Carjacking: A Comparison Between Campinas,Brazil and Detroit,Michigan

Abstract

Carjacking calls for a specialized analysis because it occurs at lightning speed within very brief windows of opportunity. This article describes the geographic distribution of carjacking in Campinas, Brazil, and then compares this pattern with Detroit, Michigan. We learn that raw numbers of carjackings are widely dispersed in both cities. Although these cities differ greatly in daily transportation systems, in both cities the carjacking risk estimate is greatest where motorists stop for other reasons—at gas stations or corners (Detroit); on short street blocks (Campinas). We make very limited inferences about carjacking around the world due to the limited number of cities studied thus far. However, we do suggest that carjacking reduction is most amenable to product-based forms of situational crime prevention.

Keywords

carjacking routine activities spatial patterns crime analysis crime in Brazil

Introduction

Virtually all direct-contact predatory crimes involve the convergence of at least one offender with a suitable target in the absence of a capable guardian. Carjacking is a very specific crime that enhances these routine activity requirements. That enhancement occurs because carjacking occurs so suddenly and violently within a very brief window of opportunity. Although ordinary street robberies are also sudden and violent, a street robber has a longer time period to follow a pedestrian along the road, seize the moment, and effect the attack. In contrast, a carjacker normally attacks the victim during a very brief car stop, quickly deciding to act then and there (Bernhardt & Topalli, 2016; Jacobs et al., 2003). These brief car stops occur most often at corners, traffic signals, or at gas stations; while motorists are about to enter a shop, during a pick up or delivery; or while they are getting into their cars to leave the site. To be sure, many motorists spend hours per week in their cars. Yet, carjacking only occurs within very narrow time slices while they are stopped in unfortunate circumstances. The crime must occur in the blink of an eye. Indeed, carjacking is a special routine activity application: The target vehicle may only be available to an offender exactly when the driver is stopped, still inside, with the key still in the ignition, often with the motor still running. Some offenders overcome these limitations by trapping the target vehicle between two other stolen vehicles; yet, these exceptional carjackings require more offenders and more steps to carry out. Davis (2005) discusses South African cases in which carjackers force motorists off the road with another vehicle or bump the victim’s car to cause an accident, and then attack. This method has been observed in other nations, too.

In any case, carjacking is a special crime category, requiring additional analysis beyond such offenses as burglary and street robbery. Unlike burglary, which usually occurs when victims are absent from the property, carjacking requires their presence. Unlike street robbery, carjackers have much shorter moments that they must seize. These brief moments are more likely to occur in cities. Given global urbanization and automotive use, carjackings could occur in multiple nations. Unfortunately, carjacking research is infrequent and its cross-national comparisons are incomplete. Given the limited number of studies available, carjacking is not ready for a meta-analysis. Later, we review available studies and discern some cross-national findings. However, the main purpose of the current research is to compare the phenomenon of carjacking in two different cities: Campinas, São Paulo, Brazil and Detroit, Michigan, in the United States.

Past Literature

These two cities were not chosen for any special substantive reason, but rather because data were available. However, the main scientific problem in the current international carjacking literature is that too few cities have been studied. Lacking a substantive basis for city selection, the best long-term research strategy is to extend the number and diversity of cities included in this literature in expectation that substantive patterns will emerge over time. In other words, the goal is to grow the carjacking literature. The current study adds to that research, without claiming to represent other cities around the world. Nor do we claim that these two cities each represent the nations within which they are located. It is entirely premature to draw substantial inferences about carjacking in many nations and cities based on the limited number of available studies. The current study makes no broad claims about carjacking around the world, but rather adds another piece to the larger puzzle.

Despite the sparse research on carjacking, some studies exist and can offer some guidance for what to look for when studying carjacking. We begin with a review of what is known about carjacking in Brazil, and then continue with studies from elsewhere, including the United States. We focus on available studies in Brazil, South Africa, and the United States before proceeding to the empirical analysis.

Carjacking Studies in Brazil

Madalozzo and Furtado (2011) indicate that carjackings in Brazil are well reported to police because owners must register incidents with police to activate their insurance or to recover the car or other stolen property. According to official public security data in Brazil (http://dados.mj.gov.br/), between March 2017 and July 2020, there were more than 1.2 million carjackings in the country—a mean of approximately 350,000 on an annual basis. It appears that risk of carjacking is quite a bit greater in Brazil than in the United States (Klaus, 2004).

Despite the magnitude of the problem, carjackings in Brazil are seldom studied. In the Portuguese-language academic literature, we found five studies that deal (directly or indirectly) with carjacking in four state capitals in Brazil. In a victimization survey of 20,000 city residents of São Paulo in 2003 and 2008 (Madalozzo & Furtado, 2011), about 19% of respondents were victims of carjacking or nonviolent vehicle theft. Unfortunately, the survey did not treat carjacking separately, and the prior literature does not tell us about the location of carjacking incidents within Brazilian cities.

Batista (2016) conducted an ethnographic study of the carjacking modus operandi in the Zona Leste of São Paulo, a peripheral urban zone east of the Tamanduateí River. Batista reported that carjackings are backed by organized criminals who gain most of the profit, hiring teenagers to carry out the carjacking itself. Often, two offenders on a motorcycle find the victim, one of whom approaches the victim with a gun and takes the car, which is then cut up and sold at auto parts stores.

Porcino (2015) studied carjackings in Rio de Janeiro in 2013, learning that most carjackings occurred at night, often around 8:00 p.m., with an extra peak on Thursdays. Risks were greater in those parts of a city where police were less likely to patrol. Another study of carjacking was carried out by Silva Filho et al. (2018) in Salvador, Bahia. The authors reported that, from 2004 to 2015, the growth rate for carjacking was much higher than for nonviolent car thefts. Finally, Alves (2017) conducted a study on a sample of 760 carjackings in the city of Goiânia. The results show that the cars were generally resold for 10% of the vehicle’s market value. The police identified only 18% of the offenders, two thirds of whom were minors too young to be imprisoned for the crime.

Unfortunately, reports on carjacking in the rest of Latin America are very limited. Pérez et al. (2016) reported that there had been 13,951 carjackings in Mexico, from 2011 through 2015. (One wonders whether this number was calculated differently from the Brazilian statistics.) These researchers included a survey of 750 students in five Mexican universities. The survey did not measure carjacking distinct from other auto thefts, but 13.2% of respondents had been victims of auto theft in general. Lacking additional studies from Latin America, we turn to South Africa, which is also a developing country with substantial carjacking risks.

Carjacking Studies in South Africa

James (2017) reported that 16,717 carjacking incidents had been reported to the South African Police Service during the 2016–2017 period. Then, based on a small sample (N = 280), James learned that most of these occurred near traffic lights, stop signs, shopping centers, or gas (petrol) stations. Most occurred on weekdays during the evening rush hours. Almost all involved a gun, with many offenders using additional weapons. Violent injuries were threatened and sometimes delivered, with most offenders successfully making off with the vehicle.

An earlier set of South African studies by Linda Davis focused on the rational choice behaviors of carjacking offenders. She described carjackers as rational decision makers seeking to minimize risk and maximize gains (Davis, 2001a). In her study of 100 carjacking victims, Davis (2001b) confirms findings in other nations that offenders seek to terrify their victims. Although her subsequent studies were based on interviews with only a dozen carjackers (Davis, 2002, 2003), responses uniformly verified the conclusion that carjackers are highly selective, paying attention to such situational factors as the type of vehicle driven by the motorist, the number of passengers, the offender’s fear of victim resistance, and the perceived ease of avoiding police. As elsewhere, the element of surprise was a central feature in South African carjackings.

Carjacking in South Africa is a group activity, usually executed by two to four males. A carjacker usually approaches from the driver’s side to maximize the chance to take over the vehicle, but sometimes a team of offenders approaches from each side (Davis, 2005). Carjackers sometimes use an extra vehicle with reinforcements to assist them in overcoming the victim. Hijacker teams fill three basic roles: a “pointer” (who points the weapon at the victim), the “searcher” (who searches the victim and vehicle for weapons or anti-hijack devices), and the “driver” (who drives the vehicle after it is carjacked). Although many carjackers in South Africa find unsuspecting victims in residential areas, it still appears that commercial spots enhance carjacking risk, as in other countries.

Confirming that carjacking can be a special problem in developing countries, one can compare South Africa with Australia, whose carjacking risk is far smaller. Young and Borzycki (2008) reported that only a few hundred vehicles were carjacked in Australia every year, overwhelmed by a much larger number of ordinary auto thefts. Yet, the modus operandi for these incidents was quite similar to those in South Africa. Methods include grabbing the keys from a driver entering the vehicle, ramming the victim with another vehicle, using a pistol to threaten the driver, and pulling a driver at a traffic light out of the vehicle.

Carjacking Studies in the United States

Recent press reports indicate that the changing activity patterns after Covid led to increasing numbers of carjackings, counter to decreases in several other crime types (Fies, 2020; Rasa, 2020). However, most research on carjacking in the United States preceded the Covid era. The most continuous statistical source comes from the state of New Jersey, which offered an annual carjacking report from 1993 through 2016 (Donahue et al., 1994; New Jersey Uniform Crime Reporting Unit, 2002, 2016). In 2002, New Jersey suffered 452 carjackings; the numbers declined to half that number by 2016. Carjackings were concentrated in the most urban (Northern) part of the state, and especially likely to occur after dark.

The most consequential series of academic papers on carjacking in the United States were written by Bruce Jacobs and colleagues. Spanning a 17-year period, their work began with semi-structured ethnographic interviews with 28 active carjackers in St. Louis, Missouri (Jacobs et al., 2003). Jacobs and colleagues (2003) noted that

Carjacking’s appeal is not obvious. Unlike other forms of robbery, it seldom nets cash directly and, unlike burglary, it involves a potentially dangerous confrontation with the victim. Carjacking does provide short-term transport, but on its face at least, auto theft seems a less risky way to accomplish the same thing. So why choose carjacking? Part of the answer to this question is fairly straightforward; while carjacking clearly has some drawbacks compared to other common street crimes, it also has some advantages. Unlike the typical street robbery or residential burglary, the rewards associated with carjacking—That is, the car itself—are in plain sight so there is no need to guess or “intuit” what an offence might yield. And for those who can stomach the violence, carjackings are fairly quick and uncomplicated to commit: a simple matter of overpowering the driver and seizing the vehicle (p. 676).

As noted at the beginning of this study, Jacobs and colleagues learned that carjackings are uniquely dependent on immediate action: The opportunity is gone in a few seconds, affording no time for hesitation. Carjackings may establish a sense of normalcy to lull the victim before the sudden attack commences (Jacobs, 2012). For example, the offender asks the victim for the time, to help light a cigarette, or use other methods to create a false sense of security. Carjackings usually do not involve advanced planning prior to the very short time gap between noticing the opportunity and seizing the moment. According to Topalli et al. (2015), “situational awareness” is essential for carjackers to act when the opportunity suddenly appears. Sometimes the offender finds a moment when the victim is distracted. Offenders may also use a blitz method, with several accomplices and/or firearms allowing sudden seizure of control. Copes et al. (2012) explain that the offender must quickly gain the upper hand in a carjacking situation, terrorizing the victim as necessary. Recognizing that the offender may seize the victim for a period of time, Morewitz (2019) classifies carjacking as a subtype of kidnapping.

A successful carjacker must also prevent the victim from using their own vehicle as a weapon or shield (Jacobs, 2013). Part of doing so requires convincing the victim that death is imminent if compliance is not immediate. In general, carjackers are more afraid of victim resistance and/or bystander intervention than of sanctions via the criminal justice system (Jacobs & Cherbonneau, 2018). Their overall strategy includes seeking secluded slices of time-space to carry out the crime, committing it with haste, and exploiting audience indifference by seeking moments when no one will notice in time to interfere (Jacobs & Cherbonneau, 2019).

Although several of these carjacking studies are based on small samples, they offer a strong and consistent picture of carjacking modus operandi. This overall picture is consistent with earlier descriptive work by Burke and O’Rear (1993), as well as Rand (1994). As we synthesize the various threads of evidence from different sources and different nations, it is clear that carjackings depend on haste and surprise, occur in isolated slices of time and space, and depend on circumstances allowing offenders to take overt and risky action with a good chance of success. Although we recognize that these requirements may have different manifestations in different cities, we apply similar thinking in two distinct empirical analyses.

Two Empirical Analyses

Our empirical analyses focus on how carjacking risks concentrates in small spatial units for two very different cities. As noted earlier, the main differences between these cities have to do with climate and daily transport systems.

Table 1 compares these two cities on eight dimensions. Campinas is characterized by a growing population, with a relatively higher level of prosperity than much of the rest of Brazil. Campinas is located in southeastern Brazil, approximately 500 km west of Rio de Janeiro. It is part of the relatively affluent São Paulo state. In contrast, Detroit is an impoverished city with declining population, located in what is sometimes called the “Rust Bowl” in North Central United States. It is the largest city in Michigan, a state with a declining economic base.

Table 1.

Comparing Campinas, Brazil, to Detroit, Michigan, USA.

Features of city		Campinas	Detroit
1	Prosperity relative to larger nation	High	Low
2	Population trend	Growing	Declining
3	Population size	1,213,792	670,000
4	Metropolitan area	795 km²	360 km²
5	Public transit	Extensive metro system	Low-quality public transit
6	Culture	Brazilian mix	North American mix
7	Global location	Near Tropic of Capricorn	Well north of Tropic of Cancer
8	Extreme weather	Winter rainy season	Winter cold, snow, and ice

Campinas is almost twice as populous as Detroit, covering more than twice the area, thus giving the two cities similar population densities. Yet, people tend to move about differently on a daily basis in the two cities. Campinas has an extensive public transit system. In contrast, public transit quality is low in Detroit, a city that grew with the auto industry and declined with that same industry. Yet, both cities are vulnerable to carjacking, a largely urban phenomenon that appears to apply to very different nations.

Some characteristics of these cities are compelling from a larger cultural and historical perspective. Both cities reflect African influences over a period of centuries. Although Campinas is two thirds Caucasian, its other residents are largely of mixed races, including African origins. In Detroit, the largest ethnic group is defined as “Black,” which in terms of the U.S. culture is a fusion of many African origins and substantial cross-breeding with other races, including Caucasians and American Indian races, as well as immigrants from various islands in the West Indies. After more studies are available from around the world, future literature reviews might well discern that carjacking transcends cultural and historical origins.

More relevant to carjacking are the geographic and climatic features of these cities. Although the current study neglects seasonal variations in carjacking, we note that Campinas is located near the Tropic of Capricorn, with a temperate to warm climate—except for the winter rainy season. In contrast, Detroit is cold and icy in winter, perhaps reducing the length of its carjacking “season.” However, carjacking is a problem in both cities.

Analysis in Campinas, Brazil

We consider carjackings within the urbanized portion of Campinas, Brazil, for the period 2010–2013. Certain Brazilian governmental units are especially relevant to the current research. The município is a relatively local governmental administrative unit that contains both rural and city areas, with substantial variation in population density. The city portion of a município is called an urbano. Unfortunately, these words look like the English words municipality and urban, which fail to capture the Brazilian distinction between the larger and smaller unit. To maintain clarity in this study, we borrow the Portuguese words used by Brazilians to make these local distinctions. Campinas is one of 5,570 municípios in Brazil. It is the main city of the still larger Campinas Metropolitan Region that includes 20 municípios. The urbano within Campinas is the focus of the current research.

This study started with 7,956 carjackings in Campinas that were reported to police for the period 2017 through 2019. We subtracted 1,411 incidents from the original list because they were located in rural areas or else targeted motorcycles rather than automobiles. That left us with 6,545 carjackings in Campinas for further analysis.

Thus, our study is based entirely on police data, supplemented by geographic analysis. We limited the geographic analysis to numerical forms rather than presenting maps because we discovered that carjackings were spread over much of the city, rendering maps less able to tell the story. Our analyses are based on assigning all carjackings to blocks and then measuring the lengths of these blocks using standard geographic information system (GIS) techniques. There were no GIS innovations in this study. It is a fairly straightforward analysis linking block length to risk of carjacking. We found no evidence that Campinas carjacking locations were concentrated at police stations. This may be the case for other crime types, but not for carjackings, whose victims report the crime details for insurance purposes or help the investigations track the not insured vehicle. It is possible some victims reported an incorrect address, but we cannot assess that given the data available to us.

We had hoped to relate risk of carjacking to the proximity of corner gas stations. However, we learned in a preliminary analysis that only 46 carjackings, less than 1% of the total, involved gas stations. Our Brazilian team member noted that gas stations often have CCTV cameras, which are likely to discourage carjackers. After learning that gas stations are not the main predictor of carjacking risk in Campinas, we shifted our analysis of carjacking risk within Campinas to the length of blocks.

Risk Estimates Based on Block Lengths

In Campinas, blocks differ considerably in length. To calculate block risks of carjacking, we first classified blocks by five categories of length: (a) less than 25 m; (b) 25 to 50 m; (c) 50 to 100 m; (d) 100 to 200 m; or (e) more than 200 m. Column A of Table 2 gives the number of blocks at each length, with 13,100 of the 35,362 blocks in Campinas ranging from 50 to 100 m in length. The actual lengths of blocks were totaled in Column B, indicating that the city has 3.6 million meters of street length in all. Only 56,406 m of street length were found in blocks less than 25 m long. Column C gives the number of carjackings in each of these categories. The last column divides carjacking totals by the total street meters exposed to risk and then multiplies by 1,000 to give us a “carjacking risk estimate” for the five types of exposure, based on block length.

Table 2.

Carjackings by Block Length, Total Meters per Block Category, and Risk per Block Kilometer, Campinas, Brazil, 2010–2013.

Length of each block (in meters)	A	B	C	(C / B) × 1,000
Length of each block (in meters)	No. of blocks at this length	Total meters in each block category	Carjacking raw numbers	Risk
Below 25	3,624	56,406	280	4.96
25 to 50	5,556	213,620	516	2.42
50 to 100	13,100	947,532	2,025	2.14
100 to 200	9,304	1,307,040	2,224	1.70
Above 200	3,778	1,092,654	1,500	1.37
Total	35,362	3,617,252	6,545	—

We get a very different result when calculating raw numbers (as opposed to the carjacking risk estimate using block meters in the denominators). In raw numbers, only 280 of the 6,545 carjackings in Campinas occurred within blocks less than 25 m in length. That is, only 4% of carjackings occur in short blocks. On the contrary, by far the highest carjacking risk estimate occurs in these short blocks. Dividing Column (C) by Column (B), that is, dividing 280 carjackings occurring in short blocks by the 56,406 m of street length in these blocks, we find that approximately five carjackings occur per 1,000 m. As a 1,000 m is 1 km, that gives a short block rate of five carjackings per street kilometer. In contrast, the next category of block length has half that risk of carjacking. At the other extreme, very long blocks (200 m and over in length) report a rate of only 1.37 per 1,000 m of street length (i.e., per kilometer). The shortest blocks were much riskier than the longest blocks in relative terms, despite their small number of carjackings in absolute terms.

None of these results would have been evident without using an alternative denominator, namely, the total of meters contained in each block category. If one looks at carjacking numbers alone, 87.8% occurred away from short blocks (less than 50 m in length). Only after we divided by the block length data did it become apparent that the shortest blocks were the most vulnerable in terms of a risk rate.

We arrive at a paradoxical conclusion. On one hand, the carjacking risk estimate is relatively much greater within shorter blocks. On the other hand, in terms of raw numbers, carjackings are widely dispersed throughout the city. We were hopeful that our Detroit analysis would find a sharper prediction of carjacking location.

Analysis in Detroit, Michigan

Ours is not the first study of carjacking geography in Detroit, Michigan, USA. A prior study (Lersch, 2017) was oriented less to the street network than to diverse risk factors based on multiple land use variables. That study included 14 risk factors and a total of 19,730 features. Its focus was to mathematically fit a risk terrain over the entire city rather than to identify particular features of street segments, corners, or intersections or to isolate a few variables that identify risk. A visual inspection of that author’s risk terrain map shows clearly that most carjackings occur in lines, namely, along certain streets.

Detroit police reports include a total of 5,141 carjackings from 2009 to 2015, a mean of 571 per year. We found no evidence that Detroit carjacking locations were concentrated at police stations. Of the total carjackings in Detroit, 163 (3%) occurred at gas stations. Although that number is quite small in absolute terms, these 163 carjackings help us identify heightened carjacking risk estimate at gas stations.

Risk Estimates Based on Gas Stations

We classify Detroit carjackings into four categories: (a) at a gas station on a corner; (b) at a gas station not on a corner; (c) at a corner without a gas station; or (d) elsewhere, that is, neither at a corner nor at a gas station. We shall see that the risk is by far the greatest at gas stations, whether at corners or not.

Central to our analysis is the estimate of street length in Detroit for use in denominators to help us measure the carjacking risk estimate (Table 3). We estimated that Detroit has a total of 5,064,333 m of streets, that is, 5,064 street kilometers. These street kilometers distributed among the four categories as noted in the previous paragraph. We divided Column (B) by Column (A) to give us the risk of carjacking per kilometer of street length in each of the four categories. For example, the 112 carjackings at corner gas stations were divided by 8,207 m of street length in this category, giving us a risk estimate of 13.65 carjackings per street kilometer near corner gas stations. Gas stations not on a corner had even higher risk per street kilometer: 15.6. This risk was far greater than at corners without gas stations, whereas street lengths with neither corners nor gas stations had only about one fifteenth the risk of carjacking per street kilometer. In other words, the presence of gas stations greatly multiplied the carjacking risk estimate in Detroit.

Table 3.

Carjackings Related to Gas Stations and Corners, Detroit, Michigan, USA, 2009–2015.

Location	A	B	(B / A) × 1,000
Location	Total meters in category	Carjacking raw numbers	Risk
Gas station on a corner	8,207	112	13.65
Gas station not on a corner	3,270	51	15.60
Corner without a gas station	1,525,928	1,362	0.89
Elsewhere—not at corners or gas stations	3,526,928	3,616	1.02
Total	5,064,333	5,141

Unfortunately, this concentration of risk is not reflected in the raw numbers. Of Detroit’s 5,141 carjackings, 96.8% (4,978) occurred away from gas stations. We are left with an unsatisfactory overview of carjacking for both cities. That risk is elevated consistently when using the alternative denominators. However, carjackings occurring in the most interesting categories were numerically small. In absolute numbers, carjackings in both cities are widely distributed among a variety of location types.

Discussion of Results

Carjacking risk estimates in both Campinas and Detroit tell us that carjacking risk is relatively highest in places where motorists stop their cars. In Campinas, stops in short intersections produce the greatest relative risk; in Detroit, stops at gas stations produce the greatest relative risk. However, in both cities, carjackings are distributed over all categories and are spatially distributed well beyond the highest risk categories.

This result is unsatisfying from a policy viewpoint. We were hoping to find results that would offer place-based policies to reduce carjacking. Such policies require identifying relatively few blocks or places at which most of the crimes are located. After these are identified, place-based crime reduction becomes possible. First, it might be possible to target police patrols to those places (Braga et al., 2014). Second, it might be possible to use environmental design (Armitage & Monchuk, 2019) to make those places safer. Third, one could improve local management of privately owned places to increase their security (Eck, 2018). All three of these place-based policy employs geographic crime analysis to target efforts. However, these three policy possibilities are not as useful when a crime type occurs rarely and is widely dispersed in space. Carjacking occurs far less frequently than general motor vehicle theft, and carjacking appears to be more widely dispersed in space. We see little prospect, based on current knowledge, for place-based policies to reduce carjacking risks.

Based on the discussion at the beginning of this article, we might suggest that instantaneous opportunism is required for carjacking to occur. These instants of carjacking opportunity emerge in different parts of a city at different times. Carjacking requires ideal circumstances for the offenders, and those circumstances are fleeting. Those flashes of opportunity emerge and quickly end. Those flashes occur somewhat more often in some places than other places, but their quick disappearance poses a challenge to spatial analysis. That renders traditional crime mapping of limited value for predicting carjacking places and applying policy to reduce risk at those places. It will be a challenge to future researchers to identify how, where, and when those flashes of carjacking opportunity suddenly appear and then disappear.

It is plausible that new sources of human activity data will provide more adequate denominators for predicting carjacking risks and eventually guiding place-based methods for reducing carjacking. These sources could include social media data, cellphone data, traffic monitoring, and the like (Browning et al., 2020). Another approach to improving place-based analysis of carjacking is to disaggregate types of carjacking. For example, carjackings by offenders on motorbikes might prove to be more spatially predictable than studying a combination of all carjacking types.

Another idea for studying carjacking geography in the future draws from Bowers (2014). She studied over 30,000 theft incidents from a Metropolitan area of the United Kingdom and learned that urban crimes committed outdoors require a different spatial and place analysis from those committed indoors. Criminogenic categories—such as bars—served best to predict thefts indoors, but only predicted outdoor thefts indirectly. Those establishments with high levels of indoor theft also tended to generate more outdoor theft near-by later in the night. Bowers first identified particular facilities with very high theft rates and then learned that such risk migrated to the surrounding streets as the night wore on. This led her to conclude that risky facilities act as crime “radiators,” causing additional crime in surrounding environment. Based on her work, future carjacking researchers might engage in a two-step process: first, identify that subset of gas stations, bars, and other facilities, which generate higher levels of indoor crime, and second, check whether carjackings occur near these places later in the day or evening.

If overall reduction of carjacking is not easily accomplished using focused police patrol, environmental design, or enhanced place management, how shall society reduce these crimes? For the near future, one might turn to inside-the-car devices as the most relevant carjacking prevention method. Farrell et al. (2011) documented the contribution of these techniques in reducing motor vehicle theft, noting the extra contribution of car-tracking devices for reducing motor vehicle thefts. More generally, we might think of carjacking in terms of the history of technological change relevant to the rise and fall of diverse crime types (Clarke, 2004; Ekblom, 2017; Felson & Cohen, 1980). Given the widespread use of geographic tracking in modern motor vehicles, it is at least possible that carjacking will diminish greatly in the near future with additional applications of tracking technologies. Perhaps then carjacking will become a forgotten relic of crime in the past.

Conclusion and Perspective

Four decades have passed since Ronald V. Clarke and Patricia Mayhew (1980) introduced the notion of reducing crime by shrinking the opportunity to carry it out. This family of methods is often called situational crime prevention. During those four decades, well over a thousand situational crime reduction techniques have been developed and studied by researchers around the world (Aransiola & Ceccato, 2020; Eck & Clarke, 2019; Freilich et al., 2019; Scott, n.d.; Shariati & Guerette, 2017). We can roughly divide situational crime prevention techniques into two categories: (a) place-based techniques and (b) product-oriented techniques. Place-based techniques are often identified with John Eck (2018; Eck & Guerette, 2012). Product-based techniques are more often identified with Ronald V. Clarke (Clarke & Webb, 1999; Ekblom, 2012). The two bodies of research share a common philosophical assumption—that crime events are largely situationally generated, and that criminogenic situations can be effectively mitigated (Clarke & Eck, 2005).

Despite their shared origins, the product approach focuses more on making crime targets less attractive to offenders, as opposed to rendering places less opportune for crime. From the outset, Clarke’s objective was to produce a wide repertoire of crime prevention techniques so that at least one technique would be possible to apply if others were not. In this study, we have inferred that place-based crime reduction techniques are difficult to apply to carjacking, a nearly instantaneous crime. We are not ruling out a future discovery of place-based techniques or that CCTV could be used more effectively to reduce carjacking. However, at present, we infer that product-based techniques making automobiles less susceptible to carjacking offer the best immediate solution.

The current study covers aspects of carjacking not examined in the other studies in Brazil. We were able to map carjackings within a single city and to discern patterns not previously investigated. However, the lack of similar inquiries in different Brazilian cities means we cannot yet offer a general spatial pattern of carjacking in Brazil. We offer but one piece of a puzzle that will require future work by others. However, carjacking in this city fits many of the overall attack features found elsewhere, including the Brazilian studies cited and studies in other nations.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Author Biographies

Marcus Felson is a retired professor of criminal justice at Texas State University and resides in Austin, Texas.

Silas Nogueira de Melo is an associate professor of Geography at State University of Maranhao (Universidade Estadual do Maranhão), Brazil. His recent research focuses on the geography of crime, environmental criminology, spatial analysis, and socio-environmental conflicts.

Yanqing Xu is a professor of geography at Wuhan University, China. Her current research focuses on geographic information science, spatial analysis/statistics, quantitative methods, urban geography, public health, crime geography, economic geography and planning.

Shanhe Jiang is a professor of criminology at Wayne State University. His recent research focuses on institutional and community corrections, semiformal social control, and poverty, place and crime. He is the author or coauthor of more than 120 articles, books, and book chapters.

References

Alves

V. R.

(2017). O roubo de veículo na região sul de goiânia/GO sob a Ótica da racionalidade econômica [Vehicle theft in the southern region of Goiânia/GO under the perspective of economic rationality]. Revista Brasileira de Estudos de Segurança Pública, 10(1).

Aransiola

T. J.

Ceccato

(2020). The role of modern technology in rural situational crime prevention: A review of the literature. In Harkness

Alistair

(Ed.) Rural crime prevention: Theory, tactics and techniques (pp. 58–72).

Armitage

Monchuk

(2019). What is CPTED? Reconnecting theory with application in the words of users and abusers. Policing: A Journal of Policy and Practice, 13(3), 312–330.

Batista

L. G.

(2016). O negócio que mais cresce: Notas etnográficas sobre o mercado do roubo de carros em uma periferia de São Paulo [Fastest-growing business: Ethnographic notes on the car theft market in a suburb of São Paulo]. Florestan, 5(5), 32–49.

Bernhardt

Topalli

(2016). The situational dynamics of street crime. In Cuevas

Carlos A.

Rennison

Callie Marie

. The Wiley handbook on the psychology of violence (pp. 179–194).

Bowers

(2014). Risky facilities: Crime radiators or crime absorbers? A comparison of internal and external levels of theft. Journal of Quantitative Criminology, 30(3), 389–414.

Braga

A. A.

Papachristos

A. V.

Hureau

D. M.

(2014). The effects of hot spots policing on crime: An updated systematic review and meta-analysis. Justice Quarterly, 31(4), 633–663.

Browning

C. R.

Pinchak

N. P.

Calder

C. A.

(2020). Human mobility and crime: Theoretical approaches and novel data collection strategies. Annual Review of Criminology, 4, 99–123.

Burke

T. W.

O’Rear

C. E.

(1993). Armed carjacking: A violent problem in need of a solution. Police Chief, 60, 18–18.

10.

Clarke

R. V. G.

(2004). Technology, criminology and crime science. European Journal on Criminal Policy and Research, 10(1), 55–63.

11.

Clarke

R. V. G.

Eck

J. E.

(2005). Crime analysis for problem solvers in 60 small steps. U.S. Department of Justice, Office of Community Oriented Policing Services.

12.

Clarke

R. V. G.

Mayhew

(Eds.). (1980). Designing out crime. HM Stationery Office.

13.

Clarke

R. V. G.

Webb

(1999). Hot products: Understanding, anticipating and reducing demand for stolen goods (Vol. 112). Home Office, Policing and Reducing Crime Unit, Research, Development and Statistics Directorate.

14.

Copes

Hochstetler

Cherbonneau

(2012). Getting the upper hand: Scripts for managing victim resistance in carjackings. Journal of Research in Crime & Delinquency, 49(2), 249–268.

15.

Davis

(2001a). An application of the rational choice perspective on vehicle hijacking. Acta Criminologica: African Journal of Criminology & Victimology, 14(3), 102–113.

16.

Davis

(2001b). Consequences of vehicle hijacking. Acta Criminologica: African Journal of Criminology & Victimology, 14(1), 71–76.

17.

Davis

(2002). Planning a carjacking: Positive and negative factors that are taken into consideration when hijacking a vehicle is considered. Acta Criminologica: African Journal of Criminology & Victimology, 15(1), 31–41.

18.

Davis

(2003). Carjacking—Insights from South Africa to a new crime problem. Australian & New Zealand Journal of Criminology, 36(2), 173–191.

19.

Davis

(2005). Victims of vehicle hijacking. In Davis

Snyman

(Eds.), Victimology in South Africa (pp. 228–239). Van Schaik.

20.

Donahue

M. E.

McLaughlin

C. V.

Damm

L. V.

(1994). Accounting for carjackings: An analysis of police records in a southeastern city. American Journal of Police, 13, 91–111.

21.

Eck

J. E.

(2018). Regulation for high-crime places: Theory, evidence, and principles. The Annals of the American Academy of Political and Social Science, 679(1), 106–120.

22.

Eck

J. E.

Clarke

R. V.

(2019). Situational crime prevention: Theory, practice and evidence. In Krohn

Marvin D.

Hendrix

Nicole

Hall

Gina Penly

Lizotte

Alan J.

Handbook on crime and deviance (pp. 355–376). Springer.

23.

Eck

J. E.

Guerette

R. T.

(2012). Place-based crime prevention: Theory, evidence, and policy. In Welsh

B. C.

Farrington

D. P.

(Eds.). The Oxford handbook of crime prevention (pp. 354–383).

24.

Ekblom

(2017). Technology, opportunity, crime and crime prevention: Current and evolutionary perspectives. In Crime prevention in the 21st century (pp. 319–343). Springer.

25.

Ekblom

(Ed.). (2012). Design against crime: Crime proofing everyday products (Vol. 27). Lynne Rienner.

26.

Farrell

Tseloni

Tilley

(2011). The effectiveness of vehicle security devices and their role in the crime drop. Criminology & Criminal Justice, 11(1), 21–35.

27.

Felson

Cohen

L. E.

(1980). Human ecology and crime: A routine activity approach. Human Ecology, 8(4), 389–406.

28.

Fies

(2020). Why carjackings have skyrocketed in parts of the country during the pandemic. https://abcnews.go.com/US/carjackings-skyrocketed-parts-country-pandemic/story?id=74674597

29.

Freilich

J. D.

Gruenewald

Mandala

(2019). Situational crime prevention and terrorism: An assessment of 10 years of research. Criminal Justice Policy Review, 30(9), 1283–1311.

30.

Jacobs

(2013). The manipulation of fear in carjacking. Journal of Contemporary Ethnography, 42(5), 523–544.

31.

Jacobs

B. A.

(2012). Carjacking and copresence. Journal of Research in Crime & Delinquency, 49(4), 471–488.

32.

Jacobs

B. A.

Cherbonneau

(2018). Perceived sanction threats and projective risk sensitivity: Auto theft, carjacking, and the channeling effect. Justice Quarterly, 35(2), 191–222.

33.

Jacobs

B. A.

Cherbonneau

(2019). Carjacking and the management of natural surveillance. Journal of Criminal Justice, 61, 40–47.

34.

Jacobs

B. A.

Topalli

Wright

(2003). Carjacking, streetlife and offender motivation. British Journal of Criminology, 43(4), 673–688.

35.

James

(2017). Carjacking in South Africa: Exploring its consequences for victims. Acta Criminologica: African Journal of Criminology & Victimology, 30(2), 147–161.

36.

Klaus

P. A.

(2004). National crime victimization survey: Carjacking, 1993-2002. U.S. Department of Justice Bureau of Justice Statistics.

37.

Lersch

K. M.

(2017). Risky places: An analysis of carjackings in Detroit. Journal of Criminal Justice, 52, 34–40.

38.

Madalozzo

Furtado

G. M.

(2011). Um estudo sobre a vitimização para a cidade de São Paulo. Brazilian Journal of Political Economy, 31(1), 160–180.

39.

Morewitz

(2019). Kidnapping: Carjacking and related crimes. In Morewitz

(2019). Kidnapping and Violence: New Research and Clinical Perspectives (pp. 153–169). Springer.

40.

New Jersey Uniform Crime Reporting Unit. (2002). Tenth Annual Carjacking Offense Report. State of New Jersey, Department of Law & Public Safety Division of State Police.

41.

New Jersey Uniform Crime Reporting Unit. (2016). Twenty-Fourth Annual Carjacking Offense Report. State of New Jersey, Department of Law & Public Safety Division of State Police.

42.

Pérez

C. L.

Sotelo

K. V.

Orozco

C. M. V.

Vázquez

M. E. C.

(2016). Carjacking: Victimización generada por la delincuencia organizada en Reynosa, Tamaulipas? [Carjacking: Victimization generated by organized crime in Reynosa, Tamaulipas? Perspectivas Sociales]. Perspectivas Sociales, 18(2), 107–136.

43.

Porcino

W. C.

(2015). Análise espaço-temporal dos roubos de veículos ocorridos durante o ano de 2013 nas regiões das áreas de segurança pública integrada 4 e 6 no Rio de Janeiro [Spatio-temporal of vehicle thefts that occured during 2013 in the regions or integrated public security areas 4 and 6 in Rio de Janerio]. Revista Brasileira de Ciências Policiais, 6(1), 87–115.

44.

Rand

(1994). Bureau of Justice Statistics (BJS), U.S. Department of Justice, Office of Justice Programs, & United States of America. U.S. Department of Justice Bureau of Justice Statistics.

45.

Rasa

(2020). Carjackings are up, in the latest bleak development of 2020. https://news.yahoo.com/carjackings-latest-bleak-development-2020-143000211.html

46.

Scott

M. S.

(n.d.). ASU Center for problem-oriented policing. Arizona State University. https://popcenter.asu.edu/

47.

Shariati

Guerette

R. T.

(2017). Situational crime prevention. In Teasdale

Bradley-Engen

M. S.

(Eds.). Preventing crime and violence (pp. 261–268). Springer.

48.

Silva Filho

A. M.

Zebende

G. F.

Guedes

E. F.

(2018). Análise temporal das subtrações de veículos em Salvador (BA) [Temporal analysis of vehicle disappearances in Salvador (BA)]. Conjuntura & Planejamento, 193, 47–61.

49.

Topalli

Jacques

Wright

(2015). “It takes skills to take a car”: Perceptual and procedural expertise in carjacking. Aggression and Violent Behavior, 20, 19–25.

50.

Young

L. J.

Borzycki

(2008). Carjacking in Australia: Recording issues and future directions (Trends & Issues in Crime & Criminal Justice 351). https://www.aic.gov.au/publications/tandi/tandi351