they help to further the insurance rationality by which insurance companies operate, but can also provide benefits for other actors working according to distinct logics. However, as will become clear, the ‘ethnography of things’ approach is also sensitive to place and place-based networks, as it highlights how technologies may be used differently in different socio-spatial contexts.

These insights on security technologies and unequal specialities are particularly relevant in a city like São Paulo, the largest city in South America and one of the largest in the world, with a metropolitan population of around 21 million, and the continent’s most important financial and business center. Despite its wealth, São Paulo contains huge disparities that tend to follow patterns of spatial segregation between a wealthy center and urban peripheries with high levels of poverty and exclusion. São Paulo also hosts Brazil’s most lucrative and globalized criminal enterprises and is the birthplace of Brazil’s most powerful criminal faction, the First Command of the Capital (PCC). There is already a consolidated bibliography on the political importance of this criminal group in the regulation of order in the peripheries and prisons across the state of São Paulo (Biondi, 2016; Feltran, 2018, 2019, 2020; Willis, 2015).

Illegal markets provide opportunities of social mobility for many living in the city’s peripheries. However, the criminal world is also unevenly stratified, with low-level criminal actors earning relatively low incomes. Recent research has shown that illegalisms are interconnected with the formal and licit economies, and, contrary to what one might think, do not harm their ‘development’ or ‘progress’ (see, especially, Feltran, 2019). According to Feltran, cash economies allow ‘dirty money’ to circulate between the illegal and legal economies. Low-level participants in illicit markets spend the money they earn selling drugs, weapons, or stolen cars in fast-food restaurants or buying products from major global brands in shopping centers. In monetary terms, the ‘criminal world’, at least in São Paulo, is not opposed to or separate from the ‘official world’.

Before we start following the stolen Hyundai HB20, I would like to emphasize that the car-theft market is a good example of the many ways in which legal and illegal markets are interlocked. Not only does the risk of theft increase demand for insurance, insurers are also large buyers of parts to repair damaged cars and are the sellers in used-car auctions. They also put pressure on their franchised garages to repair their cars at the lowest possible price. Usually, those garages buy untracked and informal auto parts at illegal chop shops.

The Hyundai HB20 model was chosen for this research because it is a popular car, widely sold – and frequently stolen – in Brazil. The HB20 model was developed by Hyundai specifically for the Brazilian market, with 100% of the vehicle produced domestically, in one of the manufacturer’s factories located in São Paulo. The lack of HB20 parts in the legal industry is a factor that increases the demand for auto parts in the illegal stolen car market. In this context, security governance cannot be thought of as separate from the regulation of legal and illegal markets.

Insurance technopolitics at the street level

Retrieving a stolen Hyundai HB20

William⁸ appears to be around 17 or 18 years old. He has brown skin, the hair on the sides of his head is shaved, and he is wearing a black sweatshirt and jean pants. By means of a YouTube video, he teaches all the necessary procedures for stealing a Hyundai HB20 1.0 2014. First, he opens the car’s hood and exchanges the module for one he has at hand. This piece is considered the ‘brain’ of the engine control unit and is responsible for a series of functions needed to control the vehicle. In order to start the car, a key that is appropriately coded to fit the newly annexed module is used. In order to open the car, a chave micha (bump key) is used. William explains that, right after exchanging the module, he must cut off the car horn’s wire, ‘because when I start the car in there, the car will honk and flash its lights three times’. William explains that it takes an average of 30 seconds to execute all of the procedures and start the car. It is also common to use a chupa cabra (car jammer), a tool that inhibits the tracking device’s signal. This example shows clearly how criminal actors develop their own tactics, including sophisticated counter-technologies, to take advantage of any possible weaknesses in security systems. William will receive between R$300 (US$75)⁹ and R$500 (US$125) for the theft of a Hyundai HB20 from a car chop shop.

João is 35 years old and married. He lives in the city of São Paulo and his Hyundai HB20 1.0 2014 was stolen after he parked his car near a busy subway station in the north of the city. Upon returning to the location and realizing what had taken place, his first step was not to contact the police, but rather to contact his insurance company – one of the leading companies in the market. His main objective was to recover the economic losses he suffered as a result of the seizure of his goods. The police and the state could not guarantee this, but the insurance company could. The attendant informed João that an insurance claim had been filed, but that João still needed to contact the police immediately to file a police report, for this document would ensure that he would receive an indemnity. This would also ease the process of searching for the vehicle, for it would be reported and identified as stolen in both the government’s and the insurance company’s surveillance systems and apps. On the basis of the police report, cameras from the Detecta¹⁰ system would identify, from the car’s license plate, that the vehicle circulating throughout the city was a stolen one. Through the use of this system, it is possible to generate a dossier with images of the car captured by the cameras, identifying the places through which it has circulated.

In accordance with internal procedures, the insurance company quickly directed the claim regarding João’s HB20 to its Central Office for Monitoring and Tracking Vehicles (COMTV) by email. This section of the company, located in the inner city of São Paulo, is responsible for searching and retrieving cars insured by the company and reported as stolen or robbed. The team has two subgroups that work together and are responsible for retrieving vehicles that have a tracking device installed or not.

Theft of a Hyundai HB20 is considered a typical case for the team. In recent years, this has become one of the most commonly sold cars in Brazil and is also among the most frequently stolen in the country. One of the staff members on the team argues that this is due to a lack of car parts in stock at car dealerships. This is one of the many connections between the illegal car-theft market and legal and formal markets. Given the lack of car pieces in the legal market, most of these stolen vehicles are destined for disassembly. This car model is considered to be visado (targeted), meaning a common object for theft. This fact results in it having a higher rate for its insurance policy – for João’s profile (a man who is 35 years old and has a garage at home and at work) the price of insurance rages from R$2,500 to R$3,000 (US$625 to US$750) in the state of São Paulo. In the first half of 2018, there were about 196,253 Hyundai HB20 vehicles insured in Brazil.¹¹ Approximately 24% of these cars were located in the metropolitan region of São Paulo,¹² which is also where most of the insurance company’s claims for robbery and theft originate.

Data on João’s HB20 (license plate, chassis number, and the street in which the car theft occurred) were shared by the staff on their WhatsApp group so that the motorcyclists on the streets could begin their search. This is standard procedure, and the app is one of the main channels of communication between the team members. At the main headquarters, the staff is made up of approximately 20 technicians, who sit at their individual computers in a large room with room dividers. On the street, motorcyclists carry out the searches. These include 35 pronta-respostas (ready-responders), who search for cars with a tracking signal, and 25 caçadores (hunters) – more experienced operatives who are responsible for retrieving cars without a tracking device. As this distinction implies, the pronta-respostas are more heavily constrained by the technical system developed by the insurance company to assist car recovery, whereas the caçadores have greater room for discretion. In any case, these motorcyclists are usually paired together and responsible for specific areas of the city.

Ricardo has been a pronta-resposta on the team for about a year. His former job was as a motoboy (motorcycle courier). He claims to prefer his current job, since he considers it to be calmer and says he receives more money. On average, he earns around R$3,800 (US$950) per month. He provides services to the insurance company through temporary contracts that are renewed every three months. The insurance company does not formally employ him; therefore, he does not have access to employee rights and benefits. As soon as he is called, he receives R$26 (US$6.5) to cover fuel costs. He claims to drive about 70km per day, which amounts to a cost of R$450 (US$113) a month. When he retrieves a car, he receives R$179 (US$45) for a car with a tracking device and R$260 (US$65) for cars without the tracking device signal. His goal is to always retrieve 15 cars per month. He conducts searches in the city’s East Zone, where he was born and lives to this day, and which is known for having high rates of car theft.

His routine is to wake up early in order to leave his house by 6 a.m., ‘before they [his neighbors, who steal cars] wake up’, and drive around avenues and alleyways where he knows cars are usually left to descansar (rest). After stealing a car, it is common for the thieves to leave it parked for a few hours or days, usually near avenues, parking lots, or deserted streets, in order to be sure that the vehicle does not have a tracking device and that someone will not come looking for it. They refer to this as the period in which the car is ‘resting’. A few characteristics will draw Ricardo’s attention to possible cases and lead him to check the car’s license plate on the SINESP Cidadão (an app run by the federal government) to see if it is cantando (singing) – that is, confirmed as a stolen car. These include: cars that are beyond the neighborhood’s economic profile; broken car panels (people usually leave a garbage bag hanging beside the steering wheel in order to hide that the car panel is broken, which is a sign for the retrievers); blown airbags; and cars that are dirty, poorly parked, or lack wheels or tires. After conducting his morning search, he returns home to have lunch and, in the afternoon, only leaves the house if he is summoned by the central office to find a specific car that the tracking device identifies as being in his region.

Ricardo describes how some of his neighbors, with whom he grew up and who are now involved in crime, have come round to his house to question him about cars he retrieved. According to his account, he told these acquaintances that the car had a functioning tracking device and that the company would retrieve the car and call the police anyway. However, in reality, Ricardo was able to find the vehicle because he already knows the locations used to disassemble cars or where cars are left to rest. In other words, Ricardo was able to act tactically by drawing on his situated knowledge of both the insurance company’s system and the local criminal world. He used the tracking device as an ‘alibi’ to the thieves, presenting himself as merely a neutral tool of the technology rather than a human agent, with significant scope for discretion, who is pursuing his own objectives. On the other hand, his awareness of the threat the thieves pose to him means he can also decide when the risk of retrieving a car may be too great and calibrate his activities to achieve his recovery targets without placing himself in danger.

Ricardo’s role also carries potential benefits for the thieves, as it is best for them to avoid any encounters and possible conflict with the police. The fact that Ricardo helps retrieve a car avoids searches and police operations within a given territory. With Ricardo as the main agent of car recovery in the territory, the main risk they face is losing a recently acquired asset, rather than prison or death. On the other hand, the fact that his colleagues show up at his doorstep reveals tensions. There is a distrust that someone from the ‘community’ (in this case Ricardo) is acting against his peers’ interests and precepts.¹³ In other words, there is a tenuous line between Ricardo being considered an employee who acts correctly and, as a result of his work, helps to avoid conflicts with the police, or a traitor who helps the police or insurance companies find stolen cars and therefore hinders the business activities of his neighbors. In this case, the (imaginary) presence of the tracking device protects Ricardo from being treated with even greater suspicion by his neighbors.

Similarly, during field research, other cases of tensions and conflict between the company’s motorcyclists and agents from the world of crime occurred during searches taking place in peripheral neighborhoods. Cases of employees being chased down, or of motorcycle robberies, occurred frequently. This is the reason why all motorcycles also have tracking devices and are monitored by the central office. It is up to the motorcyclists whether or not to purchase the company’s insurance. Ricardo opted to insure his motorcycle and explains that, as soon as he was able to save some money, he bought another motorcycle to use during the weekends and periods when he was not working. This is due to the fact that he does not want the company monitoring where he goes during his leisure time. Again, he was able to use his own knowledge of how the tracking devices operate to escape their reach, potentially allowing him to carry out journeys that his employers might regard as ‘suspicious’.

Public–private (in)security governance

Ricardo claims that the company usually holds events in which those who retrieve the most cars receive prizes, such as iPhones and laptops. According to him, the motorcyclists who retrieve the most cars are those with privileged access to information from contacts or informants in the police. For example, he narrated to me the case of three cars stolen on an avenue near the airport. First, two adolescents robbed a woman who was sitting alone in the car, taking her Toyota Corolla. The Corolla was used to commit two further robberies, in which a Ford Ecosport and a Chevrolet Zafira were taken. Coincidentally, all three cars were insured by the same insurance company. The information about the robberies was provided by a caçador, before the company knew of the fact. In other words, the police had been informed through an emergency call to 190 (Brazil’s police emergency line), and the information was passed along by the police directly to the caçador before the insurance company itself had been notified of the robberies. As this suggests, individuals operating at the street level of this market traverse the lines that separate public and private security. Both caçadores and frontline police, perhaps acting together, can exploit their knowledge of the procedures and technologies of car recovery to benefit from its incentive systems.

It is a common practice in the sector that insurance companies pay police officers a reward for recovering vehicles. In the insurance company where this research was conducted, employees explained that it was common for police officers to line up in front of the company’s building to receive their rewards. They were seen in uniform and with their police cars. However, more recently, the head of the car-tracking team claimed they avoided contact with the police as much as possible, because he believed this might bring problems to the company, particularly with the implementation of a new anti-corruption law.¹⁴ For example, they no longer employ former police officers. Nevertheless, this does not stop caçadores or pronta-respostas from mediating with police networks, which are fundamental for retrieving vehicles. In this case, the company can claim it does not know of the illegal or informal arrangements established by the motorcyclists, who are not formal employees but autonomous workers.

Moreover, small companies with inactive or retired police officers permeate the field of vehicle retrieval, and the major clientele for their services are insurance companies. These private security companies also provide services to enterprises that rely on cars or to citizens who do not have insurance and want to find their stolen goods. The insurance company where I conducted fieldwork, for example, has its own team of motorcyclists with a COMTV solely for the city of São Paulo. For the rest of the country, this central office also monitors the claims but calls upon one of these partnering companies to carry out the searches. Networks that include both former and currently serving police officers control this stolen-vehicle recovery service, as well as other private security services in Brazil.

According to a former civil police officer who has worked in the field of vehicle retrieval for over 30 years in many parts of the country, Rio de Janeiro is a prime example – his company pays active police officers to retrieve stolen vehicles that are parked or being used in favelas. In an informal conversation over lunch, he explains that the amount that is paid varies according to how dangerous the territory is deemed to be. In other words, areas where criminal actors have a stronger presence, and where the police have limited access, are more expensive. The dynamic for car retrieval in Rio de Janeiro is different from that of São Paulo. The monitoring center often knows the location where the car is circulating, since the tracking device is functioning and sending out a signal. However, since the cars are circulating through streets located in the favelas of Rio de Janeiro, the caçadores and pronta-respostas do not enter these areas. According to the staff, these cars will usually be retrieved during police operations. It is common that, when the police enter the favelas, representatives from the insurance company accompany the police to retrieve stolen cars. In fact, the companies require the help of the police to retrieve such vehicles. Though further research would be needed, this suggests that in Rio de Janeiro, more than in São Paulo, the risk technology of insurance is more dependent on sovereign power and unable to operate as effectively in its absence (Lobo-Guerrero, 2011).¹⁵

João’s car did not have a tracking device, but his car was retrieved by the team¹⁶ in the city of São Paulo. Cars that have been stolen, as in João’s case, are considered harder to retrieve than those that have been taken in a robbery, since the client often takes a while to realize what has taken place and to call the police or the insurance company. In the case of robberies, because of the presence of the client and the violence they have suffered, the police and the insurance company tend to be informed immediately. This is why technology that connects the tracking device to the module is being tested on a small sample of insured cars. At the moment of the theft, when the car module is withdrawn, the central office is informed by the signal from the tracking device. At that moment, the team calls the client and checks whether the client is with their car. The idea is to give the team autonomy from the client’s call to act and thereby accelerate the search process. This technology will also help fight cases of fraud in which insured persons themselves take their cars to be disassembled.

Final considerations

This article has addressed the role of the insurance industry in the market for the protection of assets, in this case motor vehicles, in Brazil. This is a market that includes diverse private, public, and criminal actors all seeking to accumulate capital and/or exercise coercive power. It is thus characteristic of broader conditions of fragmented sovereignty observed in Latin America and other regions of the Global South (Davis, 2010). In security studies on Latin America, state sovereignty is often counterposed to – and understood as primarily challenged by – the activities of illegal armed groups. In the case presented here, professional car thieves do indeed take advantage of the normative authority of organized crime in the city’s peripheries to develop effective strategies for operating in the stolen car market. However, the case also shows that other key actors contribute to the fragmentation of sovereignty. Insurance companies, operating their own ‘risk technologies’, duplicate some sovereign functions traditionally associated with states, while subordinating these to their own entrepreneurial objectives. Meanwhile, frontline agents of the insurance companies, of the police, and of private security agencies also exploit their positions to benefit from the opportunities presented by the stolen car market and the incentive systems established for the recovery of stolen vehicles.

The insurance companies have invested significantly to address the problem of car theft, so as to minimize financial losses to themselves and their insured customers. The police force has made it clear to insurance companies that it does not have the capacity to track and hunt down stolen vehicles, and instead claims to prioritize tackling violent crime. (However, as we have seen, individual police officers do continue to make use of the structure and networks created by the public apparatus when they act as private agents in retrieving stolen vehicles.) The expanding role of insurance companies in vehicle recovery clearly alters the power balance between public and private institutions. Such activities, along with lobbying initiatives and more direct forms of collaboration with public security agencies, make evident the ways in which insurance companies increasingly ‘pose a challenge to national-state sovereignty’ (Davis, 2010: 402).

The insurance companies’ activities include the development of new technologies and the hiring of personnel to operate them. Car tracking devices, if they are to be effective, must be assembled with other technologies, such as apps (including WhatsApp, Google Maps, and SINESP Cidadão, as well as other secondary apps), SIM cards, Google Street View, and telemetrics, as well as well-trained teams who can make use of them. These technologies have been developed as a tool of insurance rationality, helping to monitor movement of vehicles between geographical locations to increase the likelihood of an insured asset being recovered. They also allow the insurance company to understand their clients’ patterns of circulation throughout the city and to identify behaviors and risky habits. This can reduce the number of claims and prevent frauds, thus allowing the insurance company to lower its costs and reward more ‘responsible’ insured members, who present a lower risk, while charging more to those considered less responsible and as representing a greater risk.

However, the analysis has shown that these systems can only go so far in ensuring successful recovery. At the limits of their capabilities, it is the motorcyclists’ knowledge of particular neighborhoods and their criminal dynamics and police informants’ access to timely information that are crucial in determining the success of the searches. Although they may be incentivized to do so, these actors do not operate straightforwardly according to insurance rationality. They also take into account informal relationships spanning public/private and legal/illegal divides; power dynamics within state institutions, legal businesses, and criminal organizations; and the scope for autonomy and impunity that the latter afford frontline agents.

A key contribution of this article is to show that material technologies are not inert participants in these relationships and struggles but actively mediate and (re)shape them. In addition to the diverse human actors, on the journey of the stolen Hyundai HB20, we also encountered various non-human ‘things’, such as car security systems, tracking devices, apps, jammers, and cameras. As the analysis has shown, at the level of everyday practices, actors pursuing different objectives and embedded in different relationships must interact with these different technologies, avoiding the risk they represent and taking advantage of their opportunities. This ‘insurance technopolitics’ constitutes a far less ordered and far more contingent world – permeated by many more actors capable of exerting meaningful agency – than dominant understandings of insurance rationality would acknowledge.

Footnotes

Acknowledgements

I would like to acknowledge the valuable comments of the blinded reviewers and, especially, all the insightful suggestions made by the editors of the special issue, Matthew Richmond and Frank Müller. This article is also the fruit of discussions within and ideas coming from the collective project ‘Global Cars: A Transnational Urban Research on Vehicle Informal Economies (Europe, Africa and South America)’, to whose team members I am very grateful. First drafts of this article also received comments from Gabriel Feltran, Erik Bähre, Irene Moretti, Tim van de Meerendonk, Nikki Mulder and Nikkie Buskermolen, and from participants of the 3rd Annual LINCS Conference held in Queen’s University of Belfast in May 2019. The responsibility for any possible mistakes is mine.

Funding

The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was funded by the Sao Paulo Research Foundation (FAPESP), process numbers 2017/24649-7 and 2020/07160-7.

ORCID iD

Deborah Fromm

Notes

Deborah Fromm is a PhD candidate in social anthropology at the University of Campinas and a researcher within the Centre for Urban Ethnography at the Brazilian Centre for Analysis and Planning (CEBRAP). Deborah studies the expansion of the insurance industry in Brazil and investigates the effects of privatization of public security and social protection on urban inequalities from an ethnographic perspective.

References

Altenhain

(2023) Networked security in the colonial present: Mapping infrastructures of digital surveillance and control in São Paulo. Security Dialogue 54(1): 21–38.

Amicelle

Aradau

Jeandesboz

(2015) Questioning security devices: Performativity, resistance, politics. Security Dialogue 46(4): 293–306.

Appadurai

(ed.) (1988) The Social Life of Things: Commodities in Cultural Perspective. Cambridge: Cambridge University Press.

Behrent

(2013) Foucault and technology. History and Technology 29(1): 54–104.

Biondi

(2016) Sharing This Walk: An Ethnography of Prison Life and the PCC in Brazil. Chapel Hill, NC: University of North Carolina Press.

Chan

CSC

(2009) Invigorating the content in social embeddedness: An ethnography of life insurance transactions in China. American Journal of Sociology 115(3): 712–754.

Davis

(2010) Irregular armed forces, shifting patterns of commitment, and fragmented sovereignty in the developing world. Theory and Society 39(3/4): 397–413.

Defert

(1991) ‘Popular life’ and insurance technology. In: Burchell

Gordon

Miller

(eds) The Foucault Effect: Studies in Governmentality. Chicago, IL: University of Chicago Press, 211–233.

Dewey

(2012) Illegal police protection and the market for stolen vehicles in Buenos Aires. Journal of Latin American Studies 44(4): 679–702.

10.

Ericson

Doyle

Barry

(2003) Insurance as Governance. Toronto: University of Toronto Press.

11.

Ewald

(1991) Insurance and risk. In: Burchell

Gordon

Miller

(eds) The Foucault Effect: Studies in Governmentality. Chicago, IL: University of Chicago Press, 197–210.

12.

Feltran

(2018) Irmãos: Uma história do PCC [Brothers: A History of the PCC]. São Paulo: Companhia da Letras.

13.

Feltran

(2019) (Il)licit economies in Brazil: An ethnographic perspective. Journal of Illicit Economies and Development 1(2): 145–154.

14.

Feltran

(2020) The Entangled City: Crime as Urban Fabric in São Paulo. Manchester: Manchester University Press.

15.

Feltran

(ed.) (2022) Stolen Cars: A Journey Through São Paulo’s Urban Conflict. London: John Wiley & Sons.

16.

Fromm

(2019) Creating (il)legal markets: An ethnography of the insurance market in Brazil. Journal of Illicit Economies and Development 1(2): 155–163.

17.

Garmany

Galdeano

(2018) Crime, insecurity and corruption: Considering the growth of urban private security. Urban Studies 55(5): 1111–1120.

18.

Graham

(2017) Cidades sitiadas: O novo urbanismo military [Cities Under Siege: The New Military Urbanism]. Trans. Azuma

. São Paulo: Boitempo Editorial.

19.

Grassiani

Müller

(2019) Brazil–Israel relations and the marketing of urban security expertise. Latin American Perspectives 56(3): 114–130.

20.

Hannerz

(2003) Being there . . . and there . . . and there! Reflections on multi-site ethnography. Ethnography 4(2): 201–216.

21.

Hirata

Grillo

(2017) Comparative perspectives on the retail drug markets in São Paulo and Rio de Janeiro. Tempo Social 29(2): 75–98.

22.

Knowles

(2014) Flip-flop: A Journey Through Globalisation’s Backroads. London: Pluto.

23.

Lobo-Guerrero

(2011) Insuring Security: Biopolitics, Security and Risk. Abingdon: Routledge.

24.

Marcus

(1995) Ethnography in/of the world system: The emergence of multi-sited ethnography. Annual Review of Anthropology 24: 95–117.

25.

Meyers

Van Hoyweghen

(2018) Enacting actuarial fairness in insurance: From fair discrimination to behaviour-based fairness. Science as Culture 27(4): 413–438.

26.

Meyers

Van Hoyweghen

(2020) ‘Happy failures’: Experimentation with behaviour-based personalisation in car insurance. Big Data & Society 7(1): 1–14.

27.

Müller

Richmond

(2023) The technopolitics of security: Agency, temporality, sovereignty. Security Dialogue 54(1): 3–20.

28.

O’Malley

(2009) Governing risks. Legal Studies Research Paper no. 09/86. Sydney: The University of Sydney.

29.

Raldes

(2013) Bolívia devolverá 1500 carros ao Brasil e outros países [Bolivia will return 1500 cars to Brazil and other countries]. Exame, 11 February. Available at: https://exame.abril.com.br/mundo/bolivia-devolvera-1-500-carros-ao-brasil-e-outros-paises/ (accessed 3 October 2022).

30.

Strathern

(2018) O efeito etnográfico e outros ensaios [The Ethnographic Effect and Other Essays]. São Paulo: Ubu Editora LTDA-ME.

31.

Tsing

(2015) The Mushroom at the End of the World: On the Possibility of Life in Capitalist Ruins. Princeton, NJ: Princeton University Press.

32.

Van Hoyweghen

Rebert

(2012) Your genes in insurance: From genetic discrimination to genomic solidarity. Personalized Medicine 9(8): 871–877.

33.

Willis

(2015) The Killing Consensus: Police, Organized Crime, and the Regulation of Life and Death in Urban Brazil. Oakland, CA: University of California Press.

34.

Willis

(2017) City of clones: Facsimiles and governance in São Paulo, Brazil. Current Sociology 65(2): 235–247.

35.

Wilson

Weber

(2008) Surveillance, risk and preemption on the Australian border. Surveillance & Society 5(2): 124–141.